J-.I:._:c... .-
1‘. ‘ v,_.
.. .

  
   

, flak

awn...

_ v

 

              

        

1". ,H4.
. s25: ..........w .
..v . ‘ ., ..
....u. Li. .H. ..

         

       

 

v ‘

. . _
\ ~4Vv§

l.‘|

fir:

P

 

 

3.. . O... 5'61..-
.

...-h -
“PK... .

.

f1 3x!

.r... .4 l 4.49. ‘nrx

1 H;V‘t¥/n.~n .
.

      

     

    

....uuinqv
mfimm

.V.» ...w

         

 

 

               

 

Itfli<umwfil hm .LH. ...».w.w..fl$.d.
‘0'“. £3"«fltu\quvttt ‘
. ...“.fl ., ...JR. ...}.Umi

h,“ .Jnuac. . Ln" 7.!

 

 

        

 

 

        

             

 

 

 

‘r|.n v'u‘o 095‘ .1. ..\ h I! \
."l\\|.1 k '1 w l \III
0. . lilvolr. 1.5.4. I J I u
‘ -l.
, o
‘ ‘1» .v.. Y ;| II
F... L.P’J l.‘ \ ....urfis I. ‘
u: nv\ I‘lllr.dr.tt.f .12
I. \ I! o‘. 1: 0| {‘1'
1 .. . ivv....Ol.\lsu ‘ ‘
.A....|hl...1.\'t Wh'7.»‘.l I . .-w‘l
f . . I1 trlolv 1’9: -ll... . I. ‘I.L
1.!ll d v.5”.- ru. «OO‘JJ JJ ... '
1’.5,I.|. _. {‘5 .. 2 I. y y
. 1 ‘x 2A1}. Q A114. . . 9 1 0 - -
0‘ V. v [I ...|4|‘ . U " ( I L'L
..- . I . x‘ N
w, . $H 3-! . I
13......14. . n! l c o q” IIH“
.| n . 1..» I!

re 1‘07...
lav. .

 

1.1.! \o,V .Ai ..
uvuw ..
.1 . 3

 

 

trues-91$

LIBRARY

Michigan State
University

 

This is to certify that the

thesis entitled
.. W- ,. / /
/” ’L /‘(/}'/f»»ILL, ZK/B/Lc 71;“ L— ; “It-" _,/ (I'L’¢'C

. , /
Cflbf/ (I’L'lbh’jZ’V (’47U [\{L-l;/‘l'u, LL flyh (L {52" -
.... a/ V
/ X}:/’21/7/ce¢7&,.
" presented by

’7 - .-
/A£L'//¢//I¢‘ /J /L:"// 6.; .1 21

has been ”accepted towards fulfillment
of the requirements for

—-_. .4 A i I = 1/ .
. I r a . (-7 4/ ”ff—H .1
M 5 degree in /’ (/LL Nit d (P I”!

 

L: (at ////LLUMW\

/ Major [Vlofessor
Date Il/ M/fi

 

 

MEMBRANE EXPRESSION OF THY—1.2 AND GMl

GANGLIOSIDE ON DIFFERENTIATING T LYMPHOCYTES

By

Christine P. Milewicz

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Microbiology and Public Health

1977

ABSTRACT

MEMBRANE EXPRESSION OF THY-1.2 AND GMl

GANGLIOSIDE ON DIFFERENTIATING T LYMPHOCYTES

By

Christine P. Milewicz

C3H mouse bone marrow cells were separated by discontinuous
BSA gradient centrifugation. Marrow cells from the 17—19%, l9—21%,
21—23%, 23-25%, 25-27% interfaces and the cell pellet were treated
with thymic factor (TF) or with Vibrio cholerae neuraminidase (VCN),
followed by anti-Thy—l.2 and anti-GM1 ganglioside antisera. Antigens

determined by anti—Thy—l.2 and anti—G cytotoxicity responses were

Ml
expressed either with TF or VCN within a thirty minute incubation.
Cells expressing these antigens after VCN or TF treatment were con-
centrated in the l9-21% layer and the 21-23% layer whereas there was
little or no antiserum cytotoxicity in the other layers. A small

amount of Thy-1.2 or could be detected by cytotoxicity on 19-21%

Gm
layer marrow cells within fifteen minutes of either TF or VCN treat—
ment. Treatment of 19-21% layer cells with TF or VCN had no effect
on anti-H-Z cytotoxicity. Pretreatment of gradient purified TF—

treated marrow cells with cholera toxin or choleragenoid (which bind

cell surface GMl) abrogated the cytotoxicity of anti-Thy-l.2 or

Christine P. Milewicz

anti-GMl without affecting anti—H—Z cytotoxicity. Pretreatment of
Thy—1.2 positive C3H thymocytes with cholera toxin or choleragenoid
greatly reduced the cytotoxicity of anti—Thy-l.2 antisera without
affecting the cytotoxicity of anti-H-2 antisera. Nude mouse spleno-
cytes, after treatment with VCN or TF, were susceptible to the lytic
action of anti-Thy-l.2 and anti-GMl. The possibility that nonspecific
autologous antibodies were responsible for anti-Thy—1.2 cytotoxicity

toward VCN-treated marrow cells was eliminated because anti—Thy-1.2

was not cytotoxic for VCN—treated AKR (Thy—1.1) marrow cells.

TABLE OF CONTENTS

Page

LITERATURE REVIEW ....... . ..... .......... ............................ l
Thy-l antigen......... ..... . ..... . ........ .......... ........... 1
Definition and Distribution..... .......... .. ..... .......1
Isolation.. ..... . ......... .. ........ . ........... . ..... ..l
Detergent solubilization........... ..... . .......... 2
Immunoprecipitation...... ..... . ........... . ........ 4

Enzyme solubilization. ..... .......... ........... ...5

Thymic factor ............................. ....... ........... ...6
Introduction ........... . ............................. ...6
Isolation of a thymic inducer substance ................. 7
Intracellular events of TF induction ................... 10
Isolation and characterization of TF .......... . ........ 11

The target cell for thymic factor....... ...... . ........ ll
Neuraminidase... .............................................. 13
Introduction.... ..... . ................................. 13
Neuraminidase treatment of erythrocytes... ............. 13
Neuraminidase treatment of lymphocytes..... ...... ......14
Neuraminidase and tumor cells.. ........ .... ........... .18
Cellular neuraminidases.................. .............. 19

ii

Cholera toxin.. ..... ................ ..... ...... ............... 20
Introduction........... ........ ....... ........ .........20
Mechanisms of action.................... ........... ....20
The receptor for cholera toxin............... ....... ...21
Choleragenoid..... ..... ......................... ....... 23
Interaction of CT with cell membranes. ................. 24

Cholera toxin as a marker for lymphoid cell

populations............... ........ . ..... ....26

INTRODUCTION TO EXPERIMENTAL REPORT ................... . ............ 28
MATERIALS AND METHODS ...... ..... .............. . .................... 30
Cells............... ............ . ............................. 3O
Thymic factor ........ . .......... ............ .................. 31
Antisera and complement ....................................... 31
Cholera toxin ................. . ............................... 32
Neuraminidase..... ..... . ..... . ..... ....... .................... 32
Absorption of anti—Thy-l. ........ .... ......................... 33
RESULTS..... ..... . ..... ....... ..... ....... ........... . ............. 34

Neuraminidase and thymic factor induced expression of

Thy-1.2 and GMl .............. ....... ..... ....... ..... ....34

Blocking of Thy-1.2 and GMl by cholera toxin ..... . ............ 42
DISCUSSION ..... . ..... . ....................... . ..................... 47
BIBLIOGRAPHY................................................. ..... .54

iii

Table

II

III

IV

VI

LIST OF TABLES

BSA gradient distribution of bone marrow cells
sensitive to thymic factor or neuraminidase ..... ...

Treatment of bone marrow cells with neuraminidase..

Neuraminidase or thymic factor treatment of nude
mouse SplenocyteSQOOOOQICO...QOOQOOQ... ........... Q

Thy—1.2 antiserum specificity for C3H marrow cells
after neuraminidase treatment........ ........... ...

Inhibition of cytotoxic action of antisera after
treatment of Thy—1.2 positive cells with cholera
toxin or choleragenoid.... ........... . ..... . .......

Decrease in the cytotoxic effect of anti—Thy—1.2
antisera after treatment of thymocytes with cholera
toxin ............ . ........ . ........................

. ...... 36

....... 37

....... 4O

....... 41

....... 44

....... 46

LIST OF FIGURES

Figure Page
1. Rate of antigen appearance on thymic
factor or neuraminidase treated marrow cells ...... 39
2. Absorption of anti—Thy—l sera with Vibrio

cholerae neuraminidase—treated marrow cells.......43

LITERATURE REVIEW

I. Thy-1 Antigen

A. Definition and Distribution. In 1964, Reif and Allen (1)
demonstrated that thymocytes and peripheral T lymphocytes and central
nervous tissue of mice shared allogeneic antigenic determinants.

This antigen system occurs in two allelic forms resulting in two
alloantigens, Thy-1.1 (9 — AKR) and Thy—1.2 (9 - C3H). These antigens
were at first thought to be mouse strain specific, however, recent
studies show that cross-reacting or identical antigenic determinants

to the Thy-l antigen exist on thymocyte or brain of all rat tissue (2).

Thy-l is a membrane-bound antigen found in large amounts only on
thymocytes and brain of mice and rats (1). It is found in smaller
amounts on peripheral thymus-derived cells, fibroblasts of mouse or
rat (3), epidermal cells (4) and cells in the peripheral and central
nervous system (5). It is not found on bone marrow or bone marrow-
derived lymphocytes (1), plasma cells (6), granulocytes, macrophages,
or red blood cells (7).

B. Isolation. Because of the membrane-bound nature of the Thy-l
antigen, different attempts at solubilization and purification of the
molecular complex have resulted in variable isolation products. The
three most widely used isolation techniques concerning the Thy-l
molecule are detergent solubilization (8), immunoprecipitation (9), and

enzyme solubilization.QO).

