?3 ) 3. to' suppression. and in antigens developmental · 3. genome alterations and phenotypic...
TRANSCRIPT
) 3. to'
ftsnnt{Y'?3
IN CANCER AND IMMUNE SUPPRESSION.
. DEVELOPMENTAL ANTIGENS
by
B.sc. (HONS)
Ross Samuel Savvas '
.¡lhesissubmittedtotheUniversityofAdelaideinfulfilmentofthe requi"*tt" for the *"t of Doctor of Philosophy'
FebruarY 1977
$r¡r' I t\'tc\ Qlrt r"¡'"¡'i 'r": i
\
\
\
I
\
iî'1ii j" üi{îii
.'':åi:i' TI" "' u "
t
70 puD l,PmUþ^ 07 P"7vr7p7p v'r Yynuy vvul
L
t'l
TABLE OF CONTENTS
DEDICATION ..
TABLE OF CONTENTS
SUMMARY
DECLARATION
ACKNOl^lLEDGEÞlENTS
PUBLICATIONS
ABBREVIATIONS
PREFACE
CHAPTER I
SECTION I
SECTION II
Page
1
iivl'l
ix
X
xi
xlt
xtlì
I
i
t.'
1
1
2
LITERATURE REVIEt,l
General CharaÇteri sti cs of Cancer
1. The di sease
2. Tumour de'¡eìoPment
3. Genome alterations and phenotypic
expressi on
4. SummarY
Acqu'i si ti on of Foetal Propert'ies
1. Introduction
2. Cell surface characteristics and
di vi si on
3. DNA production
4. 'Foetal ' RNA
5. Enzyme Patterns
6. Devel oPmenta'l anti gens
7. SurunarY
3
5
6
6
I
I
cel J
7
7
B
8
9
9
t
l'tl
Onco-deveìoprnentaì Antigens - Foetal .:
1. Introduction
2. Foetal antigens - classificat'ion and
nomencl ature
3. Spec'ies speci f i ci tY
4. cr1-foetoprotein (AFP) ..
(a) definition
(b) carcinoma sPecificitY
(c) in the non-malignant state ..
(d) summary and conclusions
5. Carcino-embrYonic antigen
(a) definitjon
(b) human carcinoma
(c) and the non-maìignant state
(d) summary .u
Onco-deve'lopmentaì Antigens - Placental
1. Introduction
2. Carcìno-placental molecules
(a) Regan isozYme
(b) human p'lacental ìactogen
(c) human chori oni c gonadotroph'in ' '
(d) pyruvate kinase
(e) histaminase
3. Concl usions
Page
12
t2
l
)
,1
II
SECTION III
SECTION IV
13
16
18
18
1B
18
2L
22
22
22
22
24
26
27
27
27
29
29
29
29
r
tv.
SECTI0N V : Concl usi ons
SECTION VI : Introduction to Proposed Study
Page .
33
36
36
36
36
36
37
37
37
37
37
38
39
39
39
39
30
CHAPTER I I
SECTION I
SECTION II
CHAPTER IIISECTION I
MATERIALS AND METHODS
Materi al s
1. Animals .
(a) rats
(b) mi ce
(c) rabbits :.
2. Reagents ..
Methods
1. Preparation of antigen extracts
(a ) sa] i ne sol ubl e extract
(b) cel'l surface bound anti gens
(c) whol e p'lacental cel I s
2. Antisera
3. Immunol og'i caì tests
(a) immunodiffusion
(b) immunoel ectroPhoresis
MURINE FOETAL ANTIGENS
Rationa'le of tlt,-' Proposed Study 41
V
SECTION ÏI The System
1. The anti gen
2. Tumours .
SECTI0N III : Characterization of Anti-foetal Serum
SECTION IV Concl usions
1. The antibody
2. Time course studY ..
Page
42
42
42
43
44
44
48
CHAPTER IV
SECTION I
ONCO.PLACENTAL ANTI GENS
Introducti on
Rationale for the Appearance of Placental
Antigens
Saline-soluble Cytoplasmic Placental
Antigens .. ..
1. Immunodiffusion studies
2. Immunoelectrophoretic studies
Membrane-bound Placental Antigens
1. Introduction
2. Intact Placental cells
3. Placental pìasma membrane fraction
50
50
SECTION II
SECTION III
SECTION ÏV
53
53
54
59
59
59
60
64SECTION V : SummarY
vt.
CHAPTER V :
SECTION I
FOETAL ANTIGENS AND II4MUNE REPRESSION
I ntroducti on
Pagg
66
SECTION II
SECTION III
SECTION IV
Suppression of Antibody Synthesis to
Foetal Extracts
1. Rabbit antibody
2. Murine antibody
3. Conclusions
Immunogenicity of Murine Foetal Extracts
1. Cel I organe'l 1e anti gens
2. Antigen modifier
3. Conclusions
Inununol ogi caì SuPPressi on
I . Rati ona'l e .
2. Methodology
3. Results
67
67
69
73
75
75
76
7B
BO
BO
BO
B1
SECTION V : Discussion
CHAPTER VI : CONCLUDING REMARKS
ADDENDA
APPENDI CES
85
92
97
98
105REFERENCES
v]1.
SUMMARY
The malignant transformation, and the relevance of developmental
ant.igens to the cancer process, is broadly reviewed. The two
developmental antigens - foetal and placental - are then examìned
in experimental mouse and rat tumour systems'
An attempt to prepare a specific arltìserum to foetal antigens
is described. Thìs type of antiserum, reported by Stoneh-ill and
Bendi ch ( 1970) as bei ng ab'le to detect muri ne onco-foetal ant'igens '
could not be reproduced in these studies. Further examinatìon
revealed that the antìserum used by Stonehill was not detecting an
onco-foetal anti gen, but rather a ci rcul at'ing normal adul t ant'igen '
A number of attempts were al so made to demonstrate the
appearance of p'lacental antìgens in rat tumours. Although a model
vras devel oped whereby at I east 6 sal 'ine-sol ubl e carci no-p'lacental
antìgens cou'ld be detected, this system was found unsu'itable for
general laboratorY use.
The reappearance of membrane-bound placental antìgens 'in a
chemically induced hepatoma was also examjned' Ìlowever, due to the
inabiìity to produce a spec'ific anti-pìacental serum, these experiments
were unsuccessful.
During the course of these studies, it was found that rabbits
and mice were unable to raise antibod'ies to heterologous foetal
extracts unless the immunogen was freeze-dried prior to inoculation'
This unusual situation was unexpected, as lyophìlizaLion should not
si gni f .icant'ly al ter the immunogenì cì ty of the anti gens. By
viii.
experimental'ly eìiminating ail other likely possibiìities' it was
proposed that the non-freeze dried foetal extract contained a
repressor which spec,ifìcally inhjbited the production of antibodies
to foeta'l antigens. Further, it was a]so postulated that the
repressor may be a lyophilization-labile antigen-antibody complex.
The relevance of these repressor studies to foetal and t:lmour
immunology is discussed.
IX.
DECLARATION
I declare that this thesis does not contain any material which has
been submitted previously fon any degree or diploma to any university;
and to the best of my knowledge it does not contain any material
previously published or written by another person, except where due
reference is made in the text.
ss Savvas
AC KNOl^lLE DG EMENTS
I am sincereìy indebted to my supervisor, Dr. J'R. Sabine, for his
gu'idance and encouragement through this study. I would also like to
thank Professor l,l.V. Macfarlane for many hours o'f stiniulat'ing
djscussion. I am grateful to Mr. J. Zupp and the Animal House staff
for thei r di I 'igent mai ntenance of experimentaì an'imal s , and to
I'lisses D. Gibson and J. Scott for techn'ical ass'istance. Many ideas
are generated by djscussion with other laboratory members, and for
their thoughts I would like to thank Dr. R. Bais and the other
graduate students with whom I have worked.
The photograph'ic expert'ise of Mr. B.A. Pal k, the xeroxi ng work of
Mrs. B. Steedman, and the typ'ing of M'iss Pat Watson is also
appreciated. Finally, I w'ish to thank my w'ife, whose perseverance
and solid encouragement have assisted me during this research
pr0granrne .
xi.
PUBLI CATI ONS
Aspects of this study have been presented at scientific meet'ings 'in
Australia. The abstracts appear in the folìow'ing journals.
Sabine, J.R., Savvas, R.S., Murray, 4.l,l., Verma, A.K. and
Ruoslahti, E. (1975). Time course of cellular changes
during spontaneous hepatic-carcinogenesis in mice.
CLin. Eæp. Pharrn. Phas. 2 : 85.
Savvas, R.S., Bajs, R.
pì acentaì antigens' .
and Sabine,
CLin. Eæp.
J.R. (1e76).
Phøîtn. Phys.
'Carci no-
3 i 2Bl.
xt1.
AFP
BSA
CEA
DNA
FCA
FD
FIA
HCG
IEP
MeDaB
NFD
RNA
ABBREVIATIONS
q-foeto prote'in
bovi ne serum al bumi n
carci no-embryonic antigen
deoxyri bonucl ei c ac'id
Freund's complete adjuvant
freeze dried
Freund's incomplete adiuvant
human chorionic AonadotroPhin
i nmunoel ectrophores'i s
3-methyl-4-d'imethyl amino azo benzene
non freeze drì ed
ribo nucleic acid
xt11.
In this thesis, I have attempted to broadly revìew the current
concepts of cancer research, With particular enrphasis on tumour
immunology. More specifically, I have concentrated on two emerging
fields - carcinofoetal and carcinoplacental antigens.
Much of the current research is masked in a confusion of
termino'logy. Further, there has been only a small attempt to answer
the question, 'why do foetal and placental ant'igens re-appear in
tumours?' - perhaps a fundamental quest'ion ìn the understand'ing of
the cancer process. Part of the prob'lem ljes in the fact that the
foetal antigens have not been characterized in terms of a specific
non-immunol ogi ca'l functi on . The exper"imental work presented i n thi s
thesis is to some degree also gui'lty of this. However, my approach
from the start has been three-fold:
1. Demonstrate the presence of these onco-developmental antigens
in exPerimental tumours.
2. Determine at which point of t'ime folìowing carcinogen
appl i cati on theY aPPear.
3. Attempt to assign a non-immunological, biologìcaì funct'ion
to these ant'igens.
Thus, wh'ile still keeping the third area of study in mind, it was
first important to demonstrate the presence of these antigens in the
tumours, and then to answer the questìon, 'What are they doing there?''
Xi V.
In the final chapters of this thesis, I describe a new
phenomenon - that of antibody synthesis repressÍon by foetal
antigens. I say new, not because the idea is new, but its concept
and relationship to spec'ific biological problems is now presented
in the form of models as a basis for solving these problems.
l
I
I
1
G
SECTION I : GENERAL CHARACTERISTICS OF CANCER
1. The Di sease
Canceris a disease of multi-cellular organìsms. It is
characterized by uncontrolled cellular d'ivision, resulting in the
production of tumours. lrlany such cancerous grovtths are abìe to
spread to other tissues, often resulting'in death of the host due to
direct impairment of essential funct'ion, and/or wasting.
Although studies of cancer have been a'lmost exclusiveìy in
mammalian systems, the disease has also been shown to occur in
vertebrates (Halver, 1965), insects (Harshbarger and Taylor, 1968),
and plants (El Khalifa and Lippincott, 1968). In this revjew, only
mammalian cancer is considered, in order to integrate the many and
diverse studies of oncology into a broad Summary of cancer. These
studies are presented at three organ'izational levels:
(a) tumour development
(b) genetic changes which injtiate tumours and those which
result from tumour develoPment
(c) phenotypic products of these gene changes.
This thesis is principally concerned with a particular phenotypic
chafacteristic of the tumour - the reappearance' in carcinomas, of
properties normally associated with embryonic development.
il
z2
.\Ftr Et4
WAITE IN
2
2. Tumour Development - the clonal evolution of tumour cells
Before discussing any phenotypic changes which occur durìng
the malignant transformation, it is necessary to describe the
mechanism of tumour development from an initial cancerous cell.
Such a d'iscussion is warranted because many tumours share a number
of phenotypic products, and yet still maintain their indivìduaf ity.
It is the path of tumour development which may determ'ine the
characteristics of a tumour.
Nowell (1976) has formulated a model which describes the clonal
evol utì on of cancerous cel I s , and the'ir subsequent devel opment 'into
a tumour. In this model, tumour initiat'ion occurs by an induced
change in a single, previously normal , cel l. 'Neoplastic proìiferatjon
then proceeds either immediately or after a latent period. This small
number of cel 1s may escape immunol ogi ca'l surve'i I I ance due to
insufficient stimulation of the immune system (Bonmasser et aL.,1976), and
hence the number of cancer cells grows. As the mass of cells increases
some may ultimately be destroyed by the immune system. Others, however,
undergo mutation due to genet'ic instability in the tumour cell
popuìation. Although many of these mutant cells are destroyed, either
immunolog'ically or by metabolic disadvantages, one mutant may arise
which has a selective advantage and hence survjves. This cell then
becomes the precursor of a new preciom'inant sub-popu'lat'ion of cancer
cells. This sequent'ia1 selection process continues over a period of
time, with the cells becoming increas'ingly abnormal both genetìcaì'ly
and metabo'lically. However, it is important to note that thìs sequence
3
of events is not comp'leteìy random. Consequently, d'ifferent tumours
may acquire certain similarities. The divergence or tumour-specific
characteristics develop due to the effect of local cond'itions
influencing the emergence of variant sub-lines. Ultimately, the
ful ly devel oped maì i gnancy as i t appears cl i ni ca'l 'ly , has a geneti c
make-up uniquely different from the origìnal non-neopìast'ic celj.
Associated with this is an aberrant metabo'lic behaviour and "new"
ant'igeni c properti es .
3. Genomi c Al terations Durinq Transformat'ion
In the model presented above, a change in the normal cell was
required in order to initiate the neoplastic process. Except in the
case of viral transrnission, the initiator of the transformatìon has
yet to be isolated. However, it is general'ly considered that tumour
initiation occurs due to some degree of escape from nonnal controls
(e.g. local ,chalones,, hormonal, intracellular). The cause of
this "escape" may be due to an ajteration in the intracellular DNA,
e.g. a number of workers have shown that ionizing radiation, carcino'
genic chemicals, and oncogenic viruses can react with the host cell
genome in a variety of ways, in order to bring about the required
al terati ons i n gene funct'ion (El ki nd and bJhi tmore , L967; Tem'in , 1971 ;
Mi I I er and M'il I er, 1974). Phenotypi c stud'ies of f ul ly devel oped tumours
have resulted in the followìng c'ìassificatìon of genome changes
(reviewed by Anderson and Coggin, 1971):
(a) new DNA is added
(b) some DNA ìs lost
4
(c) nucleotide sequences are misread or not read
(d) DNA which is norrnal]y not read, is now coding for mRNA.
Although any one of these changes may have the potential to initiate
carcinogenesis, they may however be a product of the rselective
evolution', i.e. not necessariìy a cause of the loss of control,
but rather a secondary effect caused by the clonal evolution of the
initial cancer cells.
(a) New DNA (req:gerg:)
Due to a ,'mutation" (induced by carcinogen, vìrus, radiatjon
or by vertica'l transmission (Huebner and Todaro, 1967) ), new Dl'lA
nucleotide sequences appear and an extragenomic molecule appears.
Thus the observed phenotypìc effect is an a'ltered structuraì protein
or a neo-ant'igen.
(b) Loss of DNA
Aga.in due to a mutation, certain nucleotide sequences are
lost, resuìting 'in proteìn deletions. Consequentìy enzymes, antigens
and other molecules characteristic of the normal cell are lost.
(c) Ui:lqe9lrg-gl-rvgleellde-:esse!ee:
A'l though not speci f i cal'ly 'invol vì ng a nucl eoti de change , thi s
genetic abberation results in certain sequences being misread or not
read at all. Such a defect may be caused by producer gene expression
, be'ing al tered.
(d ) Bg:srP19::19!-gI-crgbeg:991e:
Genes which are expressed only during the embryonic
5
development of the cell, are once again 're-activatedr. These genes
begin coding for phenotypic products generally on'ly found associated
with normal embryon'ic development.
These genome changes are reflected in the phenotyp'ic alterations
known to occur in expe¡imental and human cancers. The phenotypic
profile of tumour cells reveals that there'is a loSS of certaln
"normal" cell structures and functions, and a concommitant acquisition
of 'new' cancer-character j sti c pr'operti es . Thi s i s consì stent w'ith
the genomi c prof i I e which essentì a1 ly 'invol ves a I oss or mi sread'ing
of DNA, and a gain or react'ivation of existing DNA. This loss and
acquisition of properties has been demonstrated in studies on tumour
antigens (Ingleton, t97I; Kuusel a et aL., 1965; Coll'ins and Black'
1973; Kyriazjs and Wi ssl er, 1972; Bal dwi n and Gl aves , 1972) ,
metabolism and enzymology (reviewed by Weinhouse, 1973) and cell
receptors (Sharon and Lis, 1972; Fox et aL.' I97l; Garrjdo et aL.,
re74).
4. Summary
The ultimate phenotyp'ic expression of a tumour may be influenced
by the genet'ic "evolution" of cancerous cel1s during tumour development.
Nowell's model pred'icts that tumours possess a number of tumour-spec'if ic
characteristics wh'ich have arisen due to these evolutionary processes.
The production of tumour specific ant'igens (Kyriazis and Wissler' I97L
Baldwin and Glaves, Ig72) is an examp'le of such specificity. However,
6
NowelI's model also env'isages simiIarities between a number of tumours
clue to the non-randomness of the evolution" Such sirnilarit'ies include
many common enzymes (l,.¡einhouse , Ig73) , metabol'ism features such as h'igh
g'lycolytic rates (Warburg, 1923, 1956), and the acquìsition of
embryonic-associated properties (i.e. foetal and p'lacental ). It ìs
this acquisition of embryonic-associated properties, demonstrated in
many human and experìmenta'l tumours, which may be an ìmportant event
in the sequentia'l evolut'ion, or even the initiation, of tumour
development. Alternat'ively, the acquisition of these properties may
be a consequence of the carcjnogenic state ('i.e" a secondary effect).
These two alternatiVes are exam'ined below, by consìdering foetal
properties in cancer, the non-ntalignant and healthy states.
SECTION II : ACQUISITION OF EMBRYONIC-ASSOCIATED PROPERTiES
1. Introduction
The relationsh'ip between embryonic development and neopìasia
has been a central theme of cancer research s'ince 1901' when Ehrlich
first formulated the concept of "nrisplaced cel'l rests" (see Alexander,
lg72). Embryonic and cancerous cells share many properties' These
include a number of b'ioìogical processes' ranging from DNA synthes'is
enzymes, general synthes'is of proteins, ant'igens, and of cell division.
diû+
I
t
7
2. Cell Surface Characteristics and Cell Division
The fundamental similarìty between the cancerous and
embryonic cell is their abiìity'to div'ide rapid'ly. D'ifferentiated
adult cells do not genera'liy divide, and are djffìcult to culture
in uitz,o. Ëmbryon'i c and cancerous cel I s , however, ffiâY be cul tured
with relative ease (Elkind and hlhitmore, L967).
The abiìity of cells to stop d'ivjding, when a certain cell mass
has been reached, is thought to be dependent on cell-cell surface
membrane interactions (Burger, 1973). The ab'ility of cells to bind
the plant lect'in, concanavalin A, is dependent on membrane receptors.
Becker (1974) has found that mal'ignant and foetal hepatocytes were
able to bind thjs lectin, whereas non-dividing ljver cells did not.
This differential lectin binding suggests that there may be changes
in cell surface receptors of tumour cells which are common to
embryoni c cel'l recePtors.
3. DNA Production
Stafford and Jones (tglZ) and Piddington et aL. (tglï) have
shown that thym'idine k'inase' an enzyme involved in DNA synthesis'in
the normal cell, differs to the kinase found in the cancer cell'
They have reported the presence of a foetal-specific isozyme of
thymidine kinase in human He La and KB tumour cells. This partìcular
isozyme, which catalyses the production of the DNA precursor, thymjdine
tri phosphate, i s the predom'i nant vari ant i n the devel op'ing foetal cel I ,
and is generally in very low quantit'ies in the normal cell.
I
#I
I
i
!
B
,T
ÊJ
'04
4. RNA
Aìthough there has not been any ev'idence that the enzymes
controlling mRNA synthesis in cancer cells differ to those of the
normal cell,'it has been shown that tumour cells produce a different
type of tRNA.
Gonano et aL. (I97I) have studied the tRNA species in 5123c
Morris hepatomas, and have found that these substances have different
chromatographìc characteristjcs than those of the adult liver.
Further, they have also shown that the chromatograph'ic pattern of
the tumour pheny'lalanine accept'ing tRNA (tRURphe) is the same as
the tRNAPhe isolated from rat embryonic livers (Gonano et aL,, 1973).
Similarly, Yang (1971) has reported the presence of foetal tyrosine
acceptìng IRNA in Reuber hepatoma. Again, this form of tRNAtyt was
not detectable 'in the adult liver.
5. Enz.yme Patterns
Foetal metabolism is poor'ly understood. However, it is clear
that foetal tissues contain a large number of enzymes which, although
performing the same function, vary slightly in amino acid composition
and tissue specjficity to the homologous adult enzyme. Many of these
foetal isozymes have also been found in the tumour cell. Together
with this reappearance of foetal enzyme forms, there is a simultaneous
loss or decrease ìn the activity of the adult isozyme (Knox, 1972).