2

l. Detergent solubilization. Letarte-Muirhead et a1. (11)
characterized a molecule isolated from rat thymocytes by detergents,
including Lubrol-PX or deoxycholate. In high concentrations this
molecule could inhibit the cytotoxic activity of Balb/c anti-Wista
rat brain antibodies to rat thymocytes. This "solubilized" Thy-l
antigen was said to contain 25—50% of the original antigenic activity
of intact cells or membrane. Use of the solubilized undiluted antigen
to attempt inhibition of C3H anti—AKR antisera resulted in a weak but
definite inhibition of cytotoxicity. Using anti-rat brain antisera to
measure activity, the antigen isolate was characterized by sucrose
gradient centrifugation and gel filtration. The molecular weight of
the anti-Thy—l antiserum-reactive isolate was estimated to be 28,000.
A high frictional coefficient in sucrose gradient suggested that the
Thy—l molecule was asymetric, possibly containing a hydrophobic tail.

Chemical characterization of the Thy-l molecule isolated by
Barclay and Letarte-Muirhead (12) and defined by reactivity with anti—
rat Thy-l antisera showed that the Thy-l molecules from both thymocytes
and brain are glyc0protein of 25,000 molecular weight. It was
estimated that 30% of brain or thymocyte Thy-l is carbohydrate but
that the carbohydrate composition isolated from respective tissues is
not the same. It was suggested that the antigenicity of the molecule
resides in the protein moiety. In support of this hypothesis was
the finding by Barclay that the antigenicity of the molecule was
destroyed by pronase treatment (13). However, it was also found that
the antigenicity was not significantly decreased by other proteolytic

enzymes, including trypsin and papain. This may suggest that the

3

protein may be only partially responsible for antigenicity or may be
a carrier for the carbohydrate and/or other undefined antigenic units.

Further studies by this group of investigators using other
cell lines including a Thy-l positive tumor (14) and using highly con-
centrated and purified glyc0protein illustrated the reproducibility
of the isolation procedure on other Thy-1 bearing tissues. However,
the Thy-l activity isolated was oftentimes less than 25% of the origin-
ally estimated cytotoxic values. These findings and those concerning
pronase digestion of the isolated antigen may suggest a bias in the
selection procedure used for isolation of the Thy—l molecule and that
all components of the molecule essential for specificity and optimum
antigenicity within the cell membrane were not isolated by these
procedures.

In 1966 Reif and Allen (5) demonstrated that the 9 - C3H
(Thy—1.2) antigen of thymocytes was non-dialyzable and sensitive to
lipid solvents, e.g. chloroform and methanol. Noting these findings,
Esselman and Miller (15),attempted extraction of this molecule from
mouse thymocytes and brain by lipid solvents. In association with
auxilliary lipids, the antigen could inhibit the cytotoxicity of rabbit
anti-mouse brain antisera (anti-BAG). This anti-BAG inhibitory sub-
stance was located in the upper, water soluble phase of a Folch parti-
tion, and co-migrated by thin-layer chromatography with the ganglioside
G

Moreover, it was found that isolated mouse brain G ganglio-

le' le
side could also inhibit the cytotoxicity of anti-BAG. These investiga-
tors proposed that the antigenic determinant of the Thy—l antigen of

mouse brain may be associated with glycolipid material.

4

Esselman and Miller later demonstrated the specific
inhibitory capacity of isolated AKR and C3H gangliosides for anti—Thy-l
sera. As in other antigen systems, cross-reactivity was also observed
at higher levels of antigen concentration (16). Esselman and Kato (17)
demonstrated the ability of pentasaccharides derived from AKR and C3H
gangliosides to inhibit specifically their respective antisera in a
hapten inhibition assay. Cross~reacthn1was again observed when allo-
antigen was used in higher concentrations (e.g. AKR oligosaccharide with
anti-Thy—1.2).

2. Immunopreeipitation. Vitetta et al. (18) used immunopre-
cipitation methods with specific anti—Thy—l.l antiserum to murine
thymocytes to attempt solubilization of the Thy-l molecule. In
contrast to Letarte-Muirhead, it was found that deoxycholate destroyed
the antigenic activity of the Thy-l molecule. However, the estimate of
a molecular weight of 35,000 for Thy-l was in reasonable agreement with
Letarte—Muirhead.

Vitetta et al. found that 125I-labeled Thy-l or 3H—galactose-
labeled Thy—l was released from cultured thymocytes (19). The amount
of Thy-1 found in the culture medium was proportional to the amount of
time the cells were cultured. Based on the ability to label the
antigen using lactoperoxidase, and on finding molecules shed in the
culture medium, Vitetta et al. proposed that the Thy-l molecule is
peripherally located in the cell membrane. This release of soluble
fragments from the membrane of T cells has been suggested by other
investigators as a possible mechanism for T cell-mediated cytotoxicity

and stimulation of B cells (20, 21).

5

Characterization of the molecule immunOprecipitated by
Vitetta et al. demonstrated an NP—40 labile substance of lower
density than protein which, in cell culture, could not be labeled
with 3H-labeled amino acids. Based on these findings, it was proposed
that Thy-l may contain a glycolipid portion in its antigenic unit (19).
Atwell et al. (22) utilized the lactoperoxidase method of radio—
iodination of cell surface protein followed by immunoprecipitation to
obtain a non-dissociable anti—Thy-l inhibitory molecule. Mobility on
gel electrophoresis in sodium dodecylsulfate (SDS) with reference
molecules resulted in a molecular weight estimate of 60,000. Because
of the stability of the molecule in SDS, the authors suggested that
although the molecular weight is higher than previous estimates, it is
probably not a molecular conglomerate. It was suggested that the
detergents used by other investigators to extract Thy-l from membrane
may cause dissociation of the antigenic complex.

3. Enzyme Solubilization. Kuchich et al. (10) utilized a
limit papain digest of the Thy-1.2 positive 849 murine cell line.
Limited papain digestion of S49 cells resulted in solubilization of
material containing a molecular complex capable of inhibiting the
cytotoxicity of anti-Thy-l.2 antisera against S49 cells. Further in—
creased exposure of S49 cells to pronase digestion reduced the amount
of Thy-1.2 activity, indicating further degradation of the molecule.
Protease digestion of the papain-solubilized material also resulted in
a decrease in antigenicity. The authors suggested that a protein in

the released Thy—l molecule is essential for its antigenicity.

6

Later studies suggested that sialic acid participates in the
antigenicity of the molecule (23). Using papain-solubilized digests
of the anti-Thy-l inhibitory molecule, it was found that neuraminidase
destroyed this inhibitory activity. Sialic acid in isolated form
was also shown to be an effective cytotoxic inhibitor. The authors
then suggested that not only is the Thy—l molecule glyc0protein in
character but that it may also contain lipid material

In contrast to the data of Kuchich et al., Schlesinger and
Gottesfeld (24), found the effect of neuraminidase on Thy-l antigenicity
to be minimal. However, in addition to different concentrations of
enzyme and antiserum used by the two groups of investigators, it should
be noted that Kuchich et al. used papain-solubilized antigen from S49
Thy-l positive tumor cells as a substrate for neuraminidase, while
Schlesinger utilized whole thymocytes as substrate.

This summary of findings on the nature of the Thy-l molecule
illustrates that the nature of the antigen is still undetermined. The
antigen has, at various times, been prOposed to be of low molecular
weight, high molecular weight, glyc0protein, glycolipid, or a carbo-
hydrate in combination with both lipid or protein. These differences
in findings may indicate the complexity in structure of the Thy—l
molecular complex on the cell and illustrate the controversy in
utilizing different isolation techniques to define chemically a cellular

antigen.

II. Thymic Factor
A. Introduction. Evidence for a humoral factor from the thymus

has been based on the restoration of immune function to thymectomized

7

mice either by injection of thymic extracts or by insertion of thymus
grafts encapsulated in cell-impermeable Millipore chambers (25).
Evidence for immune restoration was seen in the ability of the animal
subject to produce antibodies to sheep red blood cells (26) as well
as to reject skin allografts (27). Bach et al. (28) attempted
characterization of an in vitro model of thymic extract restoration
of immunocompetence. It was postulated that the capacity of lymphoid
cells to form rosettes correlated with restoration of immunocompetence
and the ability of the cells to react with an antigen.

B. ISOZation of'a thymic inducer substance. Komuro and Boyse
(29) demonstrated the effect of an extracted thymic inducer substance
on the expression of known T cell antigens on murine spleen and bone

marrow cells. The distribution of some of these antigens is shown as

 

follows:
ANTIGEN DETECTED IN
Thy-l (9) T lymphocytes, skin, brain (1)
TL T lymphocytes in thymus only
(30)
Ly-l, Ly-2, Ly—3 T lymphocytes (31)
MSLA Thymocytes and other T cells
(32)
GIX Thymocytes of certain mouse

strains (33)

Using Thy—1.2 antigen as a marker for thymic factor inducer
function, it was found that a significant proportion (about 20%) of
those bone marrow or spleen cells settling on the 10-30% interface of
a bovine serum albumin (BSA) discontinuous gradient could be induced

to express the Thy-l antigen. Although not tested with murine marrow

8

cells, Komuro and Boyse also demonstrated the ability of the thymic
extracts to induce an increase in the expression of TL antigen on
Splenocytes (29).

Scheid et al. (34) further studied the ability of a substance
called "thymosin" to induce T cell differentiation. Thymosin,
purified by the method of A. Goldstein (35), had also been employed
in the work of Komuro and Boyse. This substance was shown to induce

the expression of TL, Thy-l, G Ly-l, Ly—2, and Ly-3 antigens on pre-

IX’
differentiated cells of mouse bone marrow. It was suggested by Scheid
that perhaps this thymic humoral substance may have a mechanism of
action similar to other isolated hormones, that is, thymosin may function
via an adenylate cyclase system, and that cyclic AMP may be a participant
in the initiation of cellular differentiation. It was shown that use
of aminophylline, a phosphodiesterase inhibitor, or insulin, a pur-
ported adenylate cyclase inhibitor, was reported to cause a decrease in
the ability of thymosin to cause expression of Thy—l antigen on marrow
cells. Scheid then suggested that any agent able to induce or inter-
fere with the production of cyclic AMP may alter the ability of thymosin
to induce antigen expression. It was reasoned that cyclic AMP may act
as a non—specific trigger of a pre—Thy-l positive cell to convert to a
Thy-l positive cell.