The summary of a few of these foetal enzymes (Table 1.1) suggests
that there is a broad enzyme pattern established during carcinogenes'is,
i.e. there is a shift from adult to foetal isozyne synthes'is. It is
I
tII
I
t
9
I
not clear what special advantage these 'isozyres may provide to
the rnalignant cell. Maugh (7974) has suggested that they may be
better suìted to the more rap'id sustained growth of tumour and
foetal tjssues. However,'it may be equa'l1y possible that no
sel ect.i ve advantage i s ga ì ned , and the'i r appea rance , as w'i th other"
foetal properti es , i s a s'impl e consequence of the carci nogen'i c
state rather than a speci f i c necessi ty. ' ''' '
6. Developmental Antiqens
The antigenic structure of cancer cells also exhibits a
reappearance of characterjstics noimally associated wìth embryonic
development. l,lany workers have reported the appearance of antigens
which are found'in foetal cells, but not in the correspond'ing normal
adult tissues (see Sections 2 and 3). However, it is'important to
note that the presence of these antigens'in tumours ìs onìy one
exampìe of the reappearance of developmenta'l ant'igens. The other
types of developmental antigens are those whìch are found in the
placenta and tumour, but not in foetal and normal adult cells
(revìewecl by Bagshawe , 1974) -
7. Summary
From the data presented above, it can be seen that the
acquisition of embryonic properties is associated with, perhaps a
necessary association With, the malìgnant transformation' The
mechanism and reason for the re-expression of the controììing genes'
however, 'is not we'11 understood. Potter (1969, 1970) has proposed
I
10.
TABLE 1.1
SPECi FICITY
TUMOUR
REFERENCEENZYME
FOETAL ADULT
Alanine aminotransferase
Aspartate aminotransferase
Hexoki nase
Pyruvate Ki nase
DNA polymerase
Gl utami nase
Aryl esterase
M
M
H
M
T1
G
S & 1'1 Nakata et aL. (1964)
s&14 Nakata et; o.L. (1964 )
Shied et eL. (1965)
G Sato et aL. (1969)Sug'imura et aL. (1969
a,b)
Tanaka et aL. (1967)
}ve et aL. (1969)Ono & Umehara (1968)
Knox et aL. (1969)Horowitz & Knox (1968)
Mu
D
Mu
D
L
N
*
K
A
K
A
L
FaillUri e
y -CrepinI (1e73)
&
Enzymes which exhibit foetal ìsozymes during carcinogenesis'
M = membrane bound
$ = soluble, cYtoPlasmic
G = TYPe IV isozYme
tl = TYPe I, II or III isozYme
Mu = muscl e ti ssue sPec'i f i c
| = liver tissue sPecific
K = kidneY tissue sPecìfic
A = ôrYlesterase
* absent from healthy adu'lt tissues
11.
that fol 'lowì ng carc'inogen appl ì cati on , a i'ìorma'l popul ati on of cel I s
is damaged. Many of these damaged cells are replaced by "new",
rapi dìy-di vi d'ing ce'l 'l s . The cancer cel l s ari se from the di vi di ng
cel I s whi ch fai I to respond to normal d'if f erent'iati on control
mechanisms, possibìy due to some block in one or more ontogenic
pathways. Thus these cells remain essentially in the embryonic
Stage of the dìfferentiation program. Consequent'ly' cancer cells
would be expected to exhjbit a range of foetal properties, inc'luding
biochemjcal functions, antigens' and cell division propert'ies' The
information descrìbed above confjrms this expectation.
Potter's hypothesìs'is also consistent vlith the tumour, clonal
evolut'ion concepts of Nowell (1976). In this theor¡ the initial
cancer cel'l vrould exhjb'it embryonic properties, and these would
subsequentìy be passed on to future generations. During further
evolution, some of these archeo-genes would be lost due to the
"mutations" wh'ich occur. Those which rema'in act'iVe may produce
mol ecul es whi ch gi ve the tumour cel I some sel ect'ive advantage (Matlgh ,
t974; l,leinhouse, 1973). The tumour which ultimately develops, would
hence exhibit a number of embryon'ic properties, some unique to a group
of tumours and others shared by aì I tumours '
The reappearance of developmenta'l antigens in tumours is an
examp'le of such embryonic properties exhibited by a wìde range of
species. These ant'igens have been studied extensively by many
workers. It will be seen in the following sect'ion that the
reappearance of these antigens occurs 'in many tumours' This onco-
foetal associat'ion may be an important factor in understandìng the
carcinogenic Process.
t2.
SECTI()N III : ONCO-DTVELOPMENTAL ANTIGENS - FOETAL ORIGIN
1. Int roduct'ion - Tumour Cell Antiqens
Three main grori¡t:; of onco-associated ant'igens have been found
i n the ful 1.y devel oped bumour:
(i ) tumour specific or transp'lantation-associated antigens
(rnn)
(iì ) foetal antigens
( i i i ) pl acental ant'i gens .
The tumour-specific antigens differ from the other two types because
each antigen Ís specific to the particular tumour from wh'ich it has
been isolated. This type of antigen has been demonstrated jn a
variety of tumours, and is often found to be intimately assoc'iated
with the cell membrane (W'issler, 1956; Baldwi n et aL., 1971; Baldwjn
et aL., !972; Prehn, 1967).
Foetal and placental antigens may be found in cancerous and
embryonic tissue (foetus or placenta), bul: not 'in apprec'iab]e
quantitìes in normal adult cells.
In this section, the association between foetal antigens and
cancer cel I s 'is exp'lorecl by exami nì ng the nomencl ature and cl assi -
fication of these antigens, and the types of tumours which exhib'it
them
13.
2. Foetal Antiqens - Cl assi fication and Nomenclature
The first significant attempt to classify foetal ant'igens
was by Alexander (1972). His system, (Tabl e I.2),was based on three
.immunol og.ical propert'ies of the onco-foetal mol ecul e:
(i) immunogenicity in homologous and heteroìogous hosts
(i j ) cytotox.i c.i ty i nductì on i n tumour-beari ng hosts
(i'i i ) cel I ul ar I ocal i zati on '
It can be seen from Table I.2 that these properties are studied by
employing two basic techn'iques. Specifically, these are:
(a) Using heterologous antiserum to detect foetal antigens
(i ) in tumour-bearer' s serum (e.g. a-foetoprotei n (Abel ev et aL' ,
1963); carcino-embryonic antigen (Gold and Freedman, 1965) )
and tumour extracts (Stoneh'ill and Bendich, 1970). In
Alexander's system these are termed onco-foetal associated
ant'igens ;
(ii) on the cell surface membrane of tumour cells (Ishimoto ¿¿
aL., 1974; Cauchi et aL', 7974)'
(b) Examjning animals immunized w'ith either foetal extracts or cell
suspensions for their abì1ìty to resìst challenge against a
tumour forming dose of cancer cells or virus' e'g' such onco-
foetal associated transpìantation antigens have been demonstrated
by Cogg'i n et aL. (1970)'
Immunogeni ci tY
Heterologous Homologous
TABLE 1.2
CytotoxicitY Suggested NameExampìe and
ReferenceMolecules in foetusand tumour
and some normal adultti ssues
absent from al I norma'lt'issues but i nducetolerance because theYoccur in late foetallifeon'ly occur in ear'lYfoetal life, or in a
cryptic form, andhence are recognizedas foreign(a) cytoplasmic
+ 0nco-foetal associatedanti gens (OFA-assoc. )
0nco-foetal sPecificheteroanti gen(OFA-speci fi c )
co-foetal specificto antìgen (OFA-auto
p
)
e antigen
CEA (Gold andFreedman, 1965) - -^- 1\
AFP (Abelev,et F+ t teb s)
Stonehi I I andBendich (1970)
Ishimoto et aL.(re74)
Cauchi et aL. (1974)+
++ 0nau
Chemi cal l) hepatoma
-i nducedBal dwi n72)
v(
(b) plasma membrane-bound + Onco-foetal trans-p'l antati on-tY(OFA-transp'l .
++
et aL., 19
Hamster, viraltumours (Cogginet aL., 1970)
The classification of onco-foetal antigens as proposed by Alexander (lgVZ)
15.
Although Alexander's nomenclature'is specific about the
jmmunological nature of the foetal molecule, 'it does not assign
a spec'if ic functjon or role to the ant'igen under study' Further
characterization of the antigen is desirable because
(a) much of the confus'ing termÍnology could be eliminated
if each foetal antigen could be referred to by its
functional name;
(b) a knowledge of its function may provìde a clearer
understanding of the biochem'ical and physìolog'ical
processes which occur in the tumour'
spasmod'ic attempts have been made to achieve th'iS, €.9. Raftell
et aL. (1974) have detected and assigned a funct'ion to onco-
foetal ant'igens in rat tumours. Ruoslaht'i et aL. (1'976) have
demonstrated that the hepatoma-teratoma associated foetal ant'igen'
a1-foetoprotein, shares amino acid sequence homology w'ith albumin'
Thus jt may be poss'ible that alpha-foetoprotein'is a "foetal"
albumin. However, desp'ite these attempts, and a current trend
towards elucidating the functional role of foetal antigens, there
stilt remain a number of misnomers in foetal antigen terminology'
e.g. carcino-embryonìc antigen (Go]d and Freedman, 1965), is a
general term which is used to describe a specific colon carcinoma-
associated antigen.
16.
3. The Rearrpearance of Foetal Antiqens in Tumours
The reappearance of foetal antìgens in tunlours is not
restricted to one partìcular animal specìes. From Table 1.3,
it can be seen that a wide varjety of tumour types in d'ifferent
species produce these ant'igens.
It is ìmportant to ¡ote bhat one certain tumour may produce
two or three d'ifferent foetal antigens, e.9. chem'ica'l1y-jnduced
hepatomas produce cr-foetoprotein, membrane-bound and cytoplasm'ic
foetal antigens (Cauchi et aL., 7974). Converse'ly, a variety
of different tumours may produce the same foetal antìgens
(Stoneh'i11 et aL.,1971), while others have been shown to produce
only one (Banwo et aL., !974; Mukherii ancl Hìrshaut, 1973).
Thus it can be seen that the reappearance of foetal ant'igens
is a characterjstic of a large number of tumours. Perhaps the
most Studied onco-foetal ant'igens are cx1-foetoprotein and carcino-
embryonic antìgen. These antigens have been examined princ'ipally
as a di agnost j c and prognost'i c parameterin the cl i ni cal management
of cancer (rev'iewed by Laurence and Nev'ille, L972; l'laugh , 1974).
These turo a.ntigens are brief'ly djscussed in the next section in
order to clemonst.rate that the reappearance of foetal antigens 'in
the diseased state is not necessarily specific for cancer. This
observati on i s of parti.ul ul_.'importance as i ts non-speci f i ci ty
for cancer suggests that their presence may S'imply be a function
of the rapìdly-d.ividing state rather than the maf ignant
transformati on.
18.
4. a,-Foetoprotein (AFP)
(a) Defi ni ti on
AFP has been defined as the first cr-gìobulin to
appear in mammal'ian sera du¡ing development, and the dominant
serum prote'in during embryonic life (Annuaì Report of the
International Agency for Research into Cancer, 1969). The
foetal liver and yo]k sac are the primary sites of synthesis
durìng embryogenesìs. After birth, smal'l quantities of AFP are
detected only by radio-imnìunoassay, and ìts synthesis is
restrj cted to parenchyma'l I i ver cel I s (Abel ev , 1971 ) .
( u ) ôil-cl9 -scrsilere -:Peeili g1!.v
AFP is also synthesìzed by primary and metastat'ic
hepatomas (Abel ev et aL., 1963; Abelev, 797I). From Table 1.4,
it can be seen that AFP has been detected in the sera of a variety
of hepatoma-bea¡ing hosts. However, it has also been found in a
few sera of other non-hepatic carcinoma hosts (Table 1.4). From
this table it is concluded that, other than hepatonlas, only
teratomas synthesize AFP to any sign'ificant degree.
(c) App and the non-malignant state
(i) normaL sev'um uaLues
AFP has been detected in the serum of apparently
heal thy humans (Ruos'lahti and seppäl ä , I971.), mi ce (pi hko and
Ruoslahtì, 1973), and rats (Sell and Gord, 1973). The level of
AFP in these cases can only be detected by the sensit'ive radio-
immunoassay technique. However, the tumour-associated levelS may
19.
TABLE 1.4
SPEC i ES TUMOUR TYPES AFPINC I DENCE
REFERENCE
Human
Rats
Mi ce
Dogs
Monkeys
10 hepatoma
10 teratoma
other non-hepatì c
L0 hepatoma
asci tes hepatoma
transpl antable sol idnon-hepati c
1.o hepatoma
transplantablehepatoma
ascites hepatoma
non-lrepa ti c
10 hepatoma
1o hepatoma
Laurence & Neville (1972)
Laurence & Neville (1972)
Laurence & Nev'ille (I972)
Kroes et aL. (1972)
llatabe et aL. (1972)
Sell and Morris (tgz+)
Abelev et aL. (1963)
Koda et aL. (1973)
Shìnomiya et aL. (1973)
Hull et aL. (1969)
68%
45%
0.03%
n.a
n.a
a
a
a
e
a
a
a
e
n
n
n
n
n
n
n
n
Spec.ies in which the presence of AFP-bearing carcinomas have been
demonstrated.
(It is important to note that the references cited, and the specific
tumours studied, are only a representative sample of a large number
of publications in which AFP-product'ion has been demonstrated' using
a number of carcinogens to induce these tumours'
n . a. = not aPP'l i cabl e
n.e. = not examined
2t.
be measured by relatively insensitìve methods such as counter-
immunoel ectrophores i s , and immunodi ffusi on (Abel ev , 1971 ) .
(ii) rhe regenerating and eæperimentaLLy iniut'ed Liuer
Conflict'ing data has been collated concerning
AFP synthesis in the regenerating l'iver. Perova (tgAg) and
Stanislawski-Birrencwaig (1967) failecl to demonstrate the
appearance of AFP in partìally hepatectomized animals. Holever,
further investigation by Perova et aL. (tgZt) using 'immuno-
autoradiography, showed that AFP could be secreted by partially
hepatectomized rats. These results have since been confirmed
in mice by Pihko and Ruoslahti (1973).
Ruosl ahti et aL. (1973 ) have found a sharp increase in
serum AFP levels in mice with carbon tetrachloride-induced liver
injury. They have concluded that even in the notr-malignant liver,
regeneration depends on a relatively primitìve (or embryonic)
celì population which produces or passes through an AFP-producìng
stage during the differentiation phase of regeneratìon.
(i i i ) Non-maLignant diseases
Elevated serum levels of AFP have been shown to
be associated with non-cancerous hepatic diseases. Abelev (tgZi)
and Silver et aL. (1973) have shown such increases in serum hepatitis
occurring durìng the recovery phase of the disease (Ruos'lahti ei
aL., 1973). Elevated levels have also been associated with
alçoholic hepat'it'is (Meshkiqeour et aL., L974), cirrhosis and
2I.
trauma (Ruosì ahti et aL., I973; N'ishi and Hirai, 1973)'
Ruoslahtt et aL, (I974a) have shown that these AFP levels are
generally'lower than those associated with primary ìiver
carcinoma. These auLhors have further suggested that 'in some
cases of hepatic coma, caused by acute hepat'it'is, the h'igh
level nay be due to the regeneration of damaged l'iver t'issue.
These results further support Ruoslahti's concept that
regeneratìng 'liver celìs pass through an embryonic AFP-produc'ing
stage, i.e. in non-ma'lìgnant diseases, AFP productìon ìs a
f uncti on of the d j v'idi ng state.
(¿) lyrycly
There are two inrportant conclus'ions that may be drawn
from the AFP stud'ies Presented:
(i) AFP secretion is not spec'ific for hepatomas;
(ii) AFP synthesis 'is also associated with other non-
cancerous diseases.
These conclusions suggest that AFP synthesis may be a function of
the rap.idly-dìviding state (i.e. ejther due to maf ignancy or
regenerat'ion), rather than a spec'if ic property of the malignant
state.
Th.is observation has a particular signjficance in determin'ing
whether foetal antigens are a necessary pre-requisite for cancer'
or s'impìy a secondary consequence.
2?.
5. Carc'ino-emb ryonic Antiqen (CEA)
(a) Definition
ctAisdef.inedastheperch]oricacid-so]uble
gìycopr.ctein antigen fourrd in the human foetal gut between
2 and 6 months of gestatìon, but not in significant quantjties
in healthy adult gastro-'intestinal tracts'
(b) CEA and human carc'inoma
Usingthedoubleimmunodiffusiontechnique,Go]d
and Freedman (1965) found CEA'in 12 cases of colonic adeno-
carcinoma. The cEA synthesized by the tumours ìs secreted into
the serum of tumour-bearing hosts.
From Table 1.5 it can be seen that cEA secretion occurs ìn
a majori ty of gastroi ntest'i nal carci nomas . Al though 'inì ti al ly
thought to be a spec'ific product of these tumours and theìr
metastases, it may also be assoc'iated with other maf ignancìes
(Table 1.5). However, it has been proposed by warner (1976) tnat
colonic, breast and bronchial tumour cEA',s may be d'ifferent to
the others. Th'is is supportecl by e'ridence from col igan et aL'
(fgZ¡), who showed that CEA was a heterogeneous mixture rather
than a sing'le well-defined molecule'
(c) 9EA-cnd-!he-ler:[q].Isnqn!-:!q!q
l,lith the advent of the sensitive radio-jmmunoassay,
there has been a concentrated study of cEA levels in tjssues and
blood of patients with non-malignant diseases'
23.
TABLE 1.5
DISORDER N0. ¡tr CEA +\rE
PLASI4AS
/oS ITE
Gastroi ntes-ti nal tract
Breast
Respi ratorYtract
Carcinoma of:Colon and Rectum
Pancreas
Li ver
0thers
Mal i gnant
Ben i grí
Carcinoma of:Bronchus
Upper resPi ratorYtract
230/316
48/s2
72/LB
51/85
82/rsetl24
65/90
8/t5
73
92
67
60
52
4
72
53
0.3Healthy Individuals t/335
plasma cEA leveìs 'in gastrointestinal , bronch'ial and mammary
disorders. A serum is classified as CEA +ve if it contains
nore than 10 ng/ml of ant'igen'
(collated from Zamcheck et aL., Ig72; LoGerfo et aL.' t972;
Reynso et aL., Lg72; Laurence et a'L'' 1972')
24.
(i) wor:maL serum
Li ke AFP, CEA has been detected 'i n the sera of
apparently healthy adults. However, these values are very low
(.10 nglml, Laurence et aL., Lg72)' Any appearance of CEA in
the dìseased state is classified as an elevation in Serunl
quantìties above an arbitrary normal range'
(ti) cga and non-malignant diseases
CEA has also been found in the sera of patients
with non-malignant diseases (fa¡le 1.6). However, the highest
incidence of plasma CEA elevation is amongst djseases of the
gastro.intest'inal tract. The levels of cEA in these sera vary'
and are often in the range characterist'ic of colonic carcinona
(Laurence et aL., 1972).
(¿) lyryer.y
Fromthedatapresentedhere,jtcanbeseenthatCEA
is also produced in pat'ientswith a number of non-mafignant gastro-
i ntesti nal di seases , as wel I as those w'ith gastroì ntest'inal
carcinoma. This lack of specificìty for the ma'lìgnant state
suggests that the reappearance of cEA i s due to the presence of
rapidìy-dividing cells.
25;
TABLE 1.6
DI SORDER N0. 0F cEA +VE
PLASMAS
ollo
S ITE
Gastroi ntest'inaltract
Breast
Respi ratorYtract
Polyps of:Colon and Rectum
Infl ammatorY:
Ul cerati ve col i ti s 'Crohn's di sease
Di verti cul i t'i s n
peptic ulcer
Ci rrhos'is
Alcohol ic Pancreatìti s
0thers
Infl ammatorY;
PulmonarY tubercul osi s
Chronic bronchitis &
emphysema
6/67
20/95
7 /33
28/66
L7 /322/43
2t
2L
42
53
5
38
25
9
Reactìve fibroadenosì s 5170 7
Bl21
16/63
PlasmacEA]evelsinpatientswithnon'malignantdiseases
(Laurence and Nev'il1e, 1972) '
#rút
26
SECTION IV : ONCO-DEVELOPMENTAL ANTIGENS - PLACENTAL ORIGIN
1. Introducti on
The second type of developmenta'l "antigen" produced
by the cancer cell is of placental orìg.in, i.e. synthesized by
the placenta, and not foetal or healthy adult tissues' The methods
used to study p'lacental-tumour relationships d'iffer from those
used in onco-foetal antìgen studies. Foetal nlolecules have been
studied as a function of their antìgenìc'ity. Placental molecujes
have, however' been demonstrated in tumours by conventional
biochem.ical assays, rather than antigen'ic behaviour' There has
not been a spec'ific rationale for the study of onco-placental
antigens - most of these molecules were discovered because elevated
serum levels were found in tumour-bearing patients'
Theonco-placentalmo]eculessofarreportedjntheliterature
are:
(a) Regan isozYme
(b) Human chorionic gonadotrophin
(c) Placental lactogen
(d) Pyruvate kinase
(e) Hìstaminase
Thesesubstances,andtheirappearanceinhumancarcinoma'are
l"
r
discussed below.
27.
2 0nco-pl acen tal Molecules
( a ) Bggel_i:elyue-(pleçellcj-cLl:stlle-P!eePhq!i:e)
Placentala]kal.inephosphatase.isproducedby
trophoblastic cells and is secreted jnto the serum. It is not
produced by foetal or non-trophob'lastic adult cells. Its
associatìon wìth human cancer was first reported by Fishman et aL-
(1968), and has since been studied by a number of other workers
(Doellgast , Ig76; Suzuki et aL., !976; Laurence and Nevjlle'
Lg72). Regan isozyme production has sìnce been demonstrated in
a number of human tumours (fable 1.7). From Table 1.7, it can
beseenthatonlyasma]lpercentageoftumourtypesproduce
thi s enzyme. Further, the 'inc'idence wi thi n one part'i cul ar group
of tumours is low compared to CEA (c'f'Table 1'5) incidence in
colonic carcinoma, and AFP ìn hepatomas (".f. Table 1'4)'
(u) !vml-Plqeellsl-leçlgge!-(llL)
Placentalce]lsproducehumanp.lacentallactogen
durìngtheth.irdtr.imesterofpregnancy.ithasalsobeen
detected in elevated levels in the sera of patients with
trophob'last'ic tumours (Samaän et aL., 1966) and in 8% of patients
w.ith a va¡iety of non-gonadal tumours (Weintraub and Rosen' 1971)'
Aga.in the incidence of this part'icular antigen' as compared to
foetal ant'i gens , i s verY 1 ow '
driÈ
I
I
t
28
TABLE 1.7
CARCINOFIA INCI DENCEollo
I
l,
Bronchus
Mammary
Geni to-uri narY
Lymphoma
l4el anoma
Gastroi ntesti nal
7 l5L
6/ 49
10/ 5s
2/25
015
10/81
13.7
L2.24
18. 18
8.00
0
t2.35
,i,j
I
I
il'
I
Ir,
I'II
r
Regan isozyme incidence amongst cancer patients showing an
elevated serum alkaline phosphatase level (from Laurence and
Neville, 7972).