The thymic factor—induced rise in cyclic AMP was illustrated
by several investigators. Kook and Trainin (36) utilized a substance
isolated from calf thymus termed THF (thymic humoral factor). They
confirmed an establishment of immunocompetence (graft vs. host reaction)

with a concommitant rise in cyclic AMP after injection of THF. Bach

9
et al. reported that cyclic AMP (10-9) could be substituted for circu—
lating thymic factor in vivo to cause rosette forming cell production
in adult thymic-irradiated mice. The effect of thymic factor and
cyclic AMP were found to be synergistic when co—incubated with spleen
cells (37).

Several apparently non—Specific substances were then
found by Scheid et al. (38) to be capable of inducing the appearance
of Thy—l antigen on marrow cells from the 10-23% albumin layers of 3
BSA discontinuous gradient. All products could also cause increased
cyclic AMP levels. These substances included epinephrine, propranolol,
isoproterenol, endotoxin, poly AzU, and even calf muscle (38). Bach
et al. found that the prostaglandins PGEl and PGE2 could also increase
cyclic AMP levels and mimick the effect of thymic factor, and that
this effect could be inhibited by indomethacin, an inhibitor of pros-
taglandin synthetase (39). However, it was discovered that some of
these factors, including endotoxin, propranolol, and calf muscle were
also found to induce B cell differentiation, as measured by an increase
in complement receptor (C 3 receptor) expression (29).

G. Goldstein isolated a product from the thymus which he
named thymopoietin (TP)(40), a substance which was originally used for
study on its neuromuscular effects in the disease myasthenia gravis.

It was shown that this product was specific for T cell differentiation.
During the isolation of TP, a second polypeptide was isolated. This
polypeptide, named ubiquitin (UBIP) was found to be present in several
tissues and was able to induce both T and B cell differentiation in

vitro. Pre-incubation of marrow cells with propranolol inhibited the

10
induction of Thy—l antigen on pre—T cells or complement (C 3 receptor)
receptors on the surfaces of B cells. It was then suggested that
UBIP may be a non—specific inducer of Thy—l expression, and its
presence in several tissues may explain why non—thymic sources may
cause antigen expression (41).

C. Intracellular events of'TF induction. Storrie et a1.(42)
studied the intracellular effect of thymopoietin during thymocyte
differentiation. Induction of both TL and Thy-l antigen expression by
thymopoietin was blocked by actinomycin D, camptothecin, or cordycepin;
inhibitors of DNA replication did not block induction. The authors
concluded that the induction of Thy—l and TL antigen expression may
require DNA transcription and RNA translation but not DNA replication.
This was in support of earlier work by Komuro and Boyse (29) indicat-
ing that thymic factor induction could be inhibited with cycloheximide.

However, Rotter and Trainin (43) found a 50-100% increase in
the uptake of thymidine using phytohemagglutinin (PHA) or concanavalin
A (con A) as mitogens for spleen cells preincubated with THF, a pur-
ported T cell specific substance. Basch and Goldstein, using thymo-
poietin (44) observed no measurable effect of spleen cells to con A or
PHA. Using bone marrow cells, previously shown by TP induction of
Thy-1 antigen to contain precursor T cells, Goldstein observed no
thymidine uptake in response to con A or PHA incubation (4). This
discrepancy in findings may have been a consequence of different
purification procedures of the thymic extracts used by the two

investigators.

 

11

D. Isolation and characterization of'TF. Schlesinger et al. (46)
have been able to synthesize partially the amino acid sequence of
thymopoietin as isolated and determined by G. Goldstein (47). This
peptide, corresponding to a portion of the sequenced natural polypep—
tide displayed a similar selectivity of action for differentiating T
cells, although only displaying 3% of the activity of whole isolated
thymopoietin.

E. The target cell for thymic factor. Till and McCullough (48)
originally proposed the existence of a pluripotential stem cell of
bone marrow. Using sublethally irradiated mice, they demonstrated
recolonization of spleen by cell clones of injected bone marrow. These
colony forming units (CFUs) were shown to be capable of differentiating
into granulocytic or lymphocytic cell lines. It was suggested that
perhaps this pluripotential stem cell of marrow was the target cell for
thymic factors, that it migrated to the thymus and was triggered to
differentiate into a lymphoid cell.

Studies of marrow cell subpopulations using density gradient

methods delineated at least two more cell subsets. One subset is a
relatively small population of proposed "re-circulating" T cells,
demonstrated as possessing the Thy—l antigen (49), capable of initiat-
ing graft vs host responses (30), and able to respond to PHA (51). El-
Arini and Osoba (52) located a much larger population of marrow cells
which they termed "progenitor T cells". This population was weakly
susceptible to the cytotoxic activity of anti—Thy-l serum, and gave
rise to immunocompetent cells in the spleens of irradiated isogeneic
recipients within 15 days of transplantation. After isopycnic centrifu-

gation, this cell was shown to be slightly denser than a CFU (stem) cell

12
(1.050-1.059 gm/cm3) and lighter than a bone marrow Thy—l positive
cell (1.069 gm/cm3). The density of this cell was estimated at 1.063
gm/cm3. The presence of this weakly Thy-l positive cell was also
demonstrated by Loor and Roelants (53). Work by Turner et al. (54)
illustrated that cells of stem cell and pro—thymocyte densities (1.050
and 1.063 gm/cm3, respectively) fall within the 21—23% BSA layers of
a discontinuous gradient. Thus those cell populations shown by Boyse,
Sheid, and Goldstein to be inducible by various thymic extracts con-
tained both pro-thymocytes and stem cells and the target cell for the
thymic factor preparation was undefined.

Indirect evidence suggests that the thymic factor—inducible
cell from bone marrow is committed to thymocyte differentiation. The
ability of pro-thymocytes to be induced by various substances (thymic
factors, CAMP, poly A:U, endotoxin) is suggested by Boyse as a non-
specific response of pre—programmed cells to become Thy-l positive (55).
In addition, slightly Thy-l positive cells can be detected in nude mice
(mice congenitally lacking a thymus, termed nu/nu)(56) and approximately
35% of nu/nu marrow cells from the 10—23% interface of a BSA gradient
can be induced to become Thy-1 positive after thymic factor incubation.
It was reasoned that the thymus is not essential for differentiation
of stem cells, that the cells susceptible to thymic factors are "pre-
programmed" to become T cells (29). The thymic factor target cells
may not be stem cells and may instead be progenitor T cells. It is not
possible to eliminate the suggestion that these "pro—thymocytes" may
instead be "pro-immunocytes" (57), able to be triggered into functional
T or B cells depending on their environment. This may explain the

existence of cells with both B and T cell markers. The target cell(s)

13
for thymic factor has not been defined. The step-wise differentiation
from marrow stem cell to pre- and post-thymic lymphocytes to peripheral
T lymphocytes may display a differential sensitivity of these cells to
thymic factors.
III. Neuraminidase

A. Introduction. Treatment of cells with Vibrio cholerae neura-
minidase (VCN) results in the removal of 2-3 and 2-6 glycosidically-
linked sialic acid from mucopolysaccharides, glyc0protein and glycolipids
on the cell surface (58,59,60). This enzyme treatment of the cell causes
less negative charge on the cell surface (61). Several alterations can
be induced as a result of this neuraminidase treatment. These include
rendering the cell more easily phagocytosed (62), enhancing monocyte
phagocytic activity (63), inhibition of viral hemagglutination (64),
and changing homing characteristics of injected cells (65). The results
of these experiments on neuraminidase treatment of cells suggested that
cells may possess the genetic information for antigens not displayed
phenotypically. Neuraminidase will alter the antigenicity of several
cell lines, including erythrocytes, marrow cells, lymphocytes, and tumor
cells.

B. Neuraminidase treatment of’erythrocytes. Sialic acid is an
indigenous component of the cells of many different animal species (66,
67). Currie and Bagshawe (68) originally proposed that some antigens
on the cell surface may be masked by a "sialo—mucin coat". Treatment
of erythrocytes with neuraminidase releases greater than 90% of the
sialic acid (69). Barth and Singla (70) noted a differential effect
of VCN treatment on the antigenicity of sheep, horse, chicken, and

goat red blood cells. Neuraminidase treatment significantly reduced the

l4

antigenicity of both sheep and horse red blood cells, but did not
effect goat or chicken erythrocytes. The authors proposed that this
differential effect may function in determining inter—species
variations in antigenicities. This is supported by the finding that
autologous transfusion of neuraminidase-treated erythrocytes results
in their rapid removal from the circulation (69). The use of 51Cr—
labeled cells illustrated their removal by the liver and spleen (71).

The effect of neuraminidase-treated red blood cells on cell
surface interactions was studied by Schlesinger and Galili (72).
Neuraminidase treatment of sheep red blood cells resulted in the
formation of stable (non-dissociable) rosettes with peripheral T
lymphocytes or thymocytes. The authors proposed that the formation of
stable rosettes is probably due to a decrease in charge separation
between the two cells or to exposure of new cross-linkage determinants.

C. Neuraminidase treatment of lymphocytes. Several groups of
investigators have studied the effect of neruaminidase treatment on the
exposure of molecules on lymphocyte membranes. Schlesinger and Amos
(73) found an increase in guinea pig serum cytotoxicity on neuramini-
dase-treated C57BL (Thy-1.2 positive) murine thymus cells. A similar
effect was observed using neuraminidase-treated spleen cells. An
eightfold increase was observed in the effect of anti-Thy-l.2 anti-
serum on these cell populations. In addition, up to 30% of bone
marrow cells became Thy-1.2 positive as defined by cytotoxicity. The
amount of TL antigen on thymocytes increased after neuraminidase
treatment, but there was no measurable increase on marrow cells. This

illustrated a possible differential effect of the enzyme on antigen

15

exposure. From these experiments Schlesinger proposed that neuraminic
acid may cover the Thy—l molecule in its natural state and inhibit its
expression and/or detection. It was concluded that neuraminic acid is
probably not a part of the antigenic determinant(s) of the Thy—l mole—
cule, nor is it an antigenic component of the guinea pig serum-detect—
able surface molecule. It was suggested that the TL antigen may not
be "hidden" by sialic acid.