29.
(c) !stcl-sbqlielig-gsledg!rsp!il-(U99)
HCGisahormonewhichhasalsobeenfound.inthe
serum of pat'ients with teratoma, seminoma, insuloma and gastro-
intestinal carcinoma (Braunstein et aL.' 1973). It is important
tonote,however,+,hatthesynthesisofthishornronebymalìgnant
cells may be a case of ectop.ic hormone synthesjs, and not
necessarily a true reappearance of a pìacental antìgen, due to a
genera'l de-repression of archeogenes'
(d) P uvate k'inase
I
Spellman and Fottrell (1973) have examined the isozyres
ofpyruvatekinaseinplacental,foetal,adult'andfivetumour
tìssues. They found the placental-specific isozyme only ìn
extracts of 1ung, stomach and jejunum carcinoma'
(e) Histaminase
cheì-l^Je'i and Kìr'ley (1976) have shown that ascites
fluid obtained from human ovarian cancers contains a placental-
specific isozyme of histaminase. Although elevations of
histam.inase were also associated with other tumours, on]y the
ovarian carcìnoma produced the placentaf isozyme of thìs enzyme'
3. Concl usi ons
The appearance of p'lacental -spec'i f i c mol ecul es i n human
tumours supports the concept that p'lacental ant'igen reappearance
i s assoc'iated w'i th the mal'ignant state ' Hov¡ever, the l ow
r
I
30.
'incidence amongst certa'in tunlours' and complete absence 'in many
others, suggests that the reappearance of placental antigens
may not be a maior characteristic of carcinomas. Comparatively'
the association of foetal antigens and cance¡is very Strong'
The establishment of suitable animal models may provide the
necessary methodology to determine whether the appearance of
placentaì antigens is associated with cancer to the same degree
as foetal antigens.
SECTION V : CONCLUSIONS
1. Introduction
In the preceding sections, three important probìems in
tumour immunology have been ra'ised.
(a) fne current nomenclature is unsuitable and often mis-
lead'ing when discussing embryonic-associated tumour
ant'igens .
(b) There has not been a significant attempt to define the
metabolic or physiolog'ica1 functjon of foetal antìgens
i n tumours.
(c) Is the presence of these antigens a pre-requisite for
a cell to undergo mal'ignant transformation?
31.
The first quest'ion has been exam'ined by many workers, and many
meetings have been convened to exam'ine the problem (c'f' Ëirst
Conference on tmbryonic and Foetal Antigens in Cancer, Qak R'idge'
Tennessee, u.s.A., lgTI). This prob'lem may be cons'iderab'ly reduced
if a spec'ific funct'ion could be assigned to the antìgens.
The third problem is pos:'ibly of highe¡importance as it may
provide a basis for an understand'in9 of the many compiex events
which occur durìng malignant transfonnation. Essential'ly, it is
important to determ'ine whether the appearance of embryonic properties
in general is necessary before transformatìon may occur' 0r
whether these antigens appear later during the evolution of the
first cancer cell clones.
The presence of AFP in patients with non-maf ignant d'iseases,
suggests that the reappearance of this embryonìc property 'in
these cases'is probably a function of rapidly-d'iv'iding ce'l'ls
(i .e. the anti gen 'i s produced by und'if ferenti ated, embryonì c cel I s
which replace those destroyed by the disease agent). By extra-
po'lation,AFPproductionintumourslstbeproducedby
"embryonic-type" celIs wh'ich repìace adult celIs destroyed by
carcinogens. However, in this case, these rep'lacement ceìls
faìl to respond to normal differentiation control mechanìsms'
This argument has been postulated by Potter (1969, 1970)' Thus,
this proposition would support the concept that the reappearance
of embryonic properties is an 'integral characteristic of the
first cancer cell (s) which ultimately undergo a selective evolutjon'
and hence develoP into a tumour'
3?.
However, there is some evidence to suggest that' the embryonic
propert'ies may be acquired during the sequent'ia1 evolution of
the first cancer cells. Becker (L974) has shown that foetaj and
tumour cells exhibit differential lectin binding, whereas adult
and div.idìng hepatocytes did not. Thus, thìs property is not
s impty a f uncti on of the undi f ferent j ated, rap'i dly-d'i vi di ng state '
Rather it is a spec'ific characteristic of embryonìc and tumour
cells. From these results it may be concluded that this
particular onco-foetal characteristic does not necessarì'ly need
to be associated with the cells which replace those damaged by
carcinogens. If this theory'is val'id, then thìs property, and
possibly other onco-foetal characteristics, lilâY have appeared
du¡ing the sequential evolution of the first cancer cells'
The evidence in favour of either alternative, however, 'is
insufficient to produce a definjte conclusigll' It would be useful
to determìne whether there are any cancer-characteristic propert'ies
which precede and poss'ib'ly contribute to the appearance of
embryon'ic characteristics in the developing tumour. Such studies
may he]p establish a sequence of the events which occur during
tumorigenesis.
JJ.
SECTION VI : INTRODUCTION TO PROPOSED STUDY
Perhaps the fundamental questìon in experirnenta] oncoìogy
is 'What is cancer?'. At the molecular level ' cancelis often
defined as a mass of cells that have lost a number of mechanisms
which control normal cellular funct'ions. However, it has yet to
be established whether many of these lost control mechanjsms are
a prìmar"y event of carcinogenes'is ' or s'imply a secondary
consequence of the malignant transformat'ion. The appearance of
developmenta'l antigens in tumours may be an example of Such a
secondary event. The initjal obiective of the proposed study
was to heìp resolve this questìon by determinìng the role of
these ant'igens in experimental tumour systems, j,e. why do they
appear? The following quest'ions vrel"e asked in an attempt to
characterize and to determine the role of onco-developmenta'l
ant'i gens i n the ma'l ì gnant transf ormati on .
(a) At what point of time after the application of the
carcinogenìc stimulus do the developmental antigens
fi rst appear?
(b) Is their appearance corre]ated with the appearance or
d.isappearance of other parameters known to be indicatìve
of the carc'inogenic state?
(c) Do these antigens have a biological funct'ion other than
iust being antigenic?
(d) what is the celIular Iocalization of these antigens?
34.
By follow'ing the sequence of events wh'ich lead up to the first
appearance of developnrenta] antigens, it was hoped to determine
whether there was a correlation between th'is appearance and other
events which are characteristìc of the cancerous state'
The foetal antigens chosen for this study were 'tirst described
by Stoneh'iìt and Bendich (1970). They were chosen because they
could be detected by relatively simple means (ìmmunodjffusion)'
and as their appearance WaS demonstrated in 72 different tumours
of mice, this permitted theìr use in virtually any tumour system'
The p'lacental antigens of rat tumours were exarrined by raising
antisera to both saline soluble and cell-membrane bound antigens'
Before these antìgens could be studìed, it was necessary to
raise the appropriate antisera, and to detect developnlental
antigens'in the tumour systems ava'ilabte in our'laboratory'
The anti-foetal serum was not able to detect onco-foetal molecules
'in the tumours exam'ined - a dìrect contrad'iction of the work of
Sab'ine and Stonehill (tglZ). It was concluded from these studies
that stonehi I I ' s ori gi naì anti serum was probably detect'ing ant'igens
of adul t and not foetaì ori g'in. Al though the anti serum ra'ised
against saline-soluble placental ant'igens could detect these
ant'igens in rat tumours, the antigen extraction and detect'ion
methods were unsu.itable for stud.ying the probìems proposed above'
Attempts at rais'ing an antiserum to membrane bound placental
antigens were unsuccessful.
35.
The inability to produce an antiserum and a system which
could detect developmenta'l ant'igens in tumours resulted ìn the
proposed questions remaìn'ing unanswered. consequently, the
methodol ogy of thi s study was criti cal'ly re-exam'i ned, and
modifications to these techniques theoretìcally evaìuated'
As a consequence of the study on the onco-foetal antigens,
an interestìng and unusua'l observation Was made' It was found
that non-freeze-dried foetal extracts Were unable to raise
antjbodies ìn rabb'its, an effect easily reversed by lyophilization'
It was subsequent]y proposed that the foetal extracts contained
a lyophilization-labile immune repressor whìch specifìcal'ly
inhibited the product'ion of antibody to foetal antigens. This
propos'itìon was examined further and confirmed' The releVance
of these studies to tumour immunology was also discussed in the
form of a model.
36.
SECTION 1 : MATERIALS
1. Animals
(a) Bs!:
The rats used were of the Buffalo strain which had been
bred by the University's Central Animal House. Tumours of thìs strain
were obtained originaìly from Dr. Harold P. Morris of Harvard
universìty, l,lashington D.c., and subsequently ma'intained in our
Anjmal House by se¡ial transplantation into animals of the same
strai n.
Other rats used were 3 month old male and female Hooded Wistars'
also obtained from the Univers'ity's Animaì House' The rat (Wistar)
beari ng the 3-methyl -4-d'imethyì am'inoazobenzene-i nduced hepatoma
was a g.ift from l4r. K. l4atthaei of the Australian Natjonal un'iversity'
Sw1ss Albino and C3H.AvY mice were used in th1s study'
The latter were obtained originally from Dr. ld.E. Heston of the
National Cancer Institute, Bethesda, Maryland' and have since been
bred and maintained in our Animal House on a cedar-shaving bed' as
described by Sabine et aL. (igZ:).
(c) Rabb'its
All rabbit ant'isera were ra'ised in New Zealand l^Jhites'
purchased from the univers'ity of sydney veterinary Patho'logy An'imal
( Uigeb
House.
37
2. Reaqents
Co'llagenase - Sigma Chemical Co', St' Louis' Mo" U'S'A'
Sephadex G-200 - Pharmac'ia Fine chemìcals, uppsa'la, sweden'
Freund's acljuvants - Difco Laboratories, Detroit, l4ichìgan' U'S'A'
Bovine Serum Album'in - Sjgma chemicaì co., St. Louis, Mo., u's'A'
Agarose - BDH Chem'icals, Poole, Dorset, England'
SECTI0NII : METI'{0DS
1. Pre aration of anti ens
la) Saline-soluble extracts\--, -----
Soluble extracts were prepared by homogenizing the
appropriatetissuewithanequalvolumeof0.B5%saljne,inan
Ultra-turrax homogenizer at maximum speed'for 1"5 mins' The
homogenate was then centri fuged at 1500 g for 15 m'i ns, and the
supernatant dia'lysed agaìnst several changes of physiologicaì saline'
In one study with murine foetal material (Chapter III)' extracts were
first freeze-dried and then reconstituted to 15 mg solid matter per
ml, wì th d.isti t I ed water (Stoneh'il I and Bend'ich, 1970) .
(b) Cel I surface bou l9-cl!igel:
AcrudeSuspens.ionofplasmamembranesWaspreparedfrom
normal adult t'issues (heart, ìung, k'idney, spleen' liver and
intestine),byhomogen.iz.ingeachtissuewithphosphate-buffered
38
saljne jn a tefIon hontogenizer. The homogenate was cent¡ifuged at
1500 s for 30 mins and the pellet resuspended'in PBS and washed
three t'imes. Fol'lowing the f inal wash, the memþrane pel let (and some
nucle'i) was centrifuged at 105,000 g for 60 mins'
solubilìzation of the membrane-bound antigens was achieved by
the add'ition of 1 ml of 0.5% sodium deoxycholate and 1% Triton x-100
to the membrane pe'l1et and resuspendìng the peìlet' Folìowing a 5 min
incubation at 370C, the mixture was centrifuged at 105,000 g for
60 mifìSrâfìd the supernatant dialysed aga'inst PBS for 24 hours' The
presence of solubilized antigens in this extract was verified by
doubl e immunod'i ffusi on.
cellular membranes and solubiìized ant'igens were simi'la11y
prepared from rat placenta, foetus and MeDaB-induced hepatoma'
( c) I,Jhol e Pl acental cel I s
S.ingìecel.lsuspensionswerepreparedfromratplacenta
using the method detaiìed by þluller et aL. (1972), with the
fol I owi ng mod'i f i cations:
(i) The dissoc.iation medium outlined by Berry (1973) was found to
give a better yietd and viability of cells'
(ii) centrifugation at 800 g for 60 seconds was performed following
eachdìssociatjon.Eachpeìletwasresuspendedin0.5mlof
phys'io'ìog'ical sal i ne, and pool ed '
cell counts were carried out in a Neubauer haemocytometer' and viability
establ.ishedus.ingthetrypanblueexc]usiontest.
Priortoinjection,thece]lswerekilledwithl0mMsodjum
39
arsenate, and then washed three times with 0.85% sal'ine.
2. Antisera
Antisera were raised according to the schedules described in
Appendices I-V. Blood was taken from the marg'inal ear vein of
rah'bits. lulice were given live tøycobacter'íum phLei during the final
injections in order to develop an ascites fluid rich in antibodies
(Sommerville, 1967). Thjs flujd was tapped 7 days after the final
jnjection, allowed to coagulate, and centrjfuged to obtain a clear
antibody sol ution.
3 . Immunoloqical Tests
(a) Irugre9llfs:igl
This was performed on 1..5% agarose slides us'ing the double
di ffusion pattern descri bed by Quchterl ony ( 1968) . Fol I owi ng the
addjt.ion of ant'iserum and antigen (adjusted to 10 mg protein/ml ) to
the wells, the slides were incubated at room temperature for 72 hrs'
washed, dried, and stained in 0.3% Ponceau S'
(b) luuvreeles!rePhere:i:
1.0%,agarose sl'ides were used as the electrophoresing
medium. 0.04 M -do¿lrt barbitone buffer, pH 8.2, was used as the
electrode buffer, and 0.08 M sodium barbìtone, pH 8'2, as the gel
buffer . 2 v1 of antigen (concentrated to 10 mg protein/ml ) was
e'lectrophoresed for 2.25 hrs at 6 volts per cm of ge]. Care was taken
not to allow the current to exceed 20 rnAmps at any stage durìng the
run. 0n comp'letion of the eìectrophoresis, antibody was added to the'
40.
central trough and the slides allowed to incubate for 72 hrs at' room
temperature, washed, dried, and stained'in 0.3% Ponceau S.
4r.
SECTION I : RATIONAI.E OF THE PROPOSED STUDY
The b'iochemjcal and physìological events which occur in the
cancerous and pre-cancerous cell have been reported by a number of
workers in different fields. Tjme course studies before and during
the pre-cancerous stage, and ult'imately in the fully developed tumour'
have met with ljm'ited success. Such a study would be useful because
it would tead to some understandìng of the time sequence of events
which occur during the malìgnant transformation process.
The purpose of th'is study was to deternline at what stage during
the pre-cancerous stage did foetal antigens first appear' It has
been mentioned previously,that these antigens may on]y be an effect
rather than a cause of the cancerous state. To help in the
clarification of this theory, known cancer associated parameters
were examìned by other workers to determ'ine whether foetal antigens
reappeared before or after these markers of the carcìnogenic state
(SaUi ne et aL., 1975). These parameters included:
(a)AFP(Dr.E.Ruoslahti,UniversityofHe]sinki);
(b) adenyl cyclase (Professor A.l^1. Murray, Fl inders university of
South Austral i a) ;
(c) DNA polymerase II (Dr. L.A. Burgoyne, Flinders University of
South Austral ì a) ;
(d) histology (Dr. R. Rac, Institute of Medical & veterinary
Science' Adelaide) -
42
SECTION II : THE SYSTEM OF STUDY
1. The anti gen
Perhaps one of the simp]est systems of study of murìne foetal
ant'igens is that reported by Stonehil'l and Bendjch (1970)' They
produced an antiserum whìch was capable of detecting the pres'ettce of
foetal antigens 'tn 72 different tumours, from a variety of mice
strains. These onco-foetal associated antigens were chosen for this
study because such an antiserum allows a comprehensive examinat'ion
of any tumour sYstem available.
2. The tumours
We have maintained in our laboratories, a strain of mice wh'ich
have a 100% jncidence of cancer when housed under the appropriate
condj tions (Sabi ne et aL., 1973 ) . Thi s IO0% inci dence may be al tered
almost to zero if the animals are born and bred on a different beddìng'
Th'is presents an excellent test and control system for a unified study
of the malignant transformation process'
A furtherimpetus to the proposed study has been presented by
Sabine and Stonehill (Ig72). They have shown that both mammary and
'liver cancers from these tumour-prone C3H.Avy m1ce produce foetal
antigens of the stonehill type. They have also shown that these
ant.igens could be detected jn the livers of animals between 6 and 7'5
months of age, even though they were completely devoid of any obvious
signs of tumours.
IIi FIGURE 3.1
Immunodiffusion examination of unadsorbed serum from a rabbit
immunized with murine foetal extracts.
= foetal
= composite adult
= skin
= mammary tumour
= hepatoma
= unadsorbed anti-foetal serum
FIGURE 3.2
Immunodiffusion examination of exhaustively adsorbed rabbit
anti-foetal serum.
foetal
composite adult
NaCl control
skin
C3H:AvY mammary tumour
caH.AvY hepatoma
F
a
S
T
H
A
,l
l
i
¿,
Ì
{
I
i
i
IitI.
IrtII
I
F
a
c
S
T
H
3- I
H F
¡
3-2
a )
ú
C
?
I
s
þ
'iif''tj
o
a
'1,I'
FIGURE 3.3
Immunodiffusion examination of rabbit anti-mouse foeta'l extract.
This serum was given a primary adsorption with 3 mg of a composite
mouse adul t extract.
F
A
c
= foeta'l
= adul t= NaCl control (band associ ated wi th thi s we'l I i s
an artefact produced during the drying procedure)
= C3H.AvY mammary tumour
= C3H.AvY hepatoma
= skin
= Lo adsorbed anti-mouse foetal serum
T
H
S
AI
1
FIGURE 3.4
Inrnunodiffusion examination of rabbit anti-mouse foetal extract.
Thìs serum has been adsorbed twice with 3 mg of a composite mouse
adu'lt eitract.
f = foetal
c = NaCl control
. t = C3H.AvY mammary tumour
a = composite adul tþ = caH.AvY hepatoma
s = skin
t = 20 adsorbed anti-mouse foetal serum
I
i
í
I
It,
I'I
I
I
is
c'tt.s, A
ì/
^
H ì F
\
A1
I
1
I3.4
m ì ryt
"a-ffi Oa
t
3
tI
{l
!
I
a
43.
This study was desìgned to determine the poìnt of time during
carcjnogenesìs at which foetal ant'igens first reappear. This was to
be achieved by following the loss of foetal ant'igens from the neo-rlatal
liver, and the'ir subsequent reappearance in the lìvers of adult mjce
which had been maintained on the 100% tumour incìdence environment'
SECTION III : CHARACTERIZATION OF THE FOETAL ANTISERUM
Since Sabine and stonehill (Ig72) had shown that a number of
foetal antìgens reappeared in C3H.AvY hepatomas, an antibody to a
foetal extract was raised ìn our ìaboratory, in the same manner as
Stonehill's. The unadsorbed antiserum was found to react strongly
against foetal, skin, mammary and liver cancer, and a composìte
adult extract (Figure 3.1). Follow'ing exhaustìve adsorption with the
composite adult extract (i.e. 15 mg adsorbent per ml of antiserum)'
the residual serum only continued to produce double diffusion
prec.ipitati on I i nes wi th the foetal extract ( F'igure 3'2) ' Thi s res ul t
differed to that of stonehill and Bendjch (i970), who found that their
adsorbed serum reacted with skin and tumour extracts as well'
To further exam.ine thjs discrepancy, a sequentjal adsorption v'|as
performed by adding 3 mg of adsorbant to each ml of antiserum' removing
the immune precipìtate and testing the resìdual serum against the
antigens used prevjously (Figure 3.3). This adsorption and retesting
was repeated a second time (Figure 3'4)'II
I
44
Following the first adsorption, the antiserum st'ill reacted
against all the ext.racts. More importantly, the band'indicated with
an arrOw on the precìp'itation pattern has a line of identity w'ith a
band found in all the other extracts. Since there is a l'ine of
ident'ity with the adult extract, it may be concluded that it
represents a precip'itat'ion of adult antigens which may also be found
in the foetus. This type of precipìtat'ion band is similar to that
found by Stonehiì'l and his colleagues. In their system, this band
represented the tumour-associated foetal antigens (i.e' they were not
found in the adult extracts; Stonehill et aL., I97I). When a
secondary adsorption WaS performed on the Serum raised 'in our
laboratory (Figure 3.4), the arrowed band slowly faded until it
di sappeared fol ì owi ng exhausti ve adsorpt'ion ( Fi gure 3.2) .
Thus, aìthough a foetal-specific antiserum was obtained, it
failed to react aga'inst either of our two tumour extracts - a direct
contradiction of the work of Stonehill and h'is assoc'iates'
SECTION IV : CONCLUSIONS
1. The anti bod.Y
It is d.ifficult to explain the discrepancy between the results
described here and those obtained by Stonehill. There are a number of
conclusions, however, wh'ich may be drawn from the adsorpt'ion studies'
and from Stonehi 1 1 's pubì i shed materi al . These suggest that hi s
anti serum was detect'ing normal adu] t ant'i gens whi ch c'i rcul ate i n
healthy adult mice, and which are adsorbed by rapidly d'iv'id'ing cells
such as skin, tumour, and intestìne'
45.