Barton and Rosenberg (74) arrived at a somewhat different explana-
tion of the effect of neuraminidase on lymphocyte cell surfaces. It
was suggested that only the more peripheral sialyl compounds are sus-
ceptible to neuraminidase treatment. The sialolipids, e.g., may be
more deeply embedded in the cell membrane and may not be susceptible
to the action of VCN.

Reisner supported the suggestions of Schlesinger that neuramini—
dase may be exposing "hidden" membrane components to the antisera (75).
The similar reactivity of non—treated thymus cells and VCN—treated
peripheral lymphocytes was a result of the increased amount of sialic
acid released from peripheral cells. Neuraminidase-treated thymocytes
release significantly less neuraminic acid than other lymphoid cells.

Other authors confirmed the increased cytotoxic effects of a
variety of antisera to neuraminidase-treated lymphocytes. Rosenberg
and Rogentine (76) showed that normal human sera contained lytic anti-
bodies directed against autologous human lymphocytes that had been
pre—treated with VCN. The sera used were highly concentrated. These
investigators later found that pre—incubation of 8-D galactosyl residues

with normal human sera abrogated the cytotoxicity of the sera to

16
autologous VCN-treated lymphocytes (77). Within the HL-A system, the
specific antigen reactions were found to be unchanged after VCN treat-
ment. In some cases, however, an increase in the weaker antigen
reactions or the appearance of a new familial antigen (78) was observed.
The original work of Schlesinger and Amos using heteroantisera to VCN-
treated lymphoid cells was tested in other systems. Schlesinger and
Gottesfield described naturally occurring antibodies in guinea pig
serum directed against neuraminidase-treated murine cells (79).
Rosenberg et a1. discovered a similar toxicity of normal rabbit anti-
sera to neuraminidase-exposed determinants on human cells (80). This
heterospecific cytotoxicity can be absorbed with untreated mouse or
human tissue (81). It has therefore been suggested that a non-specific
diffusely occurring antigen may be unmasked in addition to those
specific antigens unmasked after neuraminidase treatment, e.g. as in
the case of the "T antigen" which is revealed on red blood cells after
neuraminidase treatment. This phenomenon was originally demonstrated
in 1947 by Burnet and Anderson (82), who noted the spontaneous agglutin-
ation of human red blood cells after VCN treatment.

Recently it has been suggested that it is not the exposure of
"cryptic" antigens that increases antiserum cytotoxicity after neuramin—
idase treatment, rather that other factors play a significant role. A
possible explanation for increased serum cytotoxicity was increased cell
fragility. However, enzyme—treated cells were shown not to be more
fragile and therefore more susceptible to antiserum treatment. For
example, human lymphocytes were shown to give negative cytotoxicity

reactions with many antisera to the human HL-A system, survive well in

l7
tissue culture, and respond as well to lectin stimulation as untreated
cells (78).

Since the fixation of complement by IgG antibody may depend on
the attachment of two molecules (83), it has been prOposed that neura—
minidase treatment may make the simultaneous attachment more feasible
by exposing additional complement receptors. Complement activation
would then be more probable. It was also suggested that a site essen-
tial for completion of the complement pathway may be further uncovered
with neuraminidase treatment (84). Because of this pr0posal some
investigators state that VCN-treated cells use more complement and
therefore appear more sensitive to antisera (84). Others suggest that
perhaps the complement is used more efficiently (75). It was suggested
by these authors that absorption studies would be helpful in clarifying
the controversy of exposure of cryptic antigens. Ray and Simmons (85)
found no difference in absorptive capacity of anti—H-2 antisera with
VCN—treated lymphocytes.

Some authors have prOposed that these naturally occurring antibodies
are directed against neuraminidase that has absorbed to the surface of
the lymphocyte (86). Rosenberg and Rogentine found that heat-inactivated
neuraminidase did not increase cellular cytotoxicity, nor did neuramini—
dase treatment of cells in the presence of sialic acid. Moreover, in
absorption experiments using large numbers of cells untreated with
enzyme, some cytotoxic activity could be removed (76). It has been
suggested that heat inactivation may not be an appropriate control for
testing neuraminidase adherence to cells, as the active site may serve

in cellular attachment (87).

18

D. Neuraminidase and tumor cells. In 1964, Kirby et al. (88)
proposed that the pregnant female fails to reject the fetal tropho-
blast as an allograft because of the presence of a protective coat.
This masking of the trophoblast cells protects the placental cells
of fetal origin from interaction with and rejection by maternal
lymphocytes. Trypsinization and subsequent in vitro culture of tropho-
blast cells allows for gross cytolysis by allogeneic or maternal
lymphocytes. A fibrinoid coat was said to be removed with trypsiniza-
tion, thus unveiling the masked antigens. This fibrinoid coat was
later shown to be an amorphous sialomucin (89). These findings led to
a theory pr0posed by Currie and Bagshaw that cancer cells maintain a
"protective coat" which inhibits cytotoxic reactivity with the immune
system of the host. This coat was proposed to be composed of "peri-
cellular sialomucins" (68).

Histochemical studies of tumor cells have illustrated that many
tumor cell lines contain a high content of acid mucopolysaccharides
(90,91,92). Moreover, Defendi and Gasic observed a significant increase
in surface sialomucin of hamster embryo cells undergoing malignant
transformation by polyoma virus (93).

As with erythrocytes and lymphocytes, the treatment of tumor cells
with neuraminidase resulted in increased lysis of the cells in hetero—
antisera (94). Neuraminidase treatment also increased the ability of
lymphoid cells or fibroblasts to stimulate allogeneic or syngeneic
lymphocytes in the one-way mixed lymphocyte culture (95,96). Several
investigators have demonstrated increased immunoreactivity and de—

creased growth of neuraminidase-treated tumor inocula (97,98,99).

l9
Hughes et al. demonstrated that this increased immunoreactivity to the
tumor after injection of VCN—treated cells could be abrogated by con—
comitant injection of anti-Thy—l antiserum (94).

Simmons et al. demonstrated the induction of immunity to a
tumor in a syngeneic system. An established methylcholanthrene-induced
fibrosarcoma could be ablated by the injection of neuraminidase-treated
cells of the same tumor. Injection of enzyme-treated cells of other
tumors would not result in the disappearance of the fibrosarcoma, thus
suggesting the specificity of the treatment (100). The possible applica-
tion of this system to clinical use has been discussed (101).

E. Cellular neuraminidases. The finding of several mammalian
cellular neuraminidases complicates the previous studies using exoge-
nous enzyme and simultaneously supports the theory that neuraminidases
may play a role in cell membrane dynamics. Carabelli (102) originally
demonstrated the presence of neuraminidase in liver, brain, spleen, and
intestine of the rat. It was later characterized in tumor cells (103)
erythrocytes, and leukocytes (104).

Neuraminidase activity has been estimated in several areas of the
cell, including microsomes (103), mitochondria (105), and in the soluble
fraction of tissue homogenates (106). Studies using rat brain illus-
trated the presence of at least two distinct cellular neuraminidases,
one lysosome-derived and one soluble (106). The enzymes display differ—
ent pH optima and ion sensitivities and perhaps different substrate
specificities. Gielin suggests that soluble neuraminidases prefer low
molecular weight substrates, e.g. neuraminlactose, whereas high molecu—

lar weight g1ycoproteins or glycolipids (in gangliosides) are preferred

20
by lysosomal neuraminidases (107). This suggestion is supported by
the findings of Schengrund et al. who found a plasma membrane-marker
enzyme associated neuraminidase which was also associated with ganglio-
side (108). This enzyme is highly active for these membrane—bound
gangliosides (109). The specificity for this substrate was suggested
by Tettamanti, the enzyme not showing as high affinity for sialo—
glyc0proteins (109). However, Yohe and Rosenberg (110) suggest that
sialolipid or sialoglyc0protein are both susceptible to membrane—bound
neuraminidase.

It has been suggested that this enzyme may participate with other
cell membrane-bound enzymes and susceptible molecules in membrane
alteration and/or cell-cell communication (107, 110).

IV. Cholera Toxin

A. Introduction. Cholera toxin is the specific exotoxin produced
by Vibrio cholerae which is active on epithelial cells of the small
intestine causing acute diarrhea (111). In its purified form the toxin
is a protein of molecular weight 84,000, composed of six subunits (112).

B. Mechanism of'Action. The ability of cholera toxin to stimulate
intracellular adenylate cyclase activity was originally investigated
using an isolated section of rabbit or dog ileum and measuring the
diarrheogenic effect of cyclic AMP levels of intestinal cells. It was
found that the toxin was rapidly fixed to these cells (113). Others
found that the activity of cholera toxin, as mediated by cyclic AMP
could be studied in other systems, including adipocytes (114), liver
cells(115), liver plasma membrane (116), adrenal cells (117), and

lymphocytes or fibroblasts in tissue culture (118).

21

Cuatrecasas (119) and Vaughn (120) studied the effect of cholera
toxin on the cyclic AMP—mediated lipolytic response of adipocytes.
Alloxan, an inhibitor of hormone-stimulated adenylate cyclase activity
in cells, would inhibit the lipolytic response caused by cholera toxin.
Inhibitors of prostaglandin synthesis, indomethacin and sodium sali-
cylate, did not block this lipolytic effect. In addition, and despite
suggestions to the contrary (121), it was found that inhibitors of RNA
and protein synthesis, including actinomycin D, cycloheximide, and
puromycin were also reported to be ineffective in blocking the lipo-
lytic effect of the toxin (119).

Van Heyningen (114) first noted that the required incubation time
for measuring the maximum lipolytic response caused by cholera toxin
was not dependent on the presence of the toxin. Washing the toxin from
the cells after ten minutes of incubation still resulted in the maximum
lipolytic response. Moreover, studies by Cuatrecasas revealed a one
hour delay in the lipolytic response (119). It was postulated that the
toxin-cell membrane receptor complex must initially undergo alteration
until it is active, and that this new active complex stimulated the
biologic reSponse, as mediated by adenylate cyclase.

C. The receptor for cholera toxin. Cuatrecasas (122) studied the
binding activity of 125I-labeled cholera toxin. At very low concentra-
tions (less than lO-IOM) of cholera toxin, greater than 50% of the
toxin bound to liver cell membranes. This could easily be replaced by
unlabeled toxin. It was found that the binding of the toxin was directly
proportional to the amount of liver cell substrate provided. The same

activity of the purified toxin could be found using isolated adipocytes.

22

It was found that certain g1yc0proteins could partially inhibit
the interaction of 125I—labeled toxin with hepatocytes. These included
fetuin and rat liver membrane glycoprotein in high concentrations.
Tests using simple sugars to attempt inhibition of the toxin were not
successful at the concentrations tested and in the form used. Some of
the sugars tested were D-glucose, D—mannose, D-fucose, D—glucosamine,
D-galactosamine, N—acetyl D-glucosamine, N—acetyl D-galactosamine, and
N-acetyl neuraminic acid. High concentrations of D—galactose effected
a small amount of inhibition. Use of a Clostridium perfringens
neuraminidase digest of fetuin resulted in a thirty percent increase in
inhibitory capacity. Studies employing purified gangliosides showed
that they were the most potent inhibitors of 125I—labeled cholera toxin
binding to hepatocytes. Simple glycosphingolipids such as glucose- and
galactose-ceramidesinamanot inhibitory. Sequential addition of terminal
galactose or N-acetylgalactosamine increased the inhibitory capacity.
Adding sialic acid to the galactose of a galactose-glucose-ceramide
moiety lead to a minimum of a twenty-fold increase in inhibition of

125

I-labeled cholera toxin binding. ganglioside was found to be

GMl
the most potent inhibitor and was found to cause inhibition at nanogram

concentrations. Addition of N—acetylneuraminic acid to GM1 ganglioside

decreased inhibition by fifty percent. Neuraminidase digestsofGDla

and GT1 gangliosides improved their inhibitory abilities. It was stated
that the relative order of potency of binding inhibition by gangliosides
was: GMl (20 ng/ml), GDla(O'7 ug/ml), GM2(1 ug/ml), GTl(l'8 ug/ml),

GL4(17 pg/ml), GM3(50 ug/ml) (122). It was also noted that predigestion

of the liver cell membranes with neuraminidase or with B-galactosidase

23

enhanced the binding of 1251 labeled toxin. Further evidence for the
binding of cholera toxin to ganglioside was obtained with membrane
extraction studies. Toxin binding to hepatocyte membranes was greatly
decreased by pre—treatment of the membranes with choloroform-methanol
solvents by methods known to be effective in extracting gangliosides.
Moreover, material in the water soluble phase of these extracts were
inhibitive of toxin binding to untreated hepatocyte membranes (122).

From these data it was suggested by Cuatrecasas that the natural
receptor for cholera toxin may be a ganglioside-containing material.
Further evidence to support this hypothesis was the finding that
exogenous gangliosides, especially GMl ganglioside, is a potent in-
hibitor of cholera toxin binding to cells and that pre-incubation of
GMl ganglioside with hepatocytes increases binding and biological
activity of toxin. This was confirmed in reports by Pierce and by
Ikflmgrenem al. (123, 124). Hollenberg et al. (125) later supported
this finding by utilizing an SV40-transformed cell line which lacks
chemically detectable G , G

M1’
‘<10% of 125I-labeled toxin.

and GDla gangliosides, and which bound

D. Choleragenoid. Finkelstein and LoSpalluto, while attempting
to purify the active cholera toxin (choleragen) also purified a bio-
logically inactive derivative of the choleragen molecule with a molecu-
lar weight of 58,000. The two molecules are antigenically identical.
The derivative now called "choleragenoid" is said to lack the biologi—
cally active subunit of the molecule (111).

125

Cuatrecasas found that the affinity of the I-choleragenoid

molecule for the liver cell membrane was identical to that of the

24
active toxin. Binding studies showed that isolated gangliosides, pre—
incubated with the choleragenoid, would inhibit the biologically
inactive toxin from binding to the cell membrane, and that pre-incuba—
tion of hepatocytes with ganglioside caused enhanced binding of the
labeled molecule to the cells. Thus this biologically inactive substance
retained binding properties virtually identical to that of the active
cholera enterotoxin. Choleragenoid is specific in its ability to in-
hibit the lipolytic effect of subsequently added cholera toxin to
adipocytes. It is not able to inhibit the lipolytic effect of
epinephrine or glucagon (126).

Choleragen is, therefore, prOposed to be composed of two subunits,
a 58,000 dalton "binding" subunit, which reacts with cell membranes,
and another 38,000 dalton "active" subunit, which is responsible for its
biological activity (127).

E. Interaction of cholera toxin with the cell membrane. Bennett
reported that the choleragen molecule forms a detergent—stable complex
with adenylate cyclase (128). The 125I-labeled cholengen is found to
be associated with adenylate cyclase activity when activated by the
molecule. It was proposed that it is unlikely that the molecule enters
the eytOplasm in soluble form, as < 1% of the 125I~labeled toxin could
be found in the cytoplasm (129).

Craig and Cuatrecasas (130) used fluorescein-labeled cholera toxin
to demonstrate temperature-dependent changes in cell membrane fluidity
of rat lymph node cells on cholera toxin binding. The membrane-bound

cholera toxin, when warmed from 0°C to 37°C for 20-30 minutes changed

from a diffuse surface fluorescence to a polar fluorescence (capping).

25
This capping phenomenon was partially inhibitable by colchicine (60%),
Con A (80%), NaN3 (96%), and NaF (78%). Antibody to the fluorescent-
labeled cholera toxin inhibited the capping. Because the capping
phenomenon was demonstrable and because it was inhibitable by anti-
cholera toxin antibodies it was suggested that cholera toxin may be at
least a bivalent molecule. This suggestion was supported by the find—

ing that G

Ml-bound agarose beads will bind holera toxin, to which

lymphocytes will then bind.

Cuatrecasas pr0posed that the membrane complex of cholera toxin
and its probable ganglioside receptor undergoes formation of an inert
toxin—ganglioside complex which is converted to an active state after
an approximate twenty to thirty minute lag phase. This active unit then
undergoes lateral diffusion within the membrane until it reacts with
adenylate cyclase to form a complex. The adenylate cyclase is thus
activated (131).

In support of this proposed mechanism of action are the following
findings:

1) The extreme persistence of the biological
effects of cholera toxin (132).

2) The irreversible binding of 125I-labeled
cholera toxin to cell membranes (119).

3) Suggestions that the toxin may become
incorporated into membrane phospholipid

(133).

4) The hydrOphobic prOperties of the active
subunit of the toxin (134).

The lag period is then the time required for the inactive ganglio-
side—cholera toxin complex to laterally diffuse to, react with, and

activate adenylate cyclase. This suggestion is in accord with the

26
pr0posed mechanisms of action of other hormones in regards to membrane
fluidity (120).

Gill and King (135) propose that the activation of adenylate
cyclase by cholera toxin is not through a one-to-one association of
the toxin with cyclase, but through an NAD—dependent enzymatic reaction.
They demonstrated that intact choleragen or its active subunit can
catalyse the hydrolysis of NAD to ADP-ribose and nicotinamide. As
previously demonstrated with diptheria toxin (136), the active subunit
of cholera toxin is proposed to mediate NAD-dependent ADP-ribosylation
of an acceptor, in this case L-arginine methyl ester or a related amino
acid residue, as contained in a protein. This acceptor protein may be
adenylate cyclase or an associated molecule (137).

Flores (138) demonstrated that the addition of NAD to toxin—
hepatocyte incubation mixtures increased the activity of adenylate
cyclase twofold. The addition of NAD eliminated the latent period re-
quired to demonstrate adenylate cyclase activity in the experiments
by Cuatrecasas. Further support for the possible NAD-mediated activa-
tion of adenylate cyclase was the finding that cholera toxin inhibits
protein synthesis in mouse spleen cells as measured by 3H-uridine
incorporation (139). This finding was analogous to information previ-
ously obtained on diptheria toxin. It is known that diptheria toxin
causes inhibition of protein synthesis as a result of NAD—dependent
ADP—ribosylation of the protein receptor molecule "elongation factor
II(EFII)" (140).

F. Cholera toxin as a marker for lymphoid cell populations.

The work by Cuatrecasas and others has demonstrated the high affinity

27
of cholera toxin for specific ganglioside material. Cholera toxin
was not found to bind strongly to oligosaccharides in the concentra-
tions used, although the inability of cholera toxin to bind to cell
membrane glycoprotein has not been excluded (141).

Studies by Esselman and Miller implicated a possible antigenic
similarity between the undefined Thy—l molecule and GMl ganglioside (16).
In possible correlation with this finding is the study by Stein-Douglas
(142) showing the value of employing anti-ganglioside antisera in
detecting thymus—derived (Thy-l positive) cell p0pulations.

Sultzer and Craig (139) demonstrated the ability of cholera toxin
to inhibit specific mitogen stimulation on T cells. This indicated
that cholera toxin receptors may be found on T cell populations. How—
ever, in much higher concentrations, cholera toxin was shown to inhibit
the effect of a specific B cell mitogen, thus indicating a lower
affinity of the toxin for this cell type. The studies by Sultzer and
Craig may indicate a high affinity receptor for cholera toxin on T cells.
The receptor may be glycolipid and/or glyc0protein in composition and
may be associated with the Thy-1 molecule. The possible ability of
cholera toxin to bind to B cells may indicate a similar receptor for
the toxin on this cell population which may exist in lower concentrations

or in a less available form.