( a) AÊs-erelier-:!s{ie:
S.ince there is a Iine of identity between a foetal
prec'ipitin band ancl the other itnmune precipìt'in bands in Figure 3'3'
it may be conclucled that the foeta'l , adult, skin, and tumour extracts
all share a comnlon antìgen(s). Further, since thjs common band
disappears from all the extracts after adsorpt'ion with adult adsorbant
(Figure 3.2), the antigens wh1ch prec'ipìtate in this band must be
normal adul t anti gens. Accord'ing to Stonehì'l 1 ' s system, antì gens
which precip'itate in th'is cross band contajn the tumour-associated
foetal antìgens. Assuming that the cross banci found in our study
detects the same ant'igen as that found by Stonehì'll and Bendjch (1970)'
and there ìs no reason to assume otherwise s'ince the same precìpitation
patterns were obta'ined, it is proposed that stonehill's antiserum was
on]y detect'ing adult antigens, and that these may have been of
jntestinal onig'in. This concl us'ion is drawn after cons'ideration of
the adul t adsorb'ing mjxtures used i n both our I aboratory and
Stonehill,s. 0ur composite m'ixture conta'ined extracts of heart, lung'
spleen, k.idney, liver and intestine. stonehill, however, did not add
intestjne to hìs adsorbant. Thus it may be that Stoneh'ill's antiserum
was detecting an antìgen(s) which is shared by both tumours and adult
intestine, i.e. a normal adult antigen(s), and one wh'ich is also
expressed in tumours-
(b) Reacti v ì tv wi th s k'i n extracts
Another possible expìanation of this d'iscrepancy l'ies in
the abil'ity of Stoneh'ill's antiserum to react against skin tissue'
Boone, commenting On stonehìII's antibody (stoneh'iII et aL" I97I)'
46
suggested that perhaps Stonehill's ant'iserum was detecting collagen
or connective tissue antigens, which, in addition to being found in
skin extracts, may also be found in tumours. in reply. Stonehill
adm.itted that the on]y dìfference he found between the tumour and
skin antjgens whjch were detected by hiS serum' was in molecular
weight. The sk'in antigen had a molecular weight of 15,000, and the
tumour antìgen, 67,000. It 'is, however, guite possjble that the
larger molecule may consist of a number of sub-unìts of the smaller
molecule. This would make the two ant'igens immunologically
indistinguishable, and hence corroborate Boone's idea that the
tumour-assocìated antigen was merely collagen' or a connective tissue-
associ ated ant'i gen.
(c) C'i rcul at'in q adul t anti S
Toexp'ìaintheappearanceofaskinantigenintumour
ce]ls, it is proposed that the skin antigen circula'tes in the tumour
bearing an.imal and ìs absorbed by the tumour cells. A similar ana'logy
exists with the Lewis blood group antigens, which are absorbed by
human red blood cells. Ev'idence to support this concept again comes
from Stonehill's own published material. Stonehill et aL' (1971)
found that freshly excised tumour cells were ant'igen positive' After
some time in tissue culture, however, they no longer produced any
anti gens capabl e of react'ing wjth thei r anti -serum' l'Jhen the cel I s
were reinjected into non-tumour bearìng mice,they aga'in began to
,produce' the antigen(s). The concept of a circulating ant'igen would
explain this phenomenon. In further corroborat'ion of this idea'
Stonehill et aL (1971), reported that on examination of healthy
mouse sera, they detected an antigen which reacted aga'inst their ant'i-
47.
serum. They dism'issed thjs as being a d'ifferent antigen than the
tumour assocìated foetal ant'igen, because it had a molecular weight
of 250,000 - 350,000. The presence of this high molecular weight
ant.igen in the serum has not been explained by stonehil'l ' It is
poss.ible that the tumour, skin and serunl antÍgens are the same' the
difference in molecular weìghts be'ing a reflectìon of the degree of
sub-unì t d'issoc'iatì on. In other words, the anti gen detected by
Stonehìll is a norma'l adult antìgen of high molecular weight. Tltis
antigen circulates in the host, and is adsorbed by rap'id]y dìv'iding
cells such as tumour, skin and intestine. Adsorption may be jn the
sub-uni t form, or al ternati ve]y, these cel I s may absorb the hi gh
molecular weìght antigen 'in the multìple aggregate form, and on
extraction of the tissue the ant'igen breaks down into the sub-unit
structure. (It is ìnteresting to note that sk'in tissue requ'ires a
more vigorous honrogenìzation than tumour, and hence a h'igh molecujar
weight ant'igen would be subiected to a greater degree of breakdown of
sub-unìt structure. This is 'in keeping wìth the observed results
that the skìn ant.igen has a molecular weight of approx'imately 15'000,
and the tumour antìgen, 67,000')
Nevertheless, whatever the explanatìon for the discrepancy between
the results presented here and those of Stonehill, 'it 'is clear that his
antìserum system could not be reproduced, and this could not be used in
th'i s s tudY.
4B
2. The time course studY
The doubtful nature of Stonehi I I 's antì gen system rai sed a
number of interest'ing points concerning the use of a heterologous
antibody. The work of Stonehìll and his col'leagues has been
extensive'ly cìted in the literature as the principal examp'le of foetal
antigen reappearance in tumours of the mouse. The'ir method, hurwêvê1"'
of inject'ing a mouse foetal extract into heterologous hosts, such as
rabbits, results in the maiority of the antibody response bejng
directed to the high concentration of adult antigens'in the extract'
The foetuses used by both Stonehill, and in the study presented here,
were of late gestation (16-19 days). A foetus of this age would be
expected to have a low foetal-specific component and a relatively h'igh
concentrat'ion of adult antìgens. Perhaps the most su'itable method of
overcoming thìs probìem would be to immunize heteroìogous hosts with
a purified foetal antigen(s). This has been done with AFP (Savvas'
unpublished results) and CEA (Gold, 1965). For this proposed study'
however, the only mouse antìgen that has been purifieci sufficient'ly
is AFP, but the levels of this were measured in Dr' Ruoslahtirs
laboratory (Sab'ine et aL.' 1975) '
Another possible substitute for Stoneh'ill rs system would be to
use membrane-bound foetal antigens as a parameter of study. This would
involve the use of early gestation foetal cells to raise an antiserum
in homologous hosts. This would ensure that there would be a relatively
high concentration of foetal antigens on the cell membrane' and by
using homologous hosts, eliminate the product'ion of antibodìes to the
49
adult antigens. Further, since membrane-bound ant'igens would be used
as the immunogen, thìs shou'ld facjljtate a strong jmmunological
reaction.
In view of the inval'idity of the Stonehill antigens, and the
unavailability of new cancerous and pre-cancerous tissue' it was
decided to abandon this parameter in the time course study' anC seek
a ne\{| system for study. carcìno-placental antigen production 'in rat
tumours was examined.
50
SECTION I : INTRODUCTION
Examples of the appearance of carcino-p'lacental nlolecules in human
carcinoma have been clescribed 'in Chapter I. The maiority of research
on these substances has been limited to the use of HCG and Regan
isozyme in the clin'ica'l management of cancer. There has been I ittle
research into the use of these substances as bio'logìcal, as opposed to
clinical, markers of the malignant transformation. Perhaps the reason
for this,ìs that these carc'ino-placental 'antigens'have so far been
found only in human tumours, and hence ethical considerations have
placed a lim'it on the expe¡imental studies that may be done.
In this chapter, a hypothesis 'is presented to exp'lain the
appearance of these antigens, and attempts to demonstrate the presence
of these antigens ìn experìmental animals are descrjbed.
SECTION II : RATIONALE FOR THE REAPPEARANCE I:i PLACENTAL ANTIGENS
An explanation for the appearance of foetal antigens in cancer
cells is summarized diagrammat'ica1]y in Figure 4.1. In terms of gene
expression, it is proposed that during the malìgnant transformat'ion
certain genes wh'ich once appeared early 'in the developmenta'l program
of the an.imal, and have since been inactive, are once again re-expressed.
Thìs gives rise to the many foetal characteristjcs known to prevail
in tumours. In arbìtrary terms, the cell has undergone a
" de-di f ferenti at'ion" .
FIGURE 4.1
The developmenta'l sequence of mammalians. The blastocyst goes
on to develop into the embryo and pìacenta. At each stage of
development, there are certain genes that are active. As
development proceeds, some genes are "swÍtched off" and new genes
acti vated.
SPERIl
ZYGOTE
BLASTOCYST
EI4BRY(]
I1ATURE ADULT
TUI'IOUR BEARING ADULT
oOVUI\1
PLACENTA
SINCE THE TUMOUR TISSUEREVERTS BACK TO AN EMBRYONICSTATE. PERHAPS THEDEDI FFERENTIATION GOESBACK FURTHER TO THEBLASTOCYST STAGE WHEREPOTENTIAL PLACENTAL ANTIGENSMAY BE FOUND
X
,l
51.
Th.is concept may also be usecl to explain the mechanism of
appearance of p'lacental ant'igens. It i s hypothesi zed that the
de-differentiation of the tumour cell goes back further, past the
embryo/foetal stage, to the blastocyst. At this stage of development,
there are four gene types wh'ich are active:
(a) blastocyst onlY
(b) pl acenta'l onlY
(c) foetal only
(d) foetal and adult
Each of these gene types code for a specifìc type of phenotypic
product. consequentìy, the placental cells which develop from the
blastocyst will synthesize all four phenotypìc products' Thus the
de-different'iation of the normal cell to the cancerous could result
ìn the reappearance of phenotypic products of all the blastocyst
gene types. It is difficult to look for blastocyst-specific products
in tumours, since th'is tissue'is so smalì in experimental an'imals
that accurate immunologìca'l stud'ies are ìmposs'ib1e' However' s'ince
foetal and placental products do appear in human tumours, there is
considerable support for this theory of de-different'iation'
In the proposed study, a d'i rect 'immunol ogi ca1 exanti ttati on of
rat tumours for the presence of placentaì ant'igens is descfiibed'
The immunologìcal approach was chosen ratl'rer than the biochemical '
because the currently known onco-placental molecules (e.g. alkaline,
phosphatase, pyruvate kjnase),may not necessarily be the placental
antigens re-expressed 'in the rat tumours. Further, s'ince the onco-
52.
placental antigens may be membrane bound, a study of cel'l-surface bound
and cytoplasmic antigens was undertaken.
The experirnents presented in this chapter have two distinct but
rel ated aims:
1. To determine whether soluble and membrane-bound p'lacental
ant'igens reappeared during carcinogenesis'
2. If so, to use these ant'igens as markers of the transition
of normal liver cells to the cancerous state, and in
particular to deter"mine the earl'iest tìme of appearance
following the application of chemical carcinogens'
Attempts to demonstrate the reappearance of these ant'igens in rat
tumours were only partìa1ìy successful. The presence of soluble
p'lacental antigens was demonstrated in three rat tumours. However,
the complexity of this system prevented its use as a means of
study'ing pl acental anti gens i n cancer.
Membrane-bound placental antigens were then examined to determine
whether these ant'igens reappeared in rat tumours'
53.
SECTION III : SALINE-SOLUBLE CYTOPLASMIC PLACEI.ITAL ANTIGENS
1. Immunodif'Fusion stud'ies
(a) Uelbsg
The methodology of th'is study was sim'ilar to that of the
foetal ant'igen program, described in the prev'ious chapter. Rat
placentae (Br¡ffalo strain) were homogen'ized 'in 0.85% sal ine, and the
soluble extract prepared as outl'ined in the Materjals and Methods.
Antibodies to this extract were raised in rabbits. Since th'is
antiserum would contain antibodies to foetal, adult and placental
antìgens, it was adsorbed fìrst with a composite adult extract and
then with a foetal extract 'in an attempt to rendelit p'lacental-
spec'ifìc. It was then planned to use this antiserum to detect
pl acenta'l anti gens 'in tumour extracts usì ng doubl e immunodì ffus i on
pl ates.
(b) Resul ts
From Figure 4.2 1t can be seen that the unadsorbed antiserum
reacted very strongly against the placenta'l, foetal, aduìt and tumour
(5123c Morrìs hepatoma) extracts. Follow'ing exhaustive adsorption wjth
a composite adult extract the serum failed to react with the adult and
tumour antigens, with only the pìacental and foetal extract exh'ibìtìng
precipìtin bands (F'igure 4.3). Further, since the foetal and pìacenta'l
'bands exhibited a react'ion of Identity (Ouchterlony' 1968 ) it seems
lìkely that the adsorbed antiserum was detectìng only the foetal
component of the placenta. Thus in th'is system, the antibody which
FIGURE 4.2
Immunodiffusion pattern of unadsorbed anti-placental serum tested
against tumour, foetal, placental and normal adult extracts'
p=p=
H=K=l=T=ap=
pl acental
foetal
heart/l ung
ki dney/spl een
I iver/intestine
transpìantable hePatoma
anti-Placental serum
= pl acental
= foetal
= heart/l ung
= kidneY/sPleen
= I iver/intestine
= Morris 5123c transplantable hepatoma
= anti-placental serum adsorbed with adult tissue
extracts
FIGURE 4.3
Irrrmunodiffusionpatternofanti-placentalserumtested
againstpìacental,foeta'l,tumourandnormaladultextracts'
P
F
H
K
L
TI
ap1
54
was raised ìn rabbits seemed to be directed aga'inst the adult and
foetal anti gens , and not aga'inst the p1 acenta'ì -spec'if i c mol ecul es '
A major consideratìon here, however, is the effect that adsorptÌon
has on the antiserum, and thìs'is considered below'
2. Immunoelectro horetic studies
(a) Rati onal e
Inthepreviousstu<ly,thereWasaposs.ibilitythat
placental-spec'ific antibodies may have been removed during adsorpt'ion
of the antìserum with adult ant'igens. This might occur, for jnstance'
if the part'icular ant'ibody jn the serum was directed agaìnst a
pìacental enzyme wh'ich has an adult isozymel Should the ant'ibody
bì nd to the antì gen.ic determi nant on both i sozymes , adsorpti on w'ith
an adult extract would remove the antiplacental activìty from the
serum. To examine this particular exampìe of adsorption effects
placental, tumour, foetal and aclult extracts were to be individually
immunoelectrophoresed. Antìgens wh'ich were found in the p'lacenta
were visualìzed by addìng unadsorbed antiplacental serum to the IEP
trough.Bycomparingtheprec.ipitinbandsofthefourextracts,it
was hoped to locate molecules which are conmon to the tumour and
placenta, but not found in either the adult or foetus' This technique
would eliminate any problems which may have been caused by adsorpLìon'
D.iscrjm'inat j on of the adul t and pl acental i sozyme shoul d occur
.because d.ifferent forms of the same enzyme often have different
el ectroPhoreti c mobi I 'i tì es '
*Such sharing of antigenic determinants has been demonstrated in the
case of ptacental and adult (intestinal) alkaline phosphatase
(Doel I gast , L976) .
55.
(b) Di rect Imnlunoel ect'ro Phoresis
Firstly, only p'lacenta'ì and foetal extracts were loaded
d.irectly on the gel and electrophoresed, 'in an attempt to determine
which of the pìacental ant'igens detected by the ant,'iserum were not
present in the foetus. A similar method has been used extensively
in the .identìfication of abnormal serum proteins (0uchterlony, 1968)'
Figure 4.4 demonstrates the unsuitabì1ìty of the system as applied here'
since there was ineffective separation of the extracts' Th'is was
probably due to steric hjndrance' caused by the'large number of molecules
which constitute the extracts exam'ined. This idea of ineffectìve
separat.ion 'is supported by the immunodiffusion patterns seen in
Figure 4.2. The precipitation bands on that plate suggest that the
unadsorbed antìserum should detect at least 6 distinct precipitation
bands when the ant'ibody is reacted aga'inst the p'lacentaì extract'
However,theìmmunoelectrophores.ispatternofF.igure4.4showsthe
presence of onìy 4 bands. (Unpublished results from our laboratory
have shown a similarly poor separation when whole placental, tumour
and foetal extracts are electrophoresed on polyacrylamide gels')
(c) Col umn Se rati on / Immunoel ectroP hores i s
(1) Method
In order to reduce the large number of molecules bejng
appl.ied to the ge]s, each extract was first fractionated on Sephadex
G-200 columns, before electrophoresis. A protein profile of the
column eluant was obtained (Figure 4'5)' and then each peak was poo'led
and adjusted to a protein concentration of 10 mg/ml' The poo'ls were
examined for the presence of placental antigens by testing aga'inst
I,ll]È
!
II
;
t
"i
FIbURE 4.4
Direct immunoelectrophoresis of pìacental and foeta'l extracts.
p
f
p'lacental
foetalill1,1
j central trough conta j ns unadsorbed anti -p'ì acenta'l serum.
II
I
þ
4-4
I
j
I
-. --.^ /^
&tiü
,tl
Ù
a'¿
I
II
;
I
I
FIGURE 4.5
Gel filtration of lrlorris transpìantable tumours, foeta'l and adult
extracts on 40 cm x 2.5 cm Sephadex G-200 col umns. Each extract was
adjusted to 50 mg protein/m1,3 mls applied to the column, and eluted
with phosphate buffered saline. 2 n\ fractions were collected and
assayed for protein by the method of Lowry et aL. (1951)' Shaded areas
represent those fractions which reacted positively agaìnst unadsorbed
anti -pì acenta'l serum.ì
i
',1
;l
I
l
V voi d vol ume
pooled fractionso
PL'P2
I
I
itr
I
lt,
I
T
I
FETAT
ADU tT
M.K 2kidney
HIGH Hb
v7794A
hepntomo
Vov5123Chepoüorro
co tu M N CHRO'I'IAT-o-GBAPHY DATA- -futvtoR - FETAT - ADutïSEPHADEX G-2OO
I
,l
I
!Fo{gz!rTl7
=l-
1
PI
FRACTION NO
56.
unadsorbed anti-placental serum, in 0uchterlony double-diffsuion p'lates.
To demonstrate that some of the antigens found in the tumour were
pì acenta'l -specì f i c, pì acenta'l -pos i ti ve peaks f rom each tumour were
immunoelectrophoresed alongs'ide the corresponding peaks from the foetal
and adult extracts (ì.e. Peak P, from each tumour, foetal and adult,
then P, from the same extracts - Figure 4.5). By comparing the precìpit'in
bands, placenta'l-specific antigens were des'ignated as those wh'ich were
found'in tumours, but not jn the foetal nor adult extracts.
(ii) ResuLts and Discussion
The precipitätion bands which formed after e'lectro-
phoresis are shown in Fìgures 4.6 (Pool 1) and 4.7 (Pool 2)- The
number of separate placental-spec'ific ant'igens (i.e.specific by the
above criteria) and their relative electrophoretic mob'ilities, are
summarized in Tables 4.1 and 4.2. Direct comparisons of e'lectrophoretic
mobilit'ies is valid because all the samples were run on the same sljde.
By comparìng the eìectrophoretic mobilities summarized in Table
4.1, it can be seen that the 5I23c and 77944 tumours share an antigen
with a common mobility and molecular vleight range (Pool 1, CPA-I and
CPA-$). This antjgen is absent, however, from the MK-2 tumour.
S'imilarly, the 5123c and MK-Z tumours share another antìgen (CPA-2 and
CPA-4), wh'ich is absent from the 7794A. In addition to these antigens'
the l4K-2 tumour also produces an additional antigen, which is not found
in either of the two hepatomas.
The electrophoretic patterns of the second pool of the tumour's
column profi'le (Table 4.2) reveal that each tumour produces a specific
FIGURE 4.6
Immunoelectrophoresis of Morris transplantable tumour, foetal and
adult extracts. Poo'l L from each of these extracts were electrophoresed
alongs'ide each other and placental antigens visualized immuno'logically
by adding unadsorbed ênti-placental serum to the central trough.
antigens founC in tumours but not adult or foetalCPA
CPA
FIGURE 4.7
Immunoelectrophoresis of l4orri.s transplantable tumour, foetal
and adult extracts. Pool 2 from each of these extracts were
electrophoresed alongside each other and pìacental ant'igens visualized
immunologically by addìng unadsorbed anti-placental serum to the
central trough
antigens found in tumours but not adult and foetal
.L'þ
z,
t1
57.
placental ant'igen wh'ich is not shared wi'L'h any of the other tumours
examjned here (Cpn-6, CPA-7, CPA-B). Such tumour specìfìcity for
developmental antigens has been described previousiy for a number of
foetal antìgens (Kroes et aL., lg72; Häkkjnen, 1974), and is
consiclered a valid conclusìon'in these studies.
Thus , based on el ectrophoreti c mob'i I i ty compari sons , 6 di sti nct
p'lacenta'l antigens have been shown to reappealin l4orris transp'lantable
tumours.
(iii) ConcLusions
Although the methodology employed here is suitable
for an initial demonstration of the presence of carcìno-placental
ant.igens, it ìs, however, unsuitable if these ant'igens are to be
monitored in a large number of samp'les. This ìs because the method of
column separatìon, followed by'immunoelectrophoresìs is a long and
tedious procedure, requ'irìng approxìmately 3 weeks to compìete per
sample. The technique also requìres a 'large quantity of tissue
(approximate'ly 1.0 gm wet weight), which is generaì]y unavailable in
the cancerous and pre-cancerous stages. Therefore, a technique was
sought wh'ich would enable':
(a) a small quantity of tissue to be utilized'
(b) simPle extractìon of tissue'
(c) detection of the antigens with relative ease'
,Membrane-bound antigens conform with these criteria' Consequentiy' thìs
typeofantìgenwaschoseninafurtherattempttodemonstratethe
reappearance of placental antigens in rat tumours'
58.
TABLE 4.1
TUMOUR ELECTROPHORETI C ÌIIOBILITY GRADIENT
ANTIGEN NO.
SLOl^l FAST
512 3c
(Li ver)
MK-2
(ri ¿ney)
77944
(Lì ver)
CPA-3
CPA. l
CPA.5
CPA-2
CPA.4
TABLE 4.2
TUMOUR ELECTROPHORETIC MOBILITY GRADIENT
FASTSLOt^l
ANTIGEN NO.
5123c
(Li ver)
MK-2
( Ki dney )
77944
(Li ver)
CPA.6
CPA-7
CPA.B
59.
SECTI0NIV:MEI4BRANE-BOUNDPLACENTALANTIGENS
1. Introductìon
l,lembrane.bound placental antìgen reappearance in rat tumours
was examined using a similar experimental approach to that previously
described for sal'ine-soluble ant'i3ens. In thìs studvr âh antiserum
was rai sed to the membrane bound anti gens, and attempts were made to
renderit pl acental -specì f i c by adsorbi ng the ant'iserum wi th adul t
tissue membrane extracts'
Brìefìy,itwasfoundthatthetwojmmunogen.s(i.e.whole
pìacental cells and a crude preparation of their cell membranes)'
stimulated the product'ion of an antiserum with activity directed
pr.incipa.llyagainsttheadu]tantigencomponentoftheplacenta,and
to a lesser degree, the foetal portion. In both cases' a placental-
specific antìbody was unobtainable. These attempts are described
bel ow.
2. Intact Pl acental Cells
(a ) lrspelcliel-ef -el!lÞedY
Placentasfrompregnantrats(Wistar)ofgestationage
13-lgdays,weremincedintosmal'lpieces.Isolatedcellswere
prepared as outlined in the lnlaterials and Methods, and antibod'ies
to the cells raised in rabb'its. Antibod'ies in the rabb'it serum were
detected by testing against antigens which had been solubìlized front
tumour (i,leOag induced hepatoma), placental and adult membranes'
60.
From Figure 4.8(a), it can be seen that the ant'iserum reacted strongìy
against all the antìgens examined.
( u ) Adsorpti on w'i th a{ul !-q!!i99!9-----L---
Antibody against adult ant'igens was removed by incubatìng
L ml of immune serum for 30 mjns at room temperature, with 50 ul of
packed, normal , adul t red bl ood cel I s. Thi s adsor'ptì on v'ras repeated
a further two times. 0n re-testing the adsorbed serum, all reactiv'ity
against the tumour, p'lacental and adult antigens had disappeared
(Fisure 4.9).
(c) Concl us i on
Sì nce al I i rrmune react'i vi ty wa s removed by adsorpt'i on
with adult red blood cells, it was concluded that all the ant'ibody
rai sed aga.inst pì acentaì cel I s was di rected agai nst e'i ther anti gens
of the adult red blood cells, or adult isozymes of p'lacental-specific
enzymes. Thjs result was sjm'i1ar to that obtained follow'ing
adsorpti on of anti serum ra'ised agai nst sal 'ine-sol ubl e extracts of
rat pìacenta (Chapter IV, Sect'ion III)'
3. Placental Plasma Memb rane Fraction
The previous experiments were repeated with an ant'iserum
raised aga.inst a crude preparation of placental cell-surface
membranes, rather than whol e cel I s '
Heterol eeev:-s!!iÞsdv(a)
(i) Pxeparation of antibodY
FÏGURT 4.8
Immunodiffusion s'lide of antiserum from rabbits immunized with whole
placental cells. The peripheral wells contain the test antigens, ì.e.
ahtigens solubilized off the appropriate tissue cell membranes, and the
central wells contain the antiserum.
(a) P
N
H
= p'lacental
= NaCl control
= heart
A=
[tl = NaCl control
| = ìung
I = intestine
anti-whole placental cells
(b) T
c
S
A
tumour
NaCì control
spl een
Anti serum
A=
Jrl =
(=
Li=
NaCl controi
kidney
liver
unadsorbed anti -whol e pl acenta'l ce'l I s
4-8a
4_ gb
N
eì
1ii
;l
I)î
o
II
¡
þ
IIGURE 4.9
Immunodiffusion slide of antÍserum from rabbits immunized with whole
placental cells. The serum had been adsorbed three times with packed rat
red blood cells before testing. The peripheral wells contain the test
antigens, i.e. antigens solubilized off the appropriate tissue cell
membranes, and the central weìls contain the adsorbed antiserum.
(a) = placental N - NaCl negative control
= NaCl negati ve control L = I ung
=heartl-intestine= adsorbed anti-whoìe p'lacental ce'lls (concentrated x10)
(b) tumour [rl = NaC] negati ve control
NaCl negative controt K = kidney
spleen Li = 'liver
adsorbed anti-whole placental ceìls (concentrated x10)
P
N
H
A
T
c
s
A
4,ga
4.9b
Na
¡
I
I
TI
Nt lt¿
N s
61.
Washed pìacenta'l membranes were prepared from 13-19
day gestational pregnant l^listar rats. Rabb'its were inttnunized with
thi s membrane preparat'ion , and the presence of anti bod'ies aga'in
verified by testing the rabbit serum agaìnst membrane solub'ilized
MeDaB hepatoma, p'lacental and adult antigens. The antibody reacted
strongly aga'inst al'l the anti gens exanl'ined by doubì e immunodi f f usi on
(Figure 4.10) .
(ii) adsorption uith aduLt antigens
1 ml of immune serum was adsorbed w'ith 50 ul of
packed normal adult red blood cells as described prev'iously' The
adsorpt.ion markedly reduced the i ntensi ty of precì p'itat'ion of the
antìgens by the immune serum (Figure 4.11). However, a residual anti-
adul t acti vi ty st'iì 1 remained i n the serum'
1 ml of this antiserum was then further adsorbed with a composÍte
mixture of adult tissue membranes. 0n repeated adsorption, all
reactivity with the adult and tumour disappeared from the antiserum
(Figure 4.I2). The resultant serum still reacted with the placenta'l
extract, although the degree of activ'ity was cons'iderabìy dim'inished.
This adsorbed antiserum was tested to determine whether the renraìn'ing
antibody was placental-specìfic, or simply directed agaìnst the foetal
component of the p'lacental cell membrane. This was done by testìng
the residual antiserum by 'immunodiffusion, against a solubil'ized
membrane extract from hlistar rat foetus. This antiserum was able to
precipitate antigens from the rat foetus (Figure 4.13). Further,
since the precipit'in bands of the placentaì and foeta'l antigens
FIGURE 4.10
Immunodiffusìon slide
membrane fractjon prepared
contajn the test antigens,
tissue cell membranes, and
anti serum.
of antiserum from rabbits immunized with a cell
from rat placental cells. The peripheral wells
i . e. anti gens sol ubi I i zed off the appropri ate
the central we]ls contain the adsorbed
(a) P
N
H
= placental N - NaCl negative control
= NaC'l negati ve control L = ì ung
= heart l= intestine
A = unadsorbed anti-placental cell membranes
=spleenK=kidney= NaCl negati ve contro'l Li = I i ver
= tumour
= NaC'l negati ve control
A = unadsorbed anti-placental ce'll membranes
(b) s
N
T
c
a
N
I
4.10a
4-l0b
i$)
a
II
I a
N\ O
a
a
Na
I
-
a
FIGURE 4.11
Immunodiffusion slide of antiserum from rabbits immunized with a cell
membrane fraction prepared from rat placental cel'ls. The serum had been
adsorbed three times wìth packed rat red blood cells prior to testing.
The peripheral wells contain the test antigens, i.e. antigens solubilized
off the appropriate tissue cel'l membranes, and the central we'lls contain
the adsorbed antiserum.
(a) = placental N = NaCl control
= NaCl control | = lung
= heart I = int,estine
A = RBC adsorbed anti-placental membranes
(b) = spl een lrl = NaCl negati ve control
= NaCl negative control K = kidney
= tumour Li = liver
A = RBC adsorbed anti-pìacental membranes
P
N
H
S
N
T
4-lla
N
Ë¡ I
t
a
4 b
N
N
-ll
a
IL
¡
a
aN
FIGURE 4.12
Immunod'iffusion slide of antiserum from rabbits immunized with rat
placental cell membrane fraction. The antiserum had been adsorbed
three times, with packed rat red blood cells, and then three ti¡nes with
a composite mixture of membrane from normal adult tissues. The
peripheral wells contain the soiubil'ized test antigens, and the
central wells the adsorbed ant'iserum.
ti
h
p
S
k
= pl acenta
= spìeen
= lung
= tumour
= i ntesti ne
= heart
FIGURE 4.13
Immunod'iffusion slide of antìserum from rabbits immunized with
placenta'l cell membrane fraction. The antiserum has been adsorbed
with rat red blood cells and a composite adult membrane suspension.
P = pì acentaì
f' = foetal
4-12
hI
a
Ða
t
4.13
P,¡
\ol
)f
p
ro
t
,lp\,f'"JT tI
o
I
o
62.
exhibited a reaction of identity, the antiserunl activ'ity was
exclusiveìy directed aga'inst the foeta'l component of the placenta'
(iii) ConcLusion
The results obtained here ind'icate that the
antiserum which was raised against the p'lacental membrane fractìon
was directed aga'inst the foetal and adult (or adult ìsozyme of a
placental enzyme) component on'ly. Th'is effect is agaìn similar to
that which occurred durìng the preparatìon of an antibody to'saline-
soluble extracts of the rat p'lacenta. consequent'ly, thjs method of
raising an ant'ibody was also unsuitable. A method of eliminatìng
the production of antibody to the large quantity of adult antigens
was thus sought.
(b) Homol ous anti bo dy
(t) Introduction
The attempts to raise an antibody to pìacenta'l-specific,
membrane-bound ant'igens had been unsuccessful so fat" since most of
the immune response was directed aga'inst the high concentration of
adult antigens on the cell surface membrane, and to a jesser degree
against those of foetal origin. In an attempt to overcome this problem'
rats were immunized with homologous pìacental nlembranes. s'ince the
rats shou'ld not recogn'ize the adult antigens as foreìgn, the ant'ibody
response to these antigens should be elìminated. This would allow the
product.ion of pl acental and foetal -spec'if i c antì sera. The amount of
antibody to these antÍgens should also be cons'iderably higher because
63.
the'immunol ogicaì response to the adult antigens (principaì'ly, the
highly immunogen'ic transplantation antigens), which would normally
"swanrp" the host's immune system, would be el imirrated.
(ii) ResuLts and discussion
The attempt to imnlunjze rats with homo'logous placentaì
membranes Was unsuccessful. The sera from the 12 immunized rats
fai I ed to react agai nst the sol ub'il i zed p'lacental membrane extracts
(F'igure 4.14). This inabi'lity to raise an antibody to the p'lacental
antigens was probab'ly due to either the small quantity of p'lacental-
specìfic antigens which constjtute the placental cell membrane, or
to the host's immunolog'icaì tolerance to these antigens. In the
first circumstance, this could possibly be overcome by cont'inuaì
jmmunization over periods of 6 months. Chang et aL. (1976) have
overcome thìs problem in such a manner, and have successfully
rai sed, i n rats , an anti body to two so'l ubl e rat p'lacenta'l enzymes .
The second prob'l em of immunol ogi cal tol erance woul d be d'i ff icul t to
sol ve. Thi s woul d requ'ire the formati on of p1 acentaì ant'igen
haptens, to facilitate their recognition by the rat's 'immune
system. A sim'ilar method has been used by Ruoslaht'i et aL. (1974)
to break homol ogous tol erance to AFP . |l|'ith pl acental ant'i gens ,
however, many technical difficultjes prevail, since a pure so'lution of
placentaì-specific antigens is not available. Any attempt to haplen'
an 'impure mixture of placental, foetal and adult antigenS' would
probably result in the product'ion of antibodies to the adult ant'igens
. as wel1, since adult antigens would be haplenated to such an extent'
ji¡
ilr-r
FIGURI 4.14
Immunodiffusion slides of sera taken from [^listar rats immunized
with homo'logous rat pìacental membranes. The peripheral wells
contain the antisera and the central wells contaìn an antigen extract
prepared from rat placental cell membranes
p = placental membrane antigens
1 ... 72 = rat sera numbers
I
4.14 a
3
-ñ.
trl46
a
a
t4-l4b
I
9
loi
a
f.,
..t
',j
iÌ
64
that i mtnunol ogi ca'l tol erance to these woul d be broketl , resul ti ng 'in
the unwanted production of ant'ibodies to adult molecules.
SECTION V SUMMARY
The aim of thjs study had been to
(a) demonstrate the appearance of placental antigens in experimental
rat tumours;
(b) to use these antigens in a time course study as markers of the
mal i gnant state.
The first aim has been achieved. However, the second has not been
attenrpted because of the inability to raise an anti-placental serum,
which could be used to detect placental antigens in cancerous tissues.
In the majority of attempts to obtajn such a Serum, two'important
observations were noted:
(i) adsorpt'ion of an anti-placental serum with adult tissues
removed most of the antibody activjty;
(ii) any residua'l activity after such adsorption was directed
against the foetal component of the placenta.
The first observat'ion was examined further and it was found that the
unadsorbed ant'iserum to saline-solubìe pìacental extracts could be
used to detect p'lacental ant'igens in tumour extracts. However, the
particular antigens against which this activity was d'irected probably
.T
¡,{
II
I
I
tII
I
!
nlì,Ë
,.1
65
had adult jsonlers, hence their detect'ion with an absorbed antiserum
was not possible because absorption with adult extracts removed the
antibodies to these antigens. Although a similar s'ituat'ion was not
shown in the membrane stud'ies, this may have occurred here also'
Thus, although placental antigen reappearance in rat tumours was
demonstratecl by indirect means, it was not possible to raise an
antibody to a placental-specìfìc antigen (j.e. one without a similar
adult form). There nray be two reasons for the jnabiljty to do this:
(a) they do not exist;
(b) aìtlrough synthes'ized by placental cells, they ex'ist in
small quantìties and are unable to raise a sufficient
ìmmune response from the inocculated host'
The complete absence of such molecules would linlit the use of placental
ant.igens i n experimental studi es of the mal 'ignant transformati on.
such studies would emp'loy placental molecules which also have an adult
form, e.g. the Regan ìsozyme, HCG, etc' of humans'
If the presence of placental-spec'ific antigens is to be verified,
however, the methods used'in these studies requ'ire considerable
modifications. The alterat'ions to this methodo'logy are revjewed
furtherin ChaPter VI.
i
I
{
t
66.
SECTION I : INTRODUCTION
Consìderable difficuìty was observed in raisìng ant'ibody to a
1500 g extract of mouse foetal homogenates (see Chapter^ III). This
difficu'lty was overcome by freeze-drying the foetal material, and
reconstituting prior t.o immunizat'ion - an unusual situation because
freeze-drying is not normally associated wìth any change in the
immunogenicìty of an antigen.
To exp'lain this unusual phenomenon, the presence, in the foetal
extract,of a'lyophilizatjon-lab'ile ìmmune repressor was proposed.
A'lthough this possib'il'ity was initia'l'ly thought to be unl i kely, the
importance of this find'ing was such as to warrant further invest'igation.
Duri ng th'i s study:
(1) the ìmmunogenìc'ity of the foetal extracts was examined;
(2) the'inabilìty to rajse antibodies to foetal extracts vras
a'lso repeated in mice immunized with non-freeze-dried rat foetal
extracts; and
(3) it was found that th'is repressìon of ant'ibody synthesis
was specific for foetal antigens on'ly.
From the data presented, it has been proposed that the inab'ilìty to
raise antibod'ies to non-freeze-drìed foetal extracts was due to the
act'ion of a'lyoph'ilization-labile repressor' possibly an ant'igen-
antibody complex.
I
67.
SECTION II : SUPPRESSION OF ANTIBODY SYNTHESIS TO FOETAL EXTRACTS
1. Rabbit Immunjt.y
(a) Ait_cld Methods
During the attempts to raise an antibody to mouse foetal
extracts, with wh'ich foetal antigens of murine tumours could be
detected (Stonehill and Bendich, 1970), two groups of rabbìts were
immunized - one with non-freeze-d¡ied (NFD) foetal extract, and the
other with freeze-dried and reconstituted (FD) extract.
Inrnunizatjons were given to the animals via a number of iniection
routes, using three different concentrations of antigen (faUles 5.1
and 5.2). Ant'ibody production was monitored by double immunodìffusion,
using non-freeze-dried foetal extracts as the precìpitating antigen.
(b) Res ul ts
The raisìng of antibodies to non-freeze-dried foetal
extracts is summarized in Table 5.1. All attempts, usìng a variety
of inoculation protocols, and antigen concentrations, were unsuccessful.
In contrast to these results, antibody product'ion was induced in the
second group of animals which had been g'iven freeze-dried extract
(Table 5.2). In this case, the response varied but was present in
all animals examined.
The d'ifference'in the responses of the two groups of animals is
difficult to expla'in since both groups apparently received the same
ant'igen, and in four cases, in comparable concentrations. It is
unusual that freeze-drying of the antigens should alter the
'immunogenicìty of the extract, since ì.yophiljzation is a standard
b'iophysicaì technique which rare'ly al ters the chemical 'integrity of
6B
TABLI 5. 1
Immune responses of rabbits inoculated with murine
foetal extracts
1 2 3RABBIT NO. ANTIGEN CONC. S ITE RESPONSE
1
2
3
4
5
6
7
I9
10
I3
3
10
10
3
3
8
3
3
s.c./i.m.i.v./s.c./i.m.i.v./s.c./i.m.
fpfplm1m
s.c./i .m.
s.c./i.m.s. c. /i .nl.
2the route of administration of the antigen
where multiple s'ites were used, the extract was divided into equal
Rabbits were immunized with various concentrations of non-freeze-dried
murine foetal extracts and their ability to raise antibodìes to the
extract examined. Protocols are given ìn Appendix IV.
lmg protein/ml Per iniection
s.c. = subcutaneous
i.v. ='intravenous
i.m. = intramuscular
f.P. = footpad
porti ons .
3The humoral immune response was measured by testing the rabbit serum
for antibodies in immunodiffus'ion pìates. The antigen used was non-
freez-e-dried mouse foetal extnact, at a concentratjon of 10 mg prote'in
per m'l .
69.
any molecule
It was decided to examine this unusual situation further, and
in particular to see if these results could be repeated usjng other
an'imal speci es .
. Muri ne Immun i t.Y2
(a) A'im and Methods
To determ.ine whether non-freeze-dried foetal extracts are
able to raìse antibodies 'in other heterologous spec'ies, m'ice were
immun.ized with NFD and FD rat (|¡listar) foetal extracts and antìbody
production examined by immunodiffusion. This system was chosen for
three reasons:
(i) pregnant rats are available at regular intervals from
the Uni versi tY Anima'l House;
(ii) the rat foetus is relatively ìarge' and a considerable
quantityoftjssuecanbeobta.inedfromasingìe
pregnancy;
(iìi) previous'ly unpubìished results from this laboratory
showed that NFD rat foetal extracts were unable to induce
antibody production in rabbits. (Immunìzation with freeze-
dried rat foeta] extracts had not, however, been
attempted, and hence a comparison of these results with
the murine foetal results was not cons'idered valid')
The rat foetal extracts were prepared as described in Chapter II and
70.
TABLE 5.2
1 2 3RABBIT NO. ANTIGEN CONC. SITE RESPONSE
1
2
3
4
5
6
I8
3
3
3
3
i.v./s.c./i.m.i.v./s.c./i.m.i.v./s.c./'i .m.
i.v./s.c./i.m.i.m.i.m.
+
++
++
++
++++
+++
lmg protein per mì (8 mg protein per ml per iniection)'
Concentration was ca]culated from a freeze-dried extract reconstituted
to 15 mg solid matter per ml with distilled water'
2
3as in Table 5.1
7r.
inoculat'ion of Swiss Albino mice performed accord'ing to the protocol
des cri bed i n ApPend'ix V -
(b) Resul ts
Two groups of animals were immunized with either freeze-
dried or non-freeze-drìed rat foetal extracts. The ascites fluid of
each anjmal was examined for the presence of antibodies to rat foetal
antigens by 'immunodiffusion" and the results and statistical analysis
recorded ìn Table 5.3. The x2 analys1s of this data showed a
signifjcant difference in the ability of the two groups to raise
antibodies to foetal antigens, Ðiz. the group gìven freeze-dried
extract contained twice the number of animaìs which produce antìbody
to the foetal extract
These results are sim'ilar to those obtained in the rabbit/murine
foetal system. However, repression of antibody production in the
rabbits occurs in alI the an.imals given non-freeze-d.ied foetal extract'
whereas repress jon ìs not comp'lete in mice g'iven rat foetus' The
statjstical anaìyses, however, do show that the difference between
both groups of mice ìs significant'
There are two possible explanations for the incomplete repression
of the murine irnrnune sYstem:
(a) fhe degree of repression may vary amongst different animal
speci es .
(b) Incomplete repressìon may aìso occur in rabbits given
NFDfoetalextracts.However,becauseoftherelative]y
72
TABLE 5.3
Freeze dried Non-freeze-dri ed
No. anjmals positive
Total no. of animalsper group
% no. raising anti-bodi es
18 6
30
?9 20
62.07
x2 Analysis : p < 0.05, therefore difference is significant'
Number and percentage of antibody positive animals inrnunized with
either freeze-dried or non-freeze-dried rat foetal extract'
73-
smal I number of rabbi ts , animal s wi th pos'it'ive antì body
productjon may not have been isolated'
Whatever the explanation for this difference, it is clear that
repression of antibody Synthesis to rat foetal extracts in mice may
only occur if the lyophilization procedure is om'itted during the
preparation of the extract.
3. Concl usions
From the results presented here,'it is clear that NFD murine
foetal extracts were unable to evoke a humoral jrnmune response in
rabbits. Similarly, a statisticalìy significant number of mice were
unable to raìse ant'ibodies to NFD rat foetal extracts. In both cases
this negatìve ant'ibody response was reversed by freeze-drying the
foetal extract prior to inoculatìon'
Four possible explanations of this phenomenon are proposed here:
( j ) Immunoqen'ic'ity of the foeta'l extract
The extract used to raise the antibody was a 1500 g supernatant
of whole foetus homogenate. consequently, the extract contains many
small cellular organelles which may not have been disrupted durìng
homogenizatjon. Followjng lyophilization, these organelles may have
been broken, hence releasing new immunogenic material ' It is proposed
here that it is this material whjch causes the production of ant'ibody
in the host an'imals. Conversely, if lyophil'izatìon is omitted' the
organelles remain intact and no immunogenjc molecules are available
to cause antibodY Production.
74.