INTRODUCTION TO EXPERIMENTAL REPORT

Enzyme treatment of intact cells has been used a a means of
releasing exposed labile components and of exposing hidden or cryptic
determinants and receptors. Treatment of lymphocytes and erythrocytes
with Vibrio cholerae neuraminidase (VCN) has been reported to expose
an antigen (or antigens) which can be detected by human or animal sera
(76,77,81,l43,l44,l45). Some of these workers (143,145) reported a
VCN exposed antigen determined by autologous antisera which was shared
by erthyrocytes and all types of leucocytes. Schlesinger (24) reported
that the treatment of mouse lymphoid cells with VCN resulted in the
increase of a guinea pig serum cytotoxicity receptor but did not change
the amounts of exposed H—2 antigen or Thy—l antigen. The cytotoxicity
of normal guinea pig serum for VCN-treated murine tumor cells was
apparently the result of exposure of glyc0peptides containing terminal
B—galactosyl residues (94).

Vibrio cholerae neuraminidase does not hydrolyze sialic acid from
cell surface gangliosides even though some gangliosides are VCN-labile
after isolation (74). The gangliosides are apparently deeply localized
in the intact surface of some cell types and are sterically inaccessible
to VCN. The sialic acid of GMl ganglioside is not hydrolyzed by VCN

even in isolated form (74). Hakomori (146,147) has shown that antiserum

28

29
to a glycolipid found in fetal red cell membrane did not cause
hemolysis even though the glycolipid was externally exposed (i.e. it
may be radiolabeled using specific cell surface labeling techniques).
Treatment of these cells with neuraminidase, however, expresses the
glycolipid antigen and renders the cells susceptible to anti-glycolipid
antisera. Glycolipids may be deeply localized in the intact membrane
where they may be sterically inaccessible to VCN or to antibodies. GMl
ganglioside has been identified as the cell surface receptor for
cholera toxin (148,149). Cholera toxin is composed of a subunit which
imparts biological activity and subunits (called choleragenoid) which
bind the toxin to the cell surface by binding GM1(150)'

It has previously been demonstrated that Thy-l antigen is exposed
on marrow stem cells after one hour of treatment with thymic factor
(151). The finding that C3H brain GMl ganglioside was capable of in-
hibiting the cytotoxicity of anti—Thy—l.2 antiserum led to the proposal
that GMl ganglioside is structurally related to the Thy—l antigen
(15,16,152). In view of the rapid appearance of Thy-l antigen on marrow
stem cells after thymic factor treatment it is suggested that Thy—1
may be expressed on marrow cells by either thymic factor or Vibrio
cholerae neuraminidase treatment and that these antigens may be

blocked by treatment with cholera toxin or choleragenoid.

MATERIALS AND METHODS

Cells. Bone marrow cells or thymocytes were harvested from
C3H/ST or C3H/HeJ mice at age 4—7 weeks (Jackson Laboratories, Bar
Harbour, Maine). Nude mice (nu/nu) on a BALB/c background were a
generous gift of Dr. M. Pienkowski of Michigan State University.
Thymuses were trimmed of all nodal tissue and were gently teased with
forcepts in Eagle's medium to free cells from connective tissue. The
thymocytes were then disaggregated using 18 and 25 gauge needles. An
average of l x 108 nucleated cells were obtained per thymus. Bone
marrow cells were harvested from the femurs and tibias using a 27
gauge needle and were suspended in Eagle's medium. Eight to ten mice
would normally yield 1 x 107 nucleated cells with 85% vilability.
These cells were then subjected to centrifugation in a discontinuous
gradient of bovine serum albumin utilizing layers of l9,21,23,25,27
and 35% BSA. The BSA was prepared from a 35% stock solution (65 m1
tris buffer and 35 g of dialyzed and lyophilized BSA, Grade V, Sigma
Chemical Co.) according to the method of Dicke, et al (153). The
NaCl concentration of the buffer was adjusted such that the osmolarity
of 17% BSA was 310 mOsmols. After centrifugation, distinct cell
layers were observed at the interfaces. Cells removed from the

respective interfaces had the following composition as observed by

30

31
light microscopy. The 17-19% BSA interface contained dead cells in
low numbers and cellular debris. The 19-21% BSA interface contained
primarily mononuclear cells of 99% vilability as did the 21-23% BSA
interface (although the latter also contained some polymorphonuclear
cells). The 25—35% BSA interface contained a diverse pOpulation of
99% viable marrow cells, including erythrocytes. The pellet contained
predominantly erythrocytes and debris. All cell layers were found to
contain only 1 to 3% Thy—1.2 positive cells when tested by cytotoxicity
with anti—Thy-l.2 antiserum and complement. Cells in the interfaces
of the 19-21% and 21-23% BSA layers were used for experimentation
either as a pooled suspension or as separate populations.

Thymic factor. Thymic factor (TF), Fraction 3, was prepared from
calf thymus according to A. Goldstein, et al. (35). Thymic factor
(Fraction 3) used in initial experiments was a generous gift of Dr.
Allyn Rule. Thymic factor was also prepared according to the protocol
of G. Goldstein (40) and was used as a mixture containing thymopoietin
I and II, as well as ubiquitin. Fifty ug of TF was added to 106 cells
in a final volume of 1.1 ml and incubated at 37°C for 60 minutes.

Antisera and complement. AKR anti-C3H (anti—Thy-1.2) antisera
was prepared according to Reif and Allen (1) or obtained from Litton-
Bionetics, Inc. Kensington, Md. (Lots 231-55—1 and 231-55-2). Anti-GMl
antisera was prepared by multiple injections of Dutch Belt rabbits
with isolated C3H brain GMl ganglioside using Freund's complete
adjuvant and BSA essentially according to Naikai, et al. (154). The

antisera were thermally inactivated at 56°C for thirty minutes and

adsorbed with C3H erythrocytes, hepatocytes, marrow cells and agarose.

32

Specificity of the antisera for GMl

tion. Anti—H—Z antisera, containing antibodies to H—2 specificities

was determined by complement fixa—

11 and 23 directed against K and I regions of H—Zk, was a gift of Drs.
Chella S. David and Donald C Shreffler. Guinea pig complement (Grand
Island Biological Co., Grand Island, N.Y.) was adsorbed with agarose.
The cytotoxicity of the antisera with added complement was determined

by trypan blue exclusion. Marrow cells or thymocytes (106 cells in 0.1
ml Eagle's medium) were first treated with 0.05 ml of heat inactivated
antisera for thirty minutes; 0.05 ml of adsorbed guinea pig complement
(diluted 1:3) was then added and the final mixture incubated for thirty
minutes at 37°C. Trypan blue (0.05 ml) was added and cells were counted
in an eosinophil counter.

Cholera toxin. Initial assays involving use of cholera toxin were
with a purified filtrate of the toxin supplied by the Cholera Advisory
Committee of the NIH, prepared from the procedure of R.A. Finkelstein
(155). Later experiments were conducted using a purified cholera toxin
filtrate obtained from Schwarz—Mann & Co. (Orangeburg, N.Y.) made accord—
ing to the preparation protocol of Finkelstein. Choleragenoid was the
generous gift of R.A. Finkelstein (150). Unless otherwise noted, 0.1 mg
of cholera toxin or choleragenoid was added to 106 cells in a total
volume of 1.1 ml. The cells and toxin were then washed twice by centri-
fugation in Eagle's medium before adding antisera.

Neuraminidase. Vibrio cholerae neuraminidase was purchased from
Koch—Light Laboratories (Burkinghampshire, England, 500 units/ml).
Marrow cells (106ml) of either the 19«21 or 21—23% fraction of BSA in

Eagle's medium were treated with 25 units of VCN for 45 minutes at 37°C.

33
Cells were then washed twice from medium by centrifugation. An aliquot
of the active enzyme preparation was inactivated by heating to 70°C
for thirty minutes in order to test for nonspecific antigen induction.
The specificity of neuraminidase was determined by incubation with
G and G l gangliosides. GDla ganglioside was found to be completely

Dla M

hydrolyzed to ganglioside and no activity was observed with GM

Gm 1
ganglioside. Neuraminidase activity was also confirmed by reaction with
neuraminlactose. No neuraminidase activity was determined in the
cholera toxin preparation using GDla ganglioside as substrate.
Absorption of'anti-Thy-I. Neuraminidase-treated C3H marrow cells
(5 x 106/0.15 ml) were added to 0.05 ml of undiluted anti-Thy-l.2 serum
and incubated at 4°C for two hours. Cells were centrifuged from the
serum and serial dilutions of the supernatant were tested for residual

cytotoxicity for C3H thymocytes. An equivalent number of untreated

marrow cells were used for absorption as a control.

RESULTS

Neuraminidase and thymic factor induced expression of’Yhy-1.2 and
CM]. Bone marrow cells, collected from the interfaces of discontinuous
BSA gradients, were washed twice and treated with thymic factor or
neuraminidase. Treated cells were washed twice and the appearance of

Thy-1.2 and GM antigens was assessed by cytotoxicity after treatment

1
with antisera and complement. The greatest amount of conversion of

TF or of neuraminidase treated stem cells to Thy-1.2 positive cells
was observed in the 19-21% interfaces of BSA (Table I). A significant
number of Thy-1.2 positive cells was also observed in the 21—23%
interface. Similarly, cells demonstrating the greatest cytotoxicity
after anti-GMl treatment were in the 19-21% interface. Cells in the
21-23% interface were also slightly positive with respect to G

M1'
Little or no conversion to Thy-1.2 or GMl positive cells was observed
for cells in other BSA layers after treatment with thymic factor or
neuraminidase.

Neuraminidase treated marrow cells of the 19-21% layer were
subjected to lysis by anti-Thy-l.2 antisera prepared in our laboratory

or to commercially prepared anti-Thy-l.2 ascites fluid (Table II).

These sources of anti—Thy—l.2 antibodies showed similar activity.