(ii) Immunocompetence of the animals
The raìs'ing of antibodies in the FD group of animals and
not in those given NFD foetal extracts may have been due to the non-
immuno-competence of the animals inoculated with NFD extracts' This
is, however, very unlikely because product'ion of rabbjt and mouse
antisera to other antigens studied 'in th'is laboratory has been
Successful (antigens include AFP, gamma globu'l'ins, placental extracts
and membrane-bound antigens).
( i i i ) Anti qen modi fì er
The third poss'ible explanation of the observed phenomenon
is that the foetus contains a modifier which renders the ant'igen non-
imrnunogenic. Further, thìs modifier may be destroyed by freeze-drying'
permi tt'ing the resumpti on of ant'ibody synthesi s '
(iv) Immunol oqical repressor
The foetus produces, or contains' an immunologìca'l suppressor
wh'ich is found in the 1500 g extract of the t'issue homogenate' This
repressor acts on the rabbit's humora'l 'immune system, preventing the
production of antibodies to the foetal antigens. The ìyoph'ilìzation
of the extract destroys the repressor, and antibody synthesis now
occurs.
In the following sections, these four concepts are exam'ined
further,to determine which may su'itably expla'in th'is unusual phenomenon
75.
SECTION III : IMMUNOGENICITY OF MURINE FOETAL EXTRACT
In the previous section, it was proposed that freeze-dryìng
may cause the djsruptìon of ce'll organeì1es, hence reìeasing rnewl
ìmmunogenic material. Alternatively, it was also theorizeC that
lyoph'i I i zatì on may destroy a modi f i er , whi ch somehow pre'rents ant'igens
from being immunogenic.
These two possibilities were examined by characterizing the
anti gens , both quanti tati vely and el ectrophoretì ca'l 'ly. Muri ne
foetal extracts were chosen for these experiments because these NFD
preparations were unable to evoke antìbody production in any
immunized rabbits, whereas rat foetal extracts did produce Some
humoral response i n i nocul ated m'i ce.
1. Cell 0rqanelle Ant'igens
To determine whether lyoph'ilization s'ignìfjcantly adds to
the immunogen poo'l by disrupting ce'|1 organe'lìes, fresh mouse
foetuses were homogenized and centrifuged at 105,000 g for one hour
to remove all und'isrupted cell organelles. The resultant supernatant
was dialysed and examined for the foetal antigens which are capable
of evoking humora'l 'immunìty. The ant'ibody used to detect these
antigens was raised to FD foetal extracts. If the concept that
immunogenjc materjal is released only after" freeze-dry'ing is correct'
then ant.igens should not be detected in the 105'000 g supernatant
since their source (the cell organel'les) has been removed. The
76.
immunodìffusion patterns of Figure 5.1, however, show that the
105,000 g extract does contain antigens which are capable of evokìng
immun'ity. Further, these antigens exhibit a reaction of identity
w'ith both FD and NFD 1500 g extracts. It may be argued that. altltough
the antigens are present'in the 105,000 g extract, theìr presence
may be due to earl'ier disrupt'ion of the organelles, and lyoph'iljzation
results in comp'lete breakage, causing the quant'ity of immunogens to
rise to a'level high enough to evoke humoral immunity. This is'
however, unlikely as the quantity of immunocompetent ant'igens in
both the 105,000 g extract and the FD,1500 g extract is comparabìe
(raute 5.4).
Thus it would seem that the breakage of cell organe'lles would
not signifìcant'ly add any more antigens to the immunogen poo'l .
2. Antiqen Modifier
The presence of an antigen modifier was exarnjned by
determining the number of ant'igens in the foetal extract, which
could be detected by the anti-foetal serum.
If an ant'igen mod'ifie¡is active'in these extracts, then it is
unlikeìy that one such molecule wjll alter the jmmunogen'ic'ity of a
variety of antigens. From Figure 5.2, it can be seen that there
are at least four, possìbly five, antigens which by their reactìv'ity
aga'inst this antiserum are able to ra'ise an ant'ibody'in rabbits.
Further, these antigens have a diverse range of electrophoret'ic
mobilities, some also being found 'in adult t'issues as well '
IGURE 5.1
Immunodiffusion examination of 105,000 g extract of murine foetus
r the presence of foetal antigen.
þ = 105,000gextract
d = 1500 g non-freeze-dried extract
f = 1500 g freeze-dried extract
a = adutt tissue extracts (This extract was used as a
positivecontrol.Previousstudiesfromthislaboratoryhaveshownthatfoetatextractsdocontain some adult antigens' )
{
p
-
I-ç
77.
TABLE 5.4
Ant'ibody Preci Pi tati onResponse of:
ANTIGEN DILUTION*
NEAT 2 4 B 16 32 64 r2B 256
**105,000 g
1500 s (FD)
***t +++ +++ +++ ++ ++ ++ +
+++ +++ +++ +++ ++ ++ ++ +
The ability of a constant concentration of anti-foetal serum to
precipitate various concentrations of antigens from a 105,000 g and
1500 g extract of mouse foetal extract'
*The antigen dilution is recorded as a reciproca'l dilution
(e.g. 2 -- >").
**Each extract was initìally adiusted to 10 mg of proteìn per ml'
f *** + r indicates the degree of precip'itation'
FTGURE 5.2.
Tmmuno-electrophoresis of murine foetal, and
The central trough contains anti-serum raisedextracts.
fcetal extracta = adult extract
adult extracts
against F. D. foet_a1
f : N.F.D.
N
ùs,
f,
z'9
78.
ilid
I
It'is unlikely that a sìng]e mod'ifier can alter the immuno-
gen.icity of such a d'iverse range of antigens. It is equally
unl.ikely that the foetus needs to produce a large number of
modifiers, and more ìmportantly, each one being 'lyophilization-
labile. The concept that freeze-dry'ing makes the antigens more
.immunogenìc would requ'ire a large number of antigens to be
lyophìlizatìon-labile - an improbable situation as such antìgens
are rare.
3. Conclusions
From these results ìt l¡||as concluded that the inab'ility of
NFD foetal extracts to raise an antibody in rabbits was not due to
the presence of an antìgen mod'ifier or to the lack of immunogen'ic
materi al due to the undi srupted state of cel 1 organel I es ' l'li th
the further improbability that the animals used in th'is study were
non-immunocompetent, it was proposed that the foetal extract contained
an immunolog'ical repressor, i.e. a substance which would 'inhibit the
productionofantibod'iestothefoetalextracts.
To accurate'ly determine whether such a repressor was present'
an isolation or partiaì purificatìon of the med'iatjng mo'lecule(s)
wasrequired.Twosuchmethodswerecons.idered:
(a)Co]umnchromatographyonSephadexG-200usingan0.85%
saline solvent.
(b) Ammonium sulphate fractionation'I
r
i
.I
n{,',,
,l
79.
The first technique'is probab'ly the better of the two, because the
inert Sephadex and saline would not be expected to alter the
activity of the repressor. The second method, although considered
"Safe" by biochemists, may have a similar effect aS freeze-drying -
particular]y since both of these techniques remove water' Thus the
,,denaturation,'of the repressor(s) by the ammonium ions may Oerntit
the production of antjbodies to non-freeze-dried foetal extracts -
a negatìVe result which would neither confirm nor reiect the presence
of a repressor.
Attempts at partialìy purifying murine foetal extracts by column
chromatography were not successful. consequent'ìy, the only remainìng
method for direct'ly showing that the foetus produced an 'immune
repressor was by using the foetal extract to 'impair antibody
production to an antigen' not produced by the foetus'
Because such an experiment required a 'large number of animals'
it was decided to inoculate m'ice with rat foetal extracts' rather
than rabbits with murine foetus. The use of a dìfferent animal
system was considered as valid because both rabbit and murine
antibody production to NFD foetal extracts was r'epressed'
tII
I
r
\
80
SECTION IV : IMMUNOLOGICAL REPRESSION
f. introduction
The presence of a lyoph'ilizatjon-labi'le immunological repressor
was examìned by immuniz'ing mice with bovine serum album'in (gSR) ' and
non-freeze-clrìed foetal extracts. If an ìmmune repressor was present
in the foetal extracts, then there should have been a repression of
antibodY Production to BSA'
2. Methodology
Threegroupsofmicewereinocu]atedwitheitherBsAand
saline(groupl),BsAandNFDratfoetalextract(group2),orBSA
and FD foetal extract (group 3). Antibody product'ion by the first
group w.iìì confirm the use of BSA as a suitable antìgen' The th'ird
group of animals serves as a control to determine whethelinject'ing
BSA and foetal extracts overloads the immune system, resutlting ìn
immune paralys'is. If thi s shoul d occur, there woul d be no ant'ibody
synthesis in the second group, causìng confusion between inh'ib'ition
by a repressor, and immune paralysis med'iated repression'
TheBsAwasmixedwiththefoetalextractandgivenintra-
peritonealìy according to the dosage and ìnoculation routes descflibed
in Appendix V.
HÌ
tÌI
I
t
81.
3. Resul ts
The antjbody product'ion by animals in the three groups is
presented in Table 5.5. All animals of the group given BSA and
saline (group 1) were able to raise ant'ibodies to this antigen. Thus
the concentration of ant'igen g'iven is suffic'ient to rajse antibodies
in mìce. The an'imals given BSA and NFD foetal extract were also able
to raise antibodies to the antigen. Conversely, the animals inlmunized
with FD foetal extract did not develop antibodies to the BSA.
Initial ly, it was thought that 'immune para'lysis had occurned in the
animals of the third group. However, this is unlikely because the
group given BSA and NFD foetal extract had the same quantìty of
antigenic material jniected as the group given BSA and FD foetal
extract.
These results are a direct contradiction of the supposition
that the NFD foetal extract contains an active immunologìcal suppressor
which should prevent the product'ion of antibodies to BSA.
Th.is apparent contradìctjon could be exp'lained by propos'ing that
the repressive process still occurred, but ìts inhibitive action was
specific only for foeta'l antigens and did not signìficantly alter the
antibody productìon to BSA. Under these condit'ions, it would be
expected that group 2 animals would raise an ant'ibody to BSA on'ly
and not foetal extracts. From Table 5.6 it can be seen that th'is
situation did occur. Similarly, group 3 animals would be expected
to ra.ise antibodies to both BSA and foetal extracts, as repressìon
has now been eliminated by freeze-dry'ing. This, however, did not occur'I
t
82.
TABLE 5.5
1
BSA + Saljne( coNTRoL )
2
BSA + NFD
3
95¡ + FD
No. positive
Total No. animals
% positive
t2
16
75
9
72
75
0
12
0
x2 Ana'lYs i s p < 0.05, therefore difference between groups 2 & 3
is sign'ificant.
Percentage number of animals able to raise antibodies to BSA, when
given concurrent immunizat'ions of saline, non-freeze-dried or freeze-
dried foetal extracts.
ANTI-BSAÃffiTVTTY
No. positiveTotal No. animals
% positive
ANTI-FOETALACTIV TY
No. pos'itì ve
Total No. animals
% Pos'i ti ve
TABLE 5.6
Group 2
BSA + NFD FoetAI
9
12
75
2
t2
t7
83.
GrouP 3
BSA + FD FoetAI
0
L2
0
11
12
92
STATIST I CS:
(f) nnti-BSA activitY :
(2) Anti-foetal activìtY :
(see Tabl e 5. 5) .
x2 analysis - p < 0.05, therefore the
two groups are si gn'if icantly di f ferent '
Percentage of animals inoculated with BSA and foetal extracts'
able to develop humoral immunity to these ant'igens.
B4
The aninrals of group 3 only raised antibody to the foetal antigens
and not BSA - an unusual result as both antìgens should have raised
an ant'ibody ì n mi ce.
A'lthough these resul ts conf i rm that repress'ion i s sti I I occurrì ng
in animals given NFD foetal extrac'L,s, it is unusual that antibodies
to BSA djd not develop in animals immunized with BSA and FD foetal
extracts (group 3). A poss'ibìe explana't'ìon is that an antigen 'in
the FD foetal extract competed wìth the BSA for the productjon of
antibodjes. Antr'genic competìtion between BSA and othet^ antigens has
been extensivel.y reported in the literature (see Liacopoulos and
Ben-Efraim, 1975). These authors have concluded that competition may
occur betvleen two antigens gìven simultaneously ìn the same Freund's
adjuvant emulsion. In the experiments desc¡ibed in thìs chapter'
the BSA and foetal extracts were m'ixed together and emulsified with
Freund's adjuvant.
Aìthough this hypothesis 'is tenable, it would be expected that
the NFD foetal antigens would also compete with the BSA' hence
elìminat'ing the latter immune response. However, this did not occur
in these experiments (Table 5.6). To expìa'in why competìtion did not
occur in thìs case, it is necessary to consider the mechanism of antigen
competit'ion. In the mouse experiments described here, ìt has been
proposed that a specific repressor prevents the production of antibody
to foetal antigens. In the next section of this chapter, it is
further proposed that this repression occurs at the macrophage site'
Gottl ieb et aL. (tglz) have reported that antigen competit'ion probably
occurs at the macrophage site, cluring ant'igen hand'ling and/or
85.
processing. consequently, when BSA is given together wjth the foetal
antigens, the immune reaction to the latter antìgens is suppressed'
hence elimjnating the conrpet'itjon wìth BSA. Thus BSA antjbody synthesis
now occurs. l,Jhen FD foetal extract is given with the BSA, however'
repression no longer ex'ists, competition l:etween both antìgen types
is inìtiated, and an immune response to the foetal ant'igens only occurs'
The concept of a competition between BSA and a foetal ant'igen'is
further supported by evìdence from Ruoslahti et aL. (1976) who have
shown that human albumin, and the foetal antigen AFP' shol considerable
amino acid sequence homology. (AFP is present in large quantìties in
rat foetal homogenates.) Further, Harel et aL. (tglz) have shown that
another globu'lin, bov'ine gamma gìobu1in, is able to compete with BSA'
resulting'in eliminat'ion of the latter's antibody production'
Thus, from these reports, jt 'is probable that ant'ibody production
aga'inst BSA when i niected simul taneously wi th FD 'îoetal extracts ' i s
suppressedbecauseofantigencompetition.Theresultspresented
here suggest that repression of antibody synthesis is specìfic for
antigens found in the foetal extract'
SECTTON V : DISCUSSION
The inability of rabb'its and mice to raise ant'ibod'ies to
heterologous foetal extracts has been demonstrated. After elimjnat'ing
al l othen 'l'ikely expl anati ons of thi s phenonlenon, 'it was proposed
that the foetal extracts contained a substance wh'ich was able to
B6
specifically repress the product'ion of antìbodies to the foetal
antigens. Further, this repressor was lyophìljzat'ion-labile'
Three .important questions have arisen from this work:
(i) What 'is the nature of the repressor?
(ii) llhat type of immune inhjbition does th'is molecule(s) exhib'it?
(ii.i) How does freeze-drying destroy the repressor(s)?
It is difficult to answer these questìons without further experi-
mentation. Holever, there is a considerable amount of information
in the lìterature from which a model of the nature and mechanism of
the repression may be formulated, and ultimately examined experimentally'
(i ) Nature of the rePt9::9r
To answer the fjrst quest'ion, it is desirable to list two
Ímportant criteria which the repressor(s) must fulfil:
1. Since the repression is specific for ant'igens in the foetal
extracts, then the antigens themselves must somehow be
medìatorsoftherepression.Itìsunìikelythatthe
antigens are immunosuppressive.* Rather' another molecule(s)
directly interacts with the ant'igens and th'is blocks the
humoral immune system from raising ant'ibodjes to these
anti gens .
2.Sincetherearea.largenumberofantjgens'eachw.itha
different electrophoretìc mobil ity (see Chapter V, Section
II), there must also be an equa'l number of repressors' each
*Th... have been a number of studies reported' wh'ich suggest that
cr1-foetoprotei n may be an 'imtnuno-regul ator (Tomasi et aL' ' I976;
Sheppard et aL., !g76; Goeken and Thompson, I976). However, these
stud'ies have been i nconcl us i ve , as yet '
87.
beìng lyophilization-labìle, and spec'ific for one ant'igen.
consequent]y, the repressors should be within one class of
chemical substance, i.ê. sharing many characterìstics, but
still able to confer specificity on a particular ant'igen - in
this case, foetaì antìgens.
Antibodies are molecules which conform to both of these criteria'
Although a different antigen specificity exists, the chemical compos'itjon
of the antibod.ies ìs essentìa'l1y the same. It'is difficult to envjsage
any other class of molecules wh'ich may conform to these establ'ished
crì teri a. Assumi ng that ant'ibocly mol ecul es may be the repressors '
it is then poss'ible to consider the mechanism of theìr actìon and
their labilitY to freeze-drying.
(i i ) Mechan'ism of action
It is unl.ikely that repress'ion is caused by antibody mask'ing
every antìgenìc determinant on all molecules 'in the extracts' A more
plaus'ible explanation is that only some of the antigens are bound to
antìbodies, and these compìexes specifically block the product'ion of
"new" antibodies. A similar situatìon has been reported by Mishell
and Dutton (tgoz), who showed that when anti-sheep red blood cell was
mjxed wjth an imrnunogen'ic dose of sheep red blood cells, thjs conjugate
was able to repress the appearance of antibody-form'ing cells to this
antigen. Nossal and Ada (197i), rev'iewjng this area of immunology'
have concluded that the repression occurs early in the immune process'
poss i b'ly at ant'igen-macrophage i nteractì on '
B8
The complexitrg ant'ibody must be of either foetal or maternal
origin. Although some specific globulìn synthesis by the foetus has
been reported (Solomon, 1971), it is improbable that this antjbody
forms the anti gen-anti body compl exes . consequent'ly, th'is ant'i body
ntay be of maternal orig'in. A number of workers have reported the
presence of maternal antibody against antjgens of the developing
foetus. Further, they have shown that antìbody synthesis in neonatal
animals may be repressed by the presence of ìarge quantities of
spec.ifìc maternal antibody (Greengard and Bernstein, i935; Cooke
et aL., 1948; Barr et aL., 1953). The proposed mechanism of action
of the repressor found ín the foetal extracts descrìbed in this chapter
is consistent w'ith the results reported by other workers, i'e' the
maternal antibody in the foetus forms antigen-antìbody complexes'
vrh'ich may prevent the productìon of ant'ibody to these antigens by
heterol ogous hosts.
(ì i i ) Lyssþill:e!ie!:liÞlulv-gf ress 1 0n
The experiments described here have shown that the
repression phenomenon is labile to freeze-drying- To explain this
labiìity, in terms of an active antigen-antìbody repressor, freeze-
dry'ing must either destroy the antibody portìon of the immune complex
or its bond with the ant'igen be broken. It is unlìke'ly that
lyophilizatìon destroys the antìbody 'itself , because freeze-dryìng is
rout.inely used to preserve antisera from degradat'ion. Thus the
- lab.ile po'int must be the bond between the ant'igen and antibody'
B9
There has not been any report in the literature which describes
a dissociation of antibody-antigen compìexes by freeze-dryjng'
There have, however, been numerous reports of high salt concentrations
caus'ing the dissocjation of these complexes into free antibody and
anti gen (Hei del berger et aL., 1936; He j de1berger ancl Kendal I , 1936) '
It is possìble that the foetal extracts used'in these studjes had a
hypertonic salt content after freeze-dryìng and subsequent
re-constitut'ion. This would result in a breakjng of the ant'igen-
antibody bond, and the subsequent elìminat'ion of the repressive
phenomenon. The valjdity of thìs observat'ion is clearer ìf the
equiì.ibrium reaction for antigen-antibody complex formation js
cons i dered:
'isotonic salt -tsiz Ag
(antigen)
Ab
( anti bo
+ Ag*Ab
(complex)ìypertonic salt
dv)
During the preparat'ion of murine foetal extracts, for rabbit inoculation'
the freeze-dried'inocu'lum was re-constituted to 15 mg of solid materials
per ml of d.ist.illed water. This results 'in the salt concentration of
the inocul um bei ng e'levated, hence shi f ti ng the equ'i ì ì bri um reacti on
to the left. Raffel (tgOt) has reported that only a 15% increase ìn
salt content is sufficient to cause this sh'ift' Consequently' it
may be poss.ible that a smalj elevation in salt concentration'in the
murjne foetal ext,racts was able to destroy the repressive antìbody-
antigen comPlexes.
90
Foetal rat extracts, for inoculatìon 'into mjce, were re-
const'ituted to their pre-'lyoph'iÌization volume. Thus they would not
be expected to contajn a sìgnificant quantity of dissociated immune
compìexes, because the concentration of the salt has not been altered.
Horvever, during the 'lyoph'il'ization procedure, there 'is a
concentrating of the salt that ìs present, and hence dissociation
may occur here. However, on reconstitution, reassociation would be
expected over a period of time because the extract solutìon has been
restored Lo isotortìc strength (see equilibrium reactìon) ' This
si tuati on .i s probab'ly ref I ec'ted experimental 1y i n Tabl e 5. 3, where
37.5% of an.imals given FD foetal extracts unexpectedly dìd not raise
antibodies. This may have been due to re-aggregat'ion of antigen and
ant'ibody i nto jrnmune comp'lexes.
Thus it is proposed here that the immune repress'ion demonstrated
in these studies is not specifical'ly reversed by lyophil'izatìon.
Rather it.is the eleva.tion of salt concentrations brought about by
lyoph'i I i zat'ion/reconsti tuti on procedure that reverses the repressi on '
( i v) !yuugl.y
In thjs chapter it has been demonstrated that antibody
synthesìs against foetal extracts occurs when 'lyoph'il i zation
is included in the preparative procedure. This phenomenon is thought
to be due to the action of an ant'igen-specific humoral immune system
- repressor. A theoretical model, formulated to expìain the mode of
repression, has been presented. However, it'is'important to note that'
'raldeqo lxau aql u! rêtlunJ passnlslp s! Lapou pasodord aql Jo 6utlsal
LPluaullJadxaaql.|.apouaqxJosldacuocaqlsl¡oddnsÁ¡6uo"t1s
p[e!J aq1 6uLra^oc arnlerallL aql 'a^!1e¡noads s¡ Lapoul stLll a[lqm
'I6
92
SECTION I : DEVELOPMENTAL ANTIGIN REAPPEARANCE IN CANCER
In this thesjs, the methorlology for the preparation of antibody
to both foetal and placental antigens has been the same, i'e' an
anti serum was rai sed to each extract, and then an attempt was made to
renderit speci f ic for ei ther foetal or pl acental antì gens by
absorptìon w'ith aclult tjssue powders. Thìs technìque was desìgned
to detect a range of foetal and placental antìgens, without
specifying the t'issue of o¡igìn of the antigen. However' one maior
prob'lem prevented the effective use of these systems to detect
developmental antigens in tumours - both foetal and placental extracts
corrtained a high concentration of adult antigens' Thus the antisera
to these extracts had a h'igh proportion of antibody whjch was
directed against the normal adult antigens' e.g. the antiserum to
intact placenta'l cells was exclusively reactive against adult red
blood cells, and not placental ant'igens'
For this type of system to be used successfully, the'immune
response to the adult antigens should be el'iminated or significantly
reduced. This may be done in one of two ways:
(a) Immun'ization of homologous hosts instead of heterologous
anìmal s.
(u) prior purif.ication of the extracts, resulting in a
relatively higher concentration of developmental antigens
be'ing available for stimulation of the immune system'
93
Both methods have been briefly d'iscussed in the earlier chapters.
The second concept of purification, however, requires Some further
discussion as immunizat'ion of homologous hosts may not always result
in antibody production (see Chapter IV).
The purification of developmental tissue extracts may be
achjeved by eìther:
(a) fractionat'ing the extract into smaller pools' or
(b) selecting a specific foetal organ or placental tissue.
The first method of purification may involve either Sephadex G-200
or ammonium sulphate fractionation. These two techniques may ensure
that each fraction contains a higher proportion of certain foetal
antigens than the unfractionated extract. Thus the immunizing host
would receive a larger quantity of foetal antigens. This method may
al so be appl i ed to p'lacental anti gens .
The second method involves the selection of one particular
foetal organ and rais'ing an antibody Eo either a soluble or membrane
extract of thjs tissue. Again, this may eliminate excess antigens'
and in particular, d large quantity of adult antigens. This type
of approach has been used successfully in thjs laboratory for the
preparatìon of anti-AFP,
tsiz, rather than used whole foetus, amnìotic fluid was used as
the vaccine for the preparation of the ant'iserum (Okita et aL., I974).
A simiìar approach may also be adopted to rajse antibody to both
membrane and cytoplasm'ic extracts of the placenta. In this method'
it is desirable to firstly perfuse the placenta with jsotonìc saline
94
in order to remove as much blood as possible. Following this' a
particular area of the placenta is selected and an extract (either
membranous or cytoplasmic) prepared for immunization. Sjm'ilarly'
other areas of t'issue may be chosen. Thus there is an eliminatìon
of excess antigen, and an effectìve increase in the concentration of
p'lacental specific antigen enter.ing the immuniz'ing host.
In conclusion, although the methods proposed above may st'i11
be effectìve as a means of preparing spec'ific antibody for detecting
developmental antigens jn tumours, it is not possible to produce a
single antiserum wjth multiple foetal and p'lacental antigen
specific'ity.
SECTI0N II : IMI'IUN0L0GICAL REPRESSI0N
Thea.imofth.isstudyhasbeentodeterm'inewhetherthefoetus
produces a 'lyophilization-labile immune repressor or antigen
mod.ifier. It has been proposed in this thesis that the rat and
mouse foetal extracts conta.in ,modi f i edr anti gens wh'ich are abl e
to specifìcally inhibjt the production of ant'ibodies to these
antigens'in xenogenic hosts. Further, this modificatjon process is
an attachment of antibod'ies to the foetal antjgens' and it is this
antigen-ant.ibody complex wh'ich represses antibody product'ion to the
foetal ant'igens.
'However,itisimportanttonotethatthisconceptislargely
specu'lative, and aìthough supported by a number of experiments
95.
reported jn the literature, no direcÈ experimental evidence has been
presented in this thesis. Further, a direct contradiction of this
theory is evident from the results of tlishi (1970), who has been able
to raise an antibody to AFP by imrnunizing horses with AFP-anti-AFP
compìexes. This evidence emphas'izes the need to establjsh the vaf idity
of the theory presented'in Chapter V and to use th'is to possibly
determine Nishi' s d'iscrepancy.
There are three important aspects of this theory which may be
examined experimental'ly:
(a) Are there sìgnifìcant numbers of antigen-antibody complexes
in the NFD foetal extracts?
(b) Can these complexes' if they exist, be dissociated during
freeze-drying and further reconstitution wjth distilled
water?
(c) l,.lill the addition of antigen-antibody complexes to a FD
foetal extract result in the inhibition of ant'ibody
product'ion to the foetal antigens?
These experiments may assist jn confirm'ing or rejectìng the proposed
theory.
The relevance of the proposed model to tumour immunity may be
considered in view of the many similarit'ies between foetal and
cancerous cells (Chapter I). Currie (fg0g) has shown that maternal
T-lymphocyte attack against the developing foetus may be repressed
by bl ock'ing factors whi ch cons i st of anti gen-ant'i body factors . Such
96
blocking factors have also been impl'icated as a defense mechanjsm
which tumours possess to prevent destruct'ion by the hostrs immune
system (Sjogren et aI 1971 Chalmers et aL., 1976).
Evidence in favour of B-ìymphocyte (i.e. ant'ibody production)
suppressi on by ant'ibody-foetal ant'igen compl exes has been I imi ted
to a few studies from our laboratory. Sjx attempts have been made
to raise antibody to NFD 1500 g extracts of hepatoma tissue. Each
attempt was found to be unsuccessful. By repeatìng the experiments
in the previous chapter with tumour rather than foetal extracts, the
possible role of antibody-foetal antigen complexes in the reguìation
of B-lymphocyte immunity to the developing tumour may be determined.
The model of B-'lymphocyte suppression presented in this thesis,
requires further examination and verification. However, ìts
importance in tumour, foetal and transplantation inrnunology, is such
that the basic concepts of this model may prov'ide an impetus for
further investigation.
97.
CHAPTER IV
Sabìne(personaìcommunjcatìon),us.ingthesamemethods
descrjbed in this chapter, has also been unable to produce a
specific antìserum capable of detecting carcino-p'lacental antigens
in syngeneic rat tumours.
CHAPTER V
An ìmportant observation has been made by a number of
coìleagues, during d'iscussions of the results presented in thìs
Chapter.Theyhavenotedthatiftherepressorisspecìfìcfor
foetalantìgensonly,thentherabbitsandmiceshouldhave
raised antìbodies to the adult component of the foetal extract'
However, it ìs 'important to note that the adult component
consists of both maternal and paternal antigens. consequently'
the foetal extract would contajn paternal antigen-ant'ibody
complexes, because these antigens would be recognized as foreign
by the maternal immune system' Hence antibody synthesis to
these ant'igens in heterologous hosts would be inhibited'
The maternal antigens should stimulate antibody product'ion in
heterologous hosts, as there shoujd be no antigen-antibdoy complexes
present.However,the]eve]oftheseantigensìnthefoetalextracts
isprobablytoolowtoproduceenoughantibodywhichcouldbe
.detected by the insensitive immunodiffus'ion system emp'loyed in
il'G
I
II
r
these studies.
*.'{
Ê+
âæ
98.
APPTNDIX I
IMMUNIZATION SCHEDULE FOR THE PRODUCTION OF RABBIT ANTI-SãRUM
TO MOUSE FOETAL EXTRACTS.
Day Iniection No. Inocul um Dose
gen Adiuvant)* TYPe : (ml )
si te***
Anti(ml
1
B
L5
29
1
2
3
4
1.0
1.0
1.0
1.0
FCA
FCA
FCA
FIA
1.0
1.0
1.0
1.0
lm
tm'lm
1m
*Antigen concentration was 15 mg solid matter (after
lyophilization).
**FCA : Freund' s ComP'l ete Adj uvant
FIA : Freund's IncomPlete Adiuvant
**:k5. ç. : SUbCutaneggs
i.m. : intramuscular
Animals were bled 10 days after the final inoculation'{l
Ir'
IT
I
I
rI
99
APPENDIX II
IMMUNIZATION SCHEDULE FOR THE PRODUCTION OF RABBIT ANTISERUM TO
SALINE SOLUBLE EXTRACTS OF RAT PLACENTA.
Day Injection No. Inocul um Dose
Anti gen* Adiuvant(mÍ) Type : (m])
Si te
I
2
3
4
1
8
15
29
1.0
1.0
1.0
1.0
FCA
FCA
FCA
FIA
1.0
1.0
1.0
1.0
s. c.
s.c
s.c
tm
*antigen concentration l,las I mg protein per ml '
Animaìs were bled 10 days after the final inocu'lation.
100.
APPENDIX III
IMMUNIZATION SCHEDULE FOR THE PRODUCTION OF ANTISERUM TO RAT
PLACENTAL ANTIGENS.
1. HETEROLOGOUS ANTISERUM TO l^lHOLE
PLACENTAL CELLS.
Day Injection No. Inocul um Dose*
(ml )
Si te**
1
6
11
2
2
2
1
2
3
nprlil'tv
IV
*concentration of cells was 3.6 x 108 per ml'
**n. p.
i.m.
i.v.
nape of neck
i ntramuscul ar
i ntravenous
Animals were bled at 7 day intervals after the final
i nocul ati on
101.
2. HETEROLOGOUS ANTISËRUM TO CRUDE
PLACENTAL CELL MEMBRANE EXTRACT
Day Iniect'ion No. Inocul urn Dose
Anti gen Adi uvant(mi ¡* TYPe : (ml )
Si te
I
2
3
4
I
15
29
43
1.0
1.0
1.0
1.0
FCA
FCA
FCA
FIA
1.0
1.0
1.0
1.0
i.m
i.m
i.m
i.m
*concentration of antigen was 10 mg prote'in per m] '
Animals were bled 14 days after the finaj inoculation'
3. HOMOLOGOUS ANTISERUM TO CRUDE
PLACINTAL MEMBRANE EXTRACTS
Day Iniection No. Inocul um Dose
Anti gen Adi uvant(ml )* Type : (ml )
Si te**
1
15
22
43
57
1
2
3
4
5
0.5
0.5
0.5
0.5
0.5
FCA
FCA
FCA
FCA
FIA
0.5
0.5
0.5
0.5
0.5
i.p
i.p
i.p
i.p
i.p
*ant'igen concentration was 10 mg protein per ml '
**i . p. : i ntraPeni toneal '
Animals were bled 14 days after the fjnal jnoculation.
102.
APPENDIX IV
IMMUNIZATION SCHEDUI-ES FOR THE PRODUCTiON OF ANTIBODY TO NFD
MURINT FOETAL EXTRACTS.
1. FOOTPAD PROTOCOL
Day Iniection No. Inoculum Dose
Anti gen Adi uvant(ml )* Type : (ml )
Si te
1
2
1
11
0.5
0.5
FCA
FiA
0.5
0.5
f.p
f. p.
*concentration of the ant'igen was 10 mg prote'in per m] .
Animals were bled 10 dals after the final jnoculation.
2. SUBCUTANEOUS. INTRAMUSCULAR
PR0T0C0L (i.e. s.c./i.m. )
Day Iniection No. Inocul um Dose
Antìgen Adiuvant(ml )* Type : (m] )
Si te
I
2
3
4
1
I15
29
1.0
1.0
1.0
1.0
FCA
FCA
FCA
FIA
1.0
1.0
1.0
1.0
s.c.
s.c
s.c.
l.m
*concentration of the antigen was either 8 or 3 mg
protein Per ml (see Table 5'1)'
Animals were bled 10 days after the final inoculation'
103
3. INTRAI'IUSCULAR PROTOCOL
Day Iniectjon No. Inocul um Dose
Antigen Adiuvant(ml )" Type : (ml )
Si te
1
2
3
1
15
29
1.0
1.C
1.0
FCA
FCA
FIA
1.0
1.0
1.0
'l
j
i
m.
m.
m.
*concentration of the ant'igen was B mg protein per m] .
Animals were bled 10 clays after the finaì inoculation.
4. INTRAVTNOUS-SUBCUTANEOUS-
INTRAMUSCULAR PROTOCOL
Day Iniection No. Inocul um Dose Si te
Anti(ml
qen Adi uvant)* Tvpe : (ml )
1
2
3
4
5
1
1
B
15
29
0.5
0.5
1.0
1.0
1.0
FCA
FCA
FCA
FIA
0.5
1.0
1.0
1.0
ì.v
s.c
s.c
s.c.
i.m
*concentration of the ant'igen was 3 mg protein per m1.
An-imals were bled 10 days after the final inoculation'
The same immun'izat'ion schedules were employed when raìsing
antibodies to FD extracts vlhich use these protocols (Table 5'2)'
104.
APPENDIX V
IMMUNIZATION SCHEDULE FOR THE PRODUCTION OF MURINE ANTISERUM TO
RAT FOETAL EXTRACTS, AND BOVINE SERUM ALBUMIN.
Day Injection No. Ant'igen
Type : Dose(ml )
Adj uvant
Type : Dose(nrl )
Site
1
15
21,
25
30
37
1
2
3
4
5
6
Sal i ne 0.2
Foetal * 0.2
Foetal 0.2
Foetal 0.2
Foetal 0.2
Foetal 0.2
FCA 0.2
FiA O.?
FIA 0.2
FIA 0.2
FCAM** 0.2
FCAM 0.2
rprprp1p
rp1p
*concentration of foetal inoculum was 2.7 ng protein per ml '
Where BSA was given also, it was added to the foetal extract (or
saline) prior to inoculation, to give a final concentration of
15 mg per ml.
**FCAM : Freund's comp'lete Adjuvant with 5 mg per cent live
Mycobacter,iunphLeiadded(Sommervi.l1e,1967).
Ascites fluid from these an'imals was tapped B days after the
final inocul ation.
ABELEV, G.I. (i965). Antigenic structure of chemically incluced
hepatomas. Prog.Exp. Tumour Res' 7: 104-157'
ABELEV, G. I . ( 1968) . Product j on of embryona'l serum o-91obuì 'in by
hepatomas: Review of experimental and clin'ical data. cancer
Res.28; 1344-1350.
ABELEV, G.I. (1971). Alpha-fetoproteìn ìn ontogenesis and its
association with mal'ignant tumours. Adv. cancer Res. 14:
295-358.
ABELEV, I., S.D. PEROVA, N:1. KHARMKOUA, Z.A. POSTNIKOVA and i.S.
IRLIN (1963). Production of embryona'l globulin by transp'lantable
mouse hepatomas. Transpl ant I : 174-180 '
ALEXANDER, P. (Ig72). Foetal antigens in cancer. Nature 235: 137-140'
ANDERSON, N.G. and J.H. COGGIN (1971). Models of d'ifferent'iation'
retrogression and cancer. Proc. 1st conf. Embryonaì and Fetal
Antigens in Cancer, PP.7-37 ' AAEC: Oak Ridge'
ARTZ, K., P. DUB0IS, D. BENNETT, M. CONDAMINE, C. BABINET and F' JACOB
(1973). Surface antigens common to mouse clearage embryos and
primit'ive teratocarcinoma cells in culture. Proc' Nat' Acad' Sci '
70;2988-2992.
BAGSHAI,,¡E, K.D. (rg74). Tumour-associated antigens. Br. Med' Bull. 30:
68-73.
BALDWIN, R.W. and D. GLAVES (1972). Deletion of liver cell-surface
membrane components from amino-azo dye 'induced rat hepatomas'
Int. J. Cancer 9: 76-85.
BALDT^jIN, R.ll|., D. GLAVES, J.R. HARRIS and M.R. PRICE (1971). Tumour-
Specific antigens assocìated witlr am'ino-azo dye induced rat
hepatomas. Transp'1. Proc. 3z 1189-1190'
BALDI^IIN, R.l,l., D. GLAVES and B.M. VOSE (1972). Embryonìc antigen
expression in chemjcally'induced rat hepatomas and sarcomas'
Int. J. Cancer lo: 233-243.
BANHQ,0., IERSEY, J. and J.R. HQBBS (1974). A new oncofoetal ant'igen
for human pancreas. Lancet i': 643-645'
BARR, M., A.T. GENNY and J.Ì¡l. HOÌ^lIE (1953). Act'ive immunìzat'ion of
ewes and their lambs. J. Pathol. Bacteriol. 65: 155.
BECKER, F.F" (1974). Differential lectin aggl utinat'ion of fetal '
diving, post natal and maf ignant hepatocytes. Proc. Nat' Acad'
Sci. 71.; 4037-4041.
BERRY , M. ( 1973 ) . H'i gh yi eì d preparat'i on of morphol og'i cal'ly i ntact
i sol ated parenchymal cel I s from rat 1 i ver. Me*"hods i n Enzymo'l ogy
32: 625-632.
BQNMASSAR, E., E. MENCQNI, A. GQLDIN and G. cuDKQl^licz (1974)' tscape
of small numbers of allogeneìc lymphoma cells from immune
surve.illance.J.Nat].CancerInst.53:475-479.
BORSOS, T., A.K. RICHARDSON, S.H. OHANIAN ANd E.J. LEONARD (1973) '
Immunochemical detection of tumour-specific and embryonic antigens
of d.i-ethyl ni trosami ne i nduced gu'inea p'ig tumours. J ' Natl '
Cancer inst. 51-: 1955-1960'
BRAUNSTTIN, G.D., J.L. VAITUKAITIS, P.P. CARBONE ANd G.T. ROSS (1973).
Ectop'ic production of human chorionic gonadotrophìn by neopìasms'
Ann. Intern. Med - 7B: 39-45'
BURGER, M.l'1. (1973). Surface changes in transformed cells detected
by lectins. Fed. Proc. 32: 91-101'
CAUCIII, M.N., J.B.hJ. HALLEY, M'G' IRViNG and J'F' blILt'IAMS (1974)'
Hepatocellular localization of fetal antìgens during'induct'ion of
rat liver tumours by 3l-methyì-4-dìmethylamino-azo-benzene'
Cancer Res. B: 1808-I8I2.
CHALMERS, p.J., N. MATTHEWS and R.c. NAIRN (1976). characterizat'ion
of serum factors modul at'ing spl en'i c cytotoxi ci ty i n a syngenei c
rat tumour system. Br. J' Cancer 34: 645-650'
CHANG, C.H., T. NISHIYAMA, D.A. ANGELLIS and C. CONNERY (1976)' Iso-
immunogenic p'lacental prote'ins. Proc. conf. 0ncodevelopmental
Gene ExPres s'i on , P .26 .
CHEI-!JEI, L. and S.D. KIRLEY (1976). Properties of onco-p'lacental
histam.inase. proc. Conf. 0ncodevelopmental Gene Expression, p'56'
COGGIN, J.H., K.R. AMBROSE and N.G. ANDERSON (1970). Fetal antigen
capable of induc'ing transplantation immunity agaìnst SV-40
hamster tumour cells' J' Immun ' L05: 524-526'
COLIGAN,J.E.'P.A.HENKART,C.W.ToDDandl^l.D.TERRY(1973).
Heterogeneity of the carcinoembryonic antigen. Immunochem' 102
59 1- 599 .
C0LLINS, J.J. and P.H. BLACTT (1973). Analysis of surface ant'igens on
simian v'irus 4o-transformed cells: I. Unique antigenic'ity of
Simian vìrus 4o-transformed outbred hamster kìdney cell lines'
J. Natl. Cancer Inst' 51-z 95-114'
COOKE, J.V., J. HOLLOI¡IACH, J.E. ATKINS ANd J.R' POI^IERS (1948)'
Antibodyformationinearly.infancyagainstdiphtheriaand
tetanus toxoid. J. Pediat ' 33i 141-146'
cuRRIE, G.A. (1969). The foetus as an allograft: the roje of
maternal "unresponsiveness to paternally derìved foetal
antigens". In: Foetal Autonomy. (E¿s.) G.E. I{olstenholme
and M. 0' Connor . Ch urch i 'l I : London ' pp '32-52 '
DQELLGAST, G.J. (1976). Homogeneous human pìacentaì aikal'ine
phosphatase shares antigenic determinants w'ith human intestinal
alkatine phosphatase. Proc. Conf. Qncodevelopmental Gene
Expressìon, P.41..
ELKHALIFA, M.D. and J.A. LIPPiNCOTT (1968). Quant'itative meastrrement
of interactions between crown ga]l tumours and the pinto bean ho:'b'
Am. J. Bot.55: 382-391.
ELKIND, M.lq. and G.F. I,IHITM0RE (1967). The radiology of cultured
mammal ian cel I s. Gordan and Breach : N 'Y '
FAILLY-CREPIN, C. and J. URIEL (1973). A fetal carboxylic esterase
of mouse orig'in common to normal embryon'ic brain and murine
experimental teratocarcinoma. FEBS Letters 33; 351-354'
FISHMAN, þJ.H., N.R. INGLISS, S. GREEN, C'LANTISS and N'K' GOSH (1968)'
Immunology and biochemistry of Regan enzylne of alkaljne phosphata'si
in human cancer. Nature 219: 697-699'
FOX, T.0., J.R. SHEPPARD and M.M' BURGER (1971)' Cyclic membrane
changes in animal cel ls. Transformed cejls pernranently display
a surface archjtecture detected 'in normal cells only during
mitosis. Proc. Nat. Acad. Sci' 6B: 244-247'
FRITZE, D., D.H. KERN, C.R. DR0CEMULLER and Y.H. PILCH (1976).
Product'ion of anti sera wi th spec j f i ci ty for ma] 'ignant mel anoma
and human fetal skin" Cancer Res' 36: 458-466'
GARRID0, J., M. BURGLEN, D. SAM0LYK, R. l/llCKER and hJ. BERNHARD (1974)'
ultrastructural comparison betvreen the distribution of concanavalin
A and wheat germ agglutin cell surface receptors of nortnal and
transformed hamster and rat cell lines. cancer Res' 34: 230-243'
GOEKEN, N.E. and J.s. TH0MPs0N (1976). Immuno-suppressive properties
of human aìpha-fetoprote'in and human cord plasma labumin' Proc'
Conf. Oncodevelopmental Gene Express'ion ' p'26'
GOLD, P. and S.0. FREEDMAN (1965). Demonstration of tumour-specific
ant'igens i n human col oni c carci nomata by 'immunol ogì cal tol erance
andabsorpt.iontechniques.J.Exp.Med.l-21:439-462.