34

35
Treatment with neuraminidase followed by either normal mouse serum or
by rabbit antivBSA antisera did not result in increased killing of
marrow cells. Minimal cytotoxic action of these antisera was observed
with non—enzyme treated cells. Marrow cells incubated with neuramini—
dase in the presence of its specific inhibitor, sialic acid, showed
minimal killing when subsequently incubated with antiserum (Table II).
A similar product inhibition of the enzyme was observed in experiments
involving anti—GM1 antiserum. These results indicate that the changes
observed are due to the specific enzymatic action of neuraminidase.
To insure that the increase in antigenicity was not due to a non—
specific mitogenic effect of neuraminidase, the enzyme was inactivated
by heat tratment and subsequently incubated with stem cells. Antisera
were then added and no significant cytotoxicity was observed above
complement controls (Table II). Neuraminidase treatment under the
conditions of the assay did not increase the lysis of cells with anti-
H-2 antisera and complement (Table II). These data indicated that VCN
treatment did not change the quantity of lytic complement receptors,
or create additional antigens on the cell surface specified by anti-
H-2 sera.

The rate of conversion of marrow cells was estimated by treating
cells for various times with thymic factor or neuraminidase (Fig. 1).
Figure l-a, illustrates that the increase in cytotoxicity of anti-
Thy-l.2 or anti-GM1 ganglioside antisera was proportional to the
length of incubation of thymic factor with the cells. Under the
conditions of the assay some antivThy-l.2 and anti~G cytotoxicity

Ml

appeared after only 15 minutes of incubation. This cytotoxicity

TABLE I

BSA GRADIENT DISTRIBUTION OF BONE MARROW CELLS

SENSITIVE TO THYMIC FACTOR OR NEURAMINIDASE

 

PERCENT DEADa

 

 

 

 

Thymic Factorb NeuraminidaseC
BSA Layer
Interface Anti-Thy-1.2 Anti-GMl Anti—Thy-l.2 Anti-GMl
17-19% 0 2 0 0
19—21% 20 32 29 34
21-23% 10 5 9 4
23—25% 0 5 0 0
25-27% 0 6 0 0
Pellet 0 0 0 0

 

a Cytotoxicity given as percent dead is corrected for complement
controls which were usually between 5% and 10% dead. The
average results of three experiments are shown. There was less
than 15% difference between individual experiments.

C3H cells (106) were incubated for 45 min at 37°C with thymic
factor (50 ug) in a total volume of 1.1 ml. Cells were washed
twice after treatment.

C3H cells (106) were incubated for 30 min at 37°C with 25 units
of Vibrio cholerae neuraminidase in a total volume of 1.1 ml.
Cells were washed twice after treatment.

36

TABLE II

TREATMENT OF BONE MARROW CELLS WITH NEURAMINIDASE

 

 

NEURAMINIDASE PERCENT DEAD
TREATMENTa ANTISERA MARROW CELLS
. b d
+ Anti—Thy-l.2 34(31-37)
+ Anti-Thy—l.2c 26(21-29)
+ Normal Mouse Serum 2(0-3)
+(Sial.)e Anti—Thy-l.2 5(0—9)
— Anti—Thy-l.2 (0-3)
+(7o°c)f Anti—Thy—1.2 3(0—3)
+ Anti-GMl 30(26—34)
- Anti-GMl 4(2—6)
+ Anti—BSA 1(0—3)
+(Sia1.)e Anti-G 5(3-6)
Ml
+ Anti-H-Z 40(35-45)
- Anti-H—2 42(38—43)

 

C3H bone marrow cells from the 19-21% BSA layer were
incubated with VCN using conditions described in Table I.

Anti—Thy-l.2 serum prepared in our laboratory (see
Materials and Methods).

Anti-Thy-l.2 ascites fluid obtained from Litton Bionetics.
Data given is corrected for complement control and is the
average of at least three experiments with range in

parenthesis.

Sialic acid (100 ug) was added to incubation buffer in a final
volume of 1.1 ml.

Neuraminidase was heated at 70°C for 30 min.

37

38
reached higher levels, however, after 30 minutes of incubation. Ex—
posure of cells to neuraminidase also resulted in a significant
increase in cytotoxicity of the antisera within 15 minutes of treatment
(Fig. l—b). Similar maximum levels of antisera cytotoxicity was
obtained for either neuraminidase or thymic factor treated cells.
Anti—BSA antisera showed minimal cytotoxicity over the period of the
assay. Cell death in untreated controls or in controls containing
complement only generally increased from less than 5% at 15 minutes
between 10% and 20% over the 60 minute period.

It has been reported that Thy-1.2 antigen could be induced to
appear on nude mouse splenocytes utilizing thymic factor (156). The
action of neuraminidase on these cells was determined as shown in Table
III. Neuraminidase increased the anti-Thy-l.2 susceptibility of nude
mouse splenocytes from a background level of 3% to maximum of 10%.

This increase was comparable to the increase induced by thymic factor.
Furthermore, the cytotoxicity of anti-GMl ganglioside antiserum was
also increased from background level of 0% to a high of 48%. This
increase was comparable to the increase induced in nude mouse spleno-
cytes with thymic factor (Table III).

The allotypic specificity of anti-Thy—l.2 antisera was demonstrated
because of reports that neuraminidase treatment of cells exposes a
nonspecific antigen found on many cell types (see discussion). AKR
(Thy—1.1) bone marrow cells of the 19—21% layer were treated with
neuraminidase in parallel with C3H cells and then incubated with anti-
Thy—l.2 antisera and complement (Table IV). Anti—Thy-l.2 antisera was

not cytotoxic for VCN treated AKR marrow cells whereas it was cytotoxic

om

.wsflaafix unmemHanu pom vmuomuuou
mucmafiuwaxm oBu mo mwmum>m mnu mum aBOCm mama .mummaucm

AoN\HV <mmuauem no . Aos\av Hzouauem . Aos\av ~.Husse

Ifiucm :uHB vmumouu was conmmz mums maamu Bouume mmmvficfiamusma
no nouomw CHE%:u zufis coaumnsocfi Houw< .mHHoo Souuma voodouu

 

 

 

 

 

 

 

 

 

 

 

Immmvflafiamusmc Ho Monomm caskau so mucmumCaam amwflucm mo Comm .N marmwm
¢gcsmus=h
on m— o cm on m— o
.‘flflflflIflflflflflflflflflflblllllfllfflfllllll}. JfiflH lflflflfinflflflflllfinflflflflflleo
(mm .35 n <wm -35 \
_l I
o\ 3 >512." .a
u .8 m
N. 352.. m
' I N
I.
till I loe_mw
Ill I _ .20 :5 W
. -.ca 1 -. I
.20 .. zu> m u: < _

 

low

39

TABLE III

NEURAMINIDASE OR THYMIC FACTOR TREATMENT OF NUDE MOUSE SPLENOCYTES

 

 

TREATMENTa ANTI SERUM PERCENT DEAD
None Anti—Thy—l.2 3
Neuraminidase Anti-Thy—l.2 10
Thymic Factor Anti-Thy-l.2 12
None Anti—GM1 0
Neuraminidase Anti—GMl 48
Thymic Factor Anti-GM1 35

 

a Neuraminidase and thymic factor were added to 10
splenocytes using conditions described in Table I
except that 75 units of VCN were used. Data given
is the average of two determinations.

40

TABLE IV

THY—1.2 ANTISERUM SPECIFICITY FOR C3H MARROW CELLS

AFTER NEURAMINIDASE TREATMENT

 

 

 

PERCENT DEADb
Neuraminidasea Sera AKR C3H
+ NMS 0 0
+ Anti-Thy—1.2 0 l7
- NMS 0 0
- Anti-Thy-l.2 9 0

 

a VCN was used with marrow cells of the 19-21% interface as
described in Table I.

Average data from three experiments. The range of percent
dead was less than 15%.

41

42
for VCN treated C3H cells. This finding supports the conclusion that
Thy-1.2 antigen was exposed by neuraminidase treatment of C3H marrow
cells. Neuraminidase—treated marrow cells were used to absorb anti-
Thy—l sera to demonstrate an increase of Thy—l antigen (Fig. 2).
Starting with a dilution of serum which gave approximately 50% cyto-
toxicity for thymocytes, the neuraminidase-treated marrow cells serially
decreased the cytotoxicity in relation to untreated marrow cells.

Blocking of'Ihy—1.2 and GM] by cholera toxin. After incubation
of pooled cells from the 19—21% and the 21—23% BSA interfaces with thy-
mic factor, approximately 20% of the population became Thy—1.2 positive
as measured by cytotoxicity. If cells treated with thymic factor were
washed and subsequently treated with cholera toxin, which binds cell
surface GM1 ganglioside (148,149) an abrogation in the cytotoxic effect
of anti-Thy-l.2 was observed (Table V). The sequence of treatment of
cells with thymic factor and cholera toxin was important to inhibit
cytotoxic action. Treatment of marrow cells with thymic factor and
subsequent incubation of cells with cholera toxin blocks the ability
of anti—Thy—l.2 antisera and complement to kill cells. Treatment of
cells in the reverse order (cholera toxin, then thymic factor) does
not cause a decrease in antiserum cytotoxicity (Table V). Cholerage-
noid, which is the GMl binding subunit of cholera toxin (150) also
exhibited the same ability to block the cytotoxicity of anti—Thy-1.2
antisera.

The ability of cholera toxin to inhibit the cytotoxic effect of
specific antisera with cells was also illustrated using anti—G

Ml

ganglioside antisera (Table V). Treatment of TF-treated marrow cells

50'!

PERCENT DEAD

o/

O/Oflx

 

 

Figure 2.

l

I I I If
0 32 64 I28 256 5|2

RECIPROCAL DILUTION

 

Absorption of anti-Thy—l sera with Vibrio cholerae
neuraminidase-treated marrow cells. C3H marrow
cells used for adsorption were obtained from the
21 and 23% BSA layer. Anti-Thy-l.2 serum was ad-
sorbed with untreated marrow cells or with
marrow cells treated with neuraminidase . Each
antiserum was titered against thymocytes starting
at a dilution which would give 50% cytotoxicity.

43

TABLE V

INHIBITION OF CYTOTOXIC ACTION OF ANTISERA AFTER TREATMENT OF

THY-1.2 POSITIVE CELLS WITH CHOLERA TOXIN OR CHOLERAGENOID

 

PERCENT DEAD MARROW CELLSb

 

EXPERIMENT EXPERIMENT
Sequence of

 

Treatmenta Antisera I II
TF — Anti—Thy-l.2 19 22
TF CT Anti-Thy-l.2 0 3
- CT Anti-Thy-l.2 o 2
CT TF Anti—Thy-l.2 18 2o
TF cc Anti-Thy-l.2 2 3
TF - Antl-GMl 18 18
TF CT Anti-GMl 0 l
- CT Anti-GMl O 2
TF - Anti-BSA O
TF CT Anti-BSA
- - Anti-H-2 38 42
CT - Anti-H-Z 35 40
CG - Anti-H-2 34 39

 

a Conditions of thymic factor treatment is given in Table I.
Cholera toxin or choleragenoid treatment (0.1 mg/lO6 cells
in 1.1 ml total volume) was performed for 15 min at 37°C.
Cells were washed twice after treatment.

b C3H bone marrow cells of the 19—21% and 21-23% BSA inter—
faces were pooled for these experiments. The average of
duplicate analysis for each experiment is given.

44

45
with either cholera toxin or choleragenoid resulted in a blocking of
the cytotoxicity of anti-G.Ml sera. Anti-BSA antisera did not have a
cytotoxic effect after treatment of marrow cells with thymic factor.
Further, the inhibitory effect by cholera toxin or choleragenoid was
not observed with anti-H—2 antisera, thus illustrating the specificity
of the treatment.

The blocking of anti-Thy-l.2 antiserum by pretreatment of Thy—1.2
positive cells with cholera toxin was further studied with C3H thymo—
cytes. The cytotoxic effect of anti-Thy—l.2 antiserum decreased after
pretreatment of thymocytes with cholera toxin (Table VI). The block—
ing, which was observed at several dilutions of antiserum, resulted
in a diminiution of cytotoxicity of 36% to 48% at antiserum dilution
of 1/100, and at antiserum dilution of 1/200 the diminutions varied
from 22% to 30%. As with marrow cells, the cytotoxic action of anti—

H—2 antisera on thymocytes was not inhibited by cholera toxin.

TABLE VI
DECREASE IN THE CYTOTOXIC EFFECT OF ANTI—THY—l.2 ANTISERA

AFTER TREATMENT OF THYMOCYTES WITH CHOLERA TOXIN

 

PERCENT DEAD THYMOCYTES

l——

 

Cholera EXPERIMENT EXPERIMENT EXPERIMENT
Toxina Anti—Thy-l.2 I II III
b
+ 1/100 37 27 19
(36) (48) (37)
- 1/100 73 75 56
C
+ 1/200 9 37 -
(30) (22)
- 1/200 39 59
+ NMS(l/20) o 0 —
( 2) ( 0)
- NMS(l/20) 2 o
+ Anti-H-Z 85 - -
- Anti-H—Z 84

 

a Cholera toxin (100 ug) was incubated with 106 C3H
thymocytes at 37°C for 30 min. Cells were washed and
anti-Thy-l.2 serum and complement were added at the
dilutions shown.

b Percent of cells killed by anti-Thy—l.2 treatment, value
in parenthesis is the decreased observed.

c Not determined.

46

DISCUSSION

The appearance of T—cell differentiation antigens on marrow or
splenic stem cells has been observed by treatment of these cells with
various substances including thymic factors, epinephrine, propranolol,
and cyclic AMP (34,38,156). Although the usual incubation time for
antigen induction has varied from one to two hours it was found that
Thy-1.2 antigen expression begins after as little as fifteen minutes
of thymic factor treatment (Fig. 1.). The cytotoxicity of anti-Thy—
l.2 antiserum increases to one half the maximal level after thirty
minutes to reach essentially maximal cytotoxicity after one hour of
treatment with thymic factor. Furthermore, the similar appearance of
antigens reactive with anti—G antisera was observed. It has been

Ml

previously preposed that GM

reactive with Thy—1 antigen (15,16,152).

l ganglioside is identical to or cross-

The appearance of Thy—1.2 or GMl on TF—treated marrow cells
could be due to: 1. de novo biosynthesis of antigen, 2. modifica-
tion of existing compounds through anabolic or catabolic reactions, or
3. expression of sequestered or cryptic antigen through membrane
changes or rearrangement. The rapid rate of appearance of Thy-1.2
(fifteen minutes) suggests that the appearance may be due to modifica-
tion of existing compounds or expression of a cryptic antigen rather
than de novo synthesis. This hypothesis has been tested using neuramin—
idase in conjunction with anti-Thy-l.2 and anti—GM1 antisera.

47

48

Neuraminidase has been reported to expose antigens specified
by autologous antibodies, histocompatability antigens on lymphocytes,
tumor antigens and a number of other antigens on a variety of cell
types. Neuraminidase has also been reported to expose cholera toxin
receptor (GMl ganglioside) on liver cells (122) and adrenal cells
(157). The reported experiments have shown that Vibrio cholerae
neuraminidase can expose or cause the expression of Thy-1.2 and GMl
antigens on bone marrow stem cells which have been separated on BSA
gradients. The expression of QMl and Thy—1.2 with neuraminidase
treatment occurs at a rate similar to expression induced by thymic
factor (Fig. 1.). Minor differences were observed in the relative
rate of appearance of Thy-1.2 with neuraminidase or thymic factor
treatment after fifteen minutes of incubation but the maximum amount
of cytotoxicity observed at sixty minutes was similar regardless of
the means of expression.

Interpretation of data involving neuraminidase treatment of cells
is complicated by at least two factors. First, neuraminidase may bind
to cells in a specific way giving rise to a "new" antigen on the cells
which may be recognized by preformed anti—neuraminidase antibodies in
sera. Second, neuraminidase treatment may expose antigens which are
common to all cells and which are recognized by autologous antibodies.
Sedlacek and Seiler (158) reported that VCN binds irreversibly to
human peripheral lymphocytes after brief VCN treatment. VCN remains
bound after four washings and was observed using anti—VCN antisera
and FITCvconjugated goat anti-rabbit Ig. It appears likely that the

binding involves the active site because heat—denatured VCN does not

49

bind to cells. This type of binding could account for an "autologous"
antibody binding to VCN treated cells observed in many human and
animal antisera. In the present study the possibility that VCN
binding to cells was responsible for increased cytotoxicity was
eliminated because cytotoxicity was observed only in the 21 and 23%
BSA and, more importantly, VCN treated AKR (Thy-1.1) marrow cells
were not killed by anti—Thy-l.2 antisera.

The possibility that VCN treatment exposes determinants that
are recognized by autologous sera has been raised by several investiga-
tors. It has been reported that VCN reveals determinants which are
recognized by autologous, undiluted normal human sera (77,143).
The same determinants appeared on all human cell types which they
tested, including erythrocytes, B and T lymphocytes, monocytes,
leukemic lymphocytes and cells of lymphoid lines. Similar autologous
antibodies have been observed for various mouse strains (159). These
normal mouse sera, in concentrated form, exhibited cytotoxicity for
every murine cell line tested. It is proposed that the Thy-1.2 and
GMl antigenicity of VCN treated bone marrow cells was not due to a non-
specific determinant, for several reasons: 1. Antisera used in the
experiments were generally more diluted (at least 1/20 or 1/40) than
autologous antisera usually used to detect nonspecific VCN exposed
antigen: 2. Normal mouse sera, (collected from bone marrow donor
mice) which was used for controls, did not have non-specific cytotoxi-
city for either marrow cells or VCN treated cells, and the cytotoxicity
of anti—H—Z antisera did not increase after VCN treatment; 3. The

cytotoxicity of anti—Thy-l.2 or anti-G was specific exclusively for

M1

50

VCN or thymic factor—treated cells of the l9v21% and 21«23% BSA layer
and not for VCN or thymic factor-treated cells of other BSA layers;
and 4. Anti—Thyvl.2 antisera retained allogeneic specificity for
VCN treated cells of the Thy-1.2 allotype.

The possibility has been raised (81) that VCN treatment of
cells might cause a greater susceptibility of the cells to complement
action. Increase in cytotoxicity of antisera for VCN-treated cells
has been observed for some antigens which were already present on the
cell surface. The reported experiments show the appearance of cyto-
toxicity of antisera for VCN marrow cells which did not previously
express antigenic specificity. Comparable cytotoxicity results were
obtained from either VCN treatment or treatment with thymic factor.
Furthermore, it was concluded that the VCN treated marrow cells were
not more susceptible to complement lysis, because under the conditions
of the assay, lysis by anti—H—Z antisera was not increased.

It has been proposed that nude mouse splenocytes contain some
Thy-1.2 bearing cells even though these cells are not susceptible to
lysis by anti-Thy-1.2 antisera (159). The results support the presence

of Thy—1.2 antigen (as well as ) on these cells because the

GM1
antigens were expressed after brief neuraminidase treatment or after
incubation with thymic factor (Table V). It is possible that Thy-1.2
antigen is sequestered on nude mouse splenocytes and C3H marrow stem

cells in the same way and the expression of Thy-1.2 on marrow stem

cells and on nude mouse splenocytes proceeds in a similar manner.

51

Schlesinger (24) has proposed that the quantity of Thy—1.2 could
not be increased on murine splenocytes, thymocytes or lymph node
cells by neuraminidase treatment. If the Thywl.2 antigen on these
cells is maximally expressed, neuraminidase treatment would have no
effect. Alternatively, using mixed populations of cells may obscure
an increase in a small population of cells. In this report, a fraction
of bone marrow cells was used which was enriched in stem cells and which
does not normally express T cell differentiation antigens. Thus it was
possible to observe increases in Thy