GONANo, F., V.P. cHIARUGI, G. PIRRO and M. MARINI (1971). Transfer
ribonucle'ic acids in rat I iver and Morris 5123 minimal deviat'ion
hepatoma. Biochemi st'ry L0: 900-908'
GgNANg, F., G. PiRRg and S. sILVETTI (1973). Foetai'liver tRNAphe in
rat hepatoma. Nature New Biol ' 242" 236-237 '
GOTTLiEB, 4.4., R.H. SCHI^IARTZ and S'R' WALDMAN (I972) ' Competition
between L- and D-synthet'ic copolymers at the level of the
macrophage. J. Immun . 10Bt 7 19-725'
GRANT, J.P., S. LADISCH and S.A. I^IELLS (1974). Immunologic similarities
between fetal cell ant'igens and tumour cell antigens ìn guinea pigs'
Cancer 33: 376-383.
GREENGARD, J. and H. BERNSTTIN (1935). Passive immunjty 'in infants
and their response to diphtheria toxoid. J. Am. Med. Assoc.
105: 34I-34?.
HALVER, J.E. (1965). Aflatoxins and rainbow trout hepatoma. In:
"Mycotoxins and foodstuffs", Ed. by G.N. Wogan. canrbridge :
MIT Press.
HÄfflnfn , I . (Lg72) . Immunochemi cal rel ati onshi p of the carc'i no-
embryonì c anti gen and the fetal su] phoglycoprote'in antì gen '
Immunochem. 9: 1115-1119.
HAREL, S., P. LIACQPQULQS and s. BEN-EFRAIM (1972). Inhib'ition of
jmmunological components by ant'igenic competition, and other
immuno-depressing procedures. Immunol . 22i 515-524.
HARSHBARGER, J.C. and R. TAYLQR (1968). Neoplasms in insects'
Ann. Rev. Entom.. l3: 159-190.
HEIDELBERGER, M. and F.E. KENDALL (1936). Quantitative studies on
ant'ibody puri f i cati on . I . The di ssoci at'ion of prec'i p'it'ates
formed by pneumococcus specific po'lysaccharides and homologous
antibodies. J. Exp. Med. 64: 761-172'
HEIDELBERGER, M., F.E. KENDALL and T. TEORELL (1936). Quantitative
studi es on the preci p'it'i n reacti on . J . Exp. Med. 63: 819-826 '
HOLLINSHEAD,4., C. McWRIGHT, T. ALFORD, D. GLEW, P. G0LD and R'B'
HERBERMAN (Ig7?). Separation of skin reactive intestinal cancer
anti gen f rom the carc'ino-embryon j c anti gen of Gol d ' Sc'ience
1.77: BB7-889.
H0ROWITZ, M.L. and l^1. E . KN0x ( 1968) . A phosphate acti ve gì utanri nase
ìn rat liver different from that in. kidney and other tissues"
Enzym. Biol. Clin. 9: 241-255.
HIEBNER, R.J. and G.J. TQDARQ (1969). Qncogenes of RNA tumour
Viruses as determinants of cancer. Proc. Nat. Acad. Sci. 642
1087- i094.
HUGHES, D.W.0'G., P.R. LAM-PQ-TANG and B. WEBSTER (1973). Foetal
haemoglobin levels ìn acute myeloici leukemia of chjlclhood:
Their value 'in predicting remissjon. Med. J. Aust. 2" iC03-1004'
HULL, E.t¡I., P. CARBQNE, D. GITLIN, R.W. 0'HARA and M.G. KELLEY (1969).
a-Fetoprotein jn monkeys with hepatoma. J. Natl' Cancer Inst'
422 1035-1044.
INGLEToN, P.M. (1971). Acute effects of hepatocarcinogens on mouse
liver microsomal antigens. Eur. J. Cancer Z: 501-504.
ISHIM0T0,4., Y.SUZUKI, T. YQSHIDA and Y. IT0 (1974). Further studies
on mouse fetal ant'igen cross reacting with Rauscher leukemia.
Cancer Res. 34; 2338-2343.
KNOX, l^J.E. (Lg72). Enzyme patterns in foetal, adult and neoplastic
tissues. pp.164-184. Karger : Basil '
KNOX, hJ.E. , M.L. HOROL.ilTZ and G.M. FRiEDELL (1969). The proportional itv
of g'lutaminase content to growth rate and morphology of rat
neoplasms. Cancer Res. 29: 669-680'
KODA, T., S. ISHIGAMI and S. TANABE (1973). Studies on liver cancer'
II. Productjon of cr-fetoprotein by experimental hepatoma, and
its characterization' Biken' J' 1-6: L29'
KROES, R.M., J.M. SONTAG, J.H. l¡JEISBURGER, P. NEWBERNE ANd G.N.
lÁlOGAN (tglz). u-Foetoprotein in rats with hepatomas induced
by Af'latoxin 81. Nature 240: 240-24L'
KUUSELA, P., E. RUOSLAHTI and A. VAHERI (1975). Polypeptides of a
glycoprotei n ant'igen ( S .F. ) present 'i n serum and s urface of
normal but not of transformed chicken fibroblasts' Biochem'
Biophys. Acta sTsz 295-303.
KYRIAZIS, A.P. and R.l,J. hlISSLER (t972). Demonstration and
purification of a specjfic antigen in Morris hepatoma 5123'
Lab. Investigation 26: 178-183'
LAURENCE, D.J. and A.M. NEVILLE (tglZ). Foetal ant'igens and their
d.iagnosi s and cl i ni ca1 management of human neop'lasms ' A revi ew '
Br. J. Cancer 26: 335-355'
LAURENCE, D.J.R., V. STEVENS, R. BETTLELHEIM, D. DARCY' C' LESSE'
C. TURBERVILLE, P. ALEXANDER, E.W. JOHNS and A..M' NEVILLE (1972)'
Role of plasma carc'ino-embryonic ant'igen in diagnosis of gastro-
intestinaì, mammary and bronchìal carcinoma. Br. Med' 'l' 3z
605-609.
LIAcop0ul0s, p. and s. BEN-EFRAIM (197s). Antigenic compet'ition.
Prog. in AllergY 18: 97-204'
LoGERF0, P., F. LoGERFO, F. HERLER, H.G. BARKER and H.J. HANSEN (1972).
Tumourassociatedant.igensinpatìentswithcarcinomaoftheco]on.
Am. J. Surg. 123i L27-L3l'
rÌil;,
,rl
L0t^lRy, 0.H. , N.J. R0SEBR0UGH, A.L. FARR and R.J. RANDALL (1951).
Prote.i n meas urement wi th the Fol i n pheno'l reagent. J ' Bi ol '
Chem. 19 3: 265'275 .
MARTIN, F., S.KNOBEL, M. MARTIN and M. BORDES (1975). A carcinofetal
antigen located on the membrane of cells from rat intestinal
carcinoma 'in culture. Cancer Res' 35" 333-336'
MAUGH, T.H. (1974). Fet,al antjgens - A biochemical assay for cancer'
Sci ence I 84 : I47 -1'48,
MESHKINP0UR, H., H.T. l¡lESPIC and w.R. SCHMALH0RST (1974) ' Alpha-1-
fetoproteìn and alcohol ic hepatìtis. Am. J. Dig' Dis ' 1-9" 709-7i3'
MILLER, E.C. and J.A. MILLTR (1974). The molecular biology of cancer'
P.377. Academic Press, N'Y'
MISHELL and DUTTON (1967). Immunization of dissociated spleen cell
cultures from normal mice' J' Exp' Med' 1-26: 423-444'
MUKHERJI, B. and Y. HIRSHAUT (1973). Evidence for fetal ant',igen in
sarcoma. Science lBl: 440'442'
MULLER, M., M. SCHREIBER, J. RARTENBECK and G. SCHREIBER (1972)'
Preparation of single-cel1 suspensions from normal lìver,
regenerating tiver and Morrjs hepatomas 9121 and s123tc' cancer
Res. 32: 2568-2576
NAKATA, y., T. SUESMATSU and y. SAKAM0T0 (1964). Transamìnase actjv'ities
inrapidlygrow.ingtiSsues.J.Biochem"Tokyoss:199-201.
NISHI, S. (tszo¡. Isolation and characterizatìon of a human fetal
cr-globulin from the sera of fetuses and a hepatoma pat'ient'
Cancer Res. 30: 2507-25L3'
t
I
1i.
i
NISHI, S. and H. HIRAI (1970). Radio-immunoassay of o-fetoprotein
in hepatoma, other diseases and pregnancy. Gann Mono. Cancer
Res. 14z 79-98.
NOssAL, G.J.V. and G.L. ADA (1971). In ant'igens, lymphoid ceìls, and
the 'immune response. Academic Press : N 'Y '
NOWELL, P.C. (1976). The clonal evolutjon of tumour cell populations'
Science 194: 23'28.
OKITA, K., M. GRUENSTEIN, M. KLAIBER and E' FARBER (1974)'
Local izat.ion of cr-fetoprote'in by immunofl uorescence 'in hyperp'last'ic
nodules during hepatocarcinogenes'is'induced by 2-acetyìaminofluorene'
Cancer Res. 34: 2758-2763.
0N0, T. and Y. UMEHARA (1968). Heterogene'ity of DNA po'lymerases in rat
liver and hepatomas. Gann Mono' 6: 97-107 '
oucHTERLONY, 0. (1968). Handbook of immuno-diffusjon.
HumphreY Sciences : London'
OVE,P.,J.LASZLo,M.D.JENKINSandH'P'MORRIS(1969)' I ncrea s ed
CancerDNA po'lymerase activjty in a series of rat hepatomas.
Res.29: 1557-1561.
PERoVA, S.D. (1969). Dissertation thesis. university of Moscow'
(CiteO from Perova et aL.' 1971)'
pER9VA, G.D., D.A. EL'GgRT and G.I. ABELEV (1971). ct-Foetoprotein
in rat sera after partial hepatectomy. Byvll. Eksp' Biol ' tiled'
7L: 45-46.
PIDDINGToN, S.K., J.E.NEhIELL and R.M. FOX (1976). Are foetal genes
expressed 'in human ìymphocytes. clin. Exp. Pharm. Phys' 3z 277'
drù,i
Ann Arbor.
II
r
PIHK0, H. and E. RU0SLAHTI (1973). High ìeve'l of alpha-fetoprotein
'in sera of adult mice. Int. J' Cancer 12: 354-360'
POTTER, V.R. (1969). Recent trends in cancer biochemistry: The
importance of studies on fetal tissue. Can. Cancer Conf. B; 9.
pSTTER, V.R. (tgZO). Environmentally induced nretabolic oscijlat'ions
as a challenge to tumour autonomy. In: Miami W'inter Symposium
2: 29t. North-Holland Publ. Co. : Amsterdam'
PREHN, R.T. (1967). The significance of tumour-distinctive hjsto-
compatabiìity antìgens. In: "Cross-reacting antigens and neo-
ant'igens." Ed. by J.J. Trentin, Willjams & Wilk'ins : Baltimore,
p . 105.
RAFFEL, S. (1961). Antigen-ant'ibody reactions in immunity, P.145.
Ed. S. Raffel. Appleton, Century, Crofts : N'Y'
RAFTELL, M., F. BLgMBERG and P. PERLMANN (1974). A comparison of
enzyme-active membrane antigens from two different 4-dìmethylamino-
azo-benzene induced rat hepatomas with those of adult and foetal
rat liver. Cancer Res. 34: 2300-2306'
REJTHAR, A. (1974). Detection of tumour-assocjated transpìantation
antigens of B0-IV carc'inoma on somatic cells of tumour-free
syngeneic mice: IV. Search for th'is antígen on syngeneic embryon'ic
tissues. Folia Biol. 20: 252'257 '
REYNSO, G., T.M. CHU, D. HOYYOK, E. OOHEN, T. NEMOTO, J.J. WANG' J.
CHUANG, p . GUINAN and G. MURPHY (1972) . Carcì noembryon'ic ant'igen
ìn patients with different cancers. J. Am. Med. Assoc.220z
361-365.
,'J
ì,f,rl
,r
i
!
tI
;
RUOSLAHTI, E. and M. sEppALA (1971). Studies on carc:jno-fetal proteins.
III. Development of a radio-immunoassay for a-fetoprote'in'
De¡nonstration of o-fetoprotein ifì serum of healthy hunran adults'
Int. J. Cancer B: 374-383.
RUOSLAHTI, E., M. SEPPALA, T.A. RASANEN, P. VUOPIO ANd T. HELSKE (1973)'
Alpha-foetoprotein and hepatitìs B antigen in acute hepatitis and
prìmary cancer of the liver. Scand. J. Gastroent. B: L97-202'
RUOSLAHTI, E. , H. PIHK0 and M. SEPPALA (tgZ+a). Aìpha-fetoprotein:
Immunochem'ical puri fi cation and chemi cal propertì es ' Expressìon
i n normal state and i n mal 'ignant and non-mal i gnant I i ver d'isease '
TransPlant. Rev. 20; 38-59'
RUOSLAHTI , E., H. PIHKO, M. BECKER and 0. MAKELA (1974b)' Rabb'it alpha-
fetoprotein: normal levels and breakage of tolerance \¡rith
haptenated homologous alpha-fetoprote'in. Eur' J' Immun' 5: 7-10'
RUOSLAHTI , E. , E. ENGVALL, H. JALANKO and H. PIHKO ( 1976) ' A] pha-
fetoproteìn: sequence homology with a'lbumin and attempts towards
use 'in tumour.immunotherapy . Proc. lst conf ' 0ncodeveì opmental
Gene ExPression , P.47.
SABINE, J.R. and E.H. STONEHILI- (1972). Appearance of "tumour-specific"
embryonal ant'igens in pre-tumourous tissue. Proc' Aust' Biochem'
Soc. 5: 44.
SABINE, J.R., B.J. HORTON and M.B. l,llCKS (1973). Spontafìeous tumours
in C3H.AVY and CgH.AvYtg mice: High inc'idence in the United States
and low incidence in Australia. J. Natl. cancer Inst' 50: 1237'
t242.
r
SABINE, J.R., R.S.SAVVAS, A.K.VERMA, A.þJ. MURRAY ANd E. RUOSLAHTI
(1975). Tinre course of cellular changes during spontaneous
hepat'ic carcinogenesis in mice. ClÍn. Exp.Pharm. Phys' e: 85'
SAMAAN, N., S.C.C. YEN, H. FRIESCEN and 0.H. PEARSQN (tg6A). Serum
pìacental 'lactogen levels durirlg pregnancy and trophoblastic
disease. J. cl in. Endocrinol. Metab . 26; 1303-1308.
sELL, S. and D. GORD (1973). Rat a'lpha-fetoprote'in III. Refinement
of radio-'immuno-assay for detection of one ng rat arF' Immunochem'
L0: 439-442.
SELL, S. and H.P. MoRRIS (1974). Rejationship of ra'; ar-fetcprotein
to growth rate and chromosome compos'ition of Mor¡is hepatomas'
Cancer Res. 34: 1413-1417.
SHARON, N. and H. LIS (t972). Lectins: Cell agglutinat'ing and sugar-
specifìc prote'ins. Science fzZ: 949-958'
SHEPPARD, H., M. P0LER, P. TREFTS and S. SELL (1976). Effect of A.F.P'
on immune phenomena in rats and mice. Proc. conf. Oncodevelopmental
Gene ExPression, P.46.
SHIED,8., H.P. M0RRIS and J.S. ROTH (1965). Distribution and act'ivity
of aspartate amino transferase ìn some rap'idly proliferating
tissues. J. Biol. Chem. 240; 3016-302?'
SHINOMIYA, Y., H. KAZUYA, K. MATSUMURA, A. IMAGAI^IA, Y. ENOMOTO' Y'
FUKAI and N. IT0 (1g73) . cr-Fetoprotei n duri ng hepatocarci nogenes'i s
ìn dogs treated with chemical carcinogens. Gann Mono. 1-4; 301-313'
sIL\/ER, H.R., P. G0LD, S. FEDER, S.0. FREEMAN and J. SHUSTE (1973)'
Proc. Nat. Acad. Sc'i . 7o: 526-530 '
SILVER, H.K., J. DENAULT, P. G0LD, t^l.G. THOMPS0N, J. SIIUSTER and
s.0. FREEDMAN (Lg74). The detection of or-fetoProte'in in pat'ients
wi th vi ral hepat'i t'is . Cancer Res ' 34: 244 '
SJOGREN, H.0., I. HELLSTROM, S.C. BANSAL and K. HELLSTROM (1971)'
Suggestìve ev'idence that blocking antibodies of tumour-bearing
i nd j vi dual s may be ant'igen-ant'ibody compl exes ' Proc ' Nat" Acad '
Sci . 6B: 1372-1375.
S0L0M0N, J.B. (197i). Maternal antibod.y. III. Protective effects
and immunologìcal hazard. In: "Foetal and Neonatal Inmunology"
North-Holland Publ. Co. : Amsterdâffi, P'115-138'
SOMMERVILLE, R.G. (1967). The production of fluorescent antìbody
reagents for virus diagnosis in the albino mice. I. Hyper-immune
anti-species serum. Arch. fin. d'ie Gesmate v'irusforschung ' 20:
445-4s2.
SPELLMAN, C.M. and p.F. FOTTRELL (1973). Similarities between pyruvate
kinase from human pìacenta and tumours. FEBS Letters ãZ: 2Bl'284'
STAFFORD, M. and 0.hl. JONES (1972). The presence of fetal thymidine
kinase in human tumours. Biochim. Biophys. Acta 2772 439'442'
STANISLAWSKI-BIRRENCWAIG, B. , E.J . URIEL and P. GRABAR ( 1967) '
Associatìon of embryonic antigens w'ith experimentally induced
hepat'ic les'ions in the rat. cancer Res. 27: 1990-1997.
ST0NEHILL, E.H. and A. BENDICH (1970). Retrogenetic expression: The
re-appearance of embryonaì antigens in cancer cells. Nature 228:
370-37 1.
STSNEHILL, E.H., E. B0RENFREUND and A. BENDICH (i971). Anachron'ist'ic
genetic expressìon recognjzed by the production of 'foetal atttigens
.in cancer cel I s . proc. Conf . on Embryona'l and Foetal Anti gens i n
cancer l: 85-104. AAEC : Oak Ridge.
SUGIMURA, T., K. IKEDA, K. HIROTA, M. HOZUMI and H.P. MORRIS (19664)'
Chemical enzymatic and cytochrome assays of mjcrosomal 'lract'ion
of hepatomas wjth d'ifferent growth rates. Cancer Res' 26:
17 11-1716 .
SUGIMURA, T., T. MATSUSHIMA, T. KAWACHI, Y. HIRATA ANd S. KAWABE (1966b)'
Molecular species of aldolases and hexokinases in experimental
hepatornas . Gann Mono . 1-; 143- 149 '
suzuKl, H., K. MIKI, S. IINo, H. NIWA and T.ODA (1976). Alkaline
phosphatase isoenzymes in cancer tissues of the stomach' Proc'
1st Conf. on Oncodevelopmental Gene Expression, p.58.
TANAKA, T., Y. HARANO, F. SUE and H. M0RIMURA (1967). crystallization'
characterization and metabolic regulation of two types of pyruvate
kinase ìsolated from rat tissues. J. Biochem' Tokyo 62: 7L-91'
TEMIN,H.M.(1971).MechanismofcelltransformatjonbyRNAturnour
viruses. Ann. Rev. Microbiol ' 25: 609-648'
THOMPSON, D.M.P. and P. ALEXANDER (1973). A cross-reacting embryon'ic
ant.igeninthemembraneofratsarcomacells,Whichìs
ìmmunogenic 'in the syngeneic host. Br. J. Cancer 27: 35-47'
TINc, C.C., J.R. ORTALDO and R.B. HERBERMAN (1973). Expression of
foetal anti gens and tumour-spec'i f i c anti gens 'in SV-40 transformed
cells. I. Serologìcaì ana'lys'is of the antigenic specific'ities'
Int. J. Cancer f2: 511-518'
TOMASI, T.8., R. THOMPSON, R. KELLOR and R. DATTWYLER (1976).
Immune regulatjon by alpha-fetoprotein" Proc. Lst. conf.
on Oncodeve'lopmental Gene txpression ' p'45'
I^JAHLSTROM, T. (1973). G]Ìa-spec'ific antigens in cell cultures from
rabbit bra'in, human foetal, adult brain and glìomas. Acta
Path. Microbiol. scand. (Micro et Immunol. ) 8L: 768-774'
I,üARBURG, 0 . ( 1923 ) . Versache en uberl ebendem carc'inogewebe '
B'iochem. 1 4: 317-333.
WAIìBURG, 0. (1956). 0n the origin of cells. Science 123: 309-314'
WARNER, N.L. (1976). Experimental tumour immunology. cl in. Exp'
Pharm. PhYs. 3: 293.
WATABE, H., H.HIRAI and H. SATOH (1972)' a-Fetoprotein in rats
transpl anted wi th asci tes hepatoma ' Gann 6 3 : r B9
l,,lEINHOUSE, S. (1973). Metabolism and ìsozyme alterat'ions in experimental
hepatomas. Fed. Proc. 322 216?'2167 '
I^IEINTRAUB , B .D. and S .W . ROSEN (1971) . Ectop'ic producti on of human
chorionic sommatomammotrop'in by non-trophoblastic cancers'
J. C'lin. Endocrinol . Metab.32" 94-101'
YANG, l,l.K. (1971) . Iso-accepting transfer RNAs in mammal ian
differentiated cells and tumour tissues' Cancer Res' 31" 639-643'
7AMCHECK, N., T.L. M00RE, P. DHAR and H. KUPCHIK (1972).
diagnosisandprognosisofhumand.igestiVecancer:
antigen. N. Eng. J. Med. 286: 83-86'
Immunol ogi ca'l
Carc'i noembrYon i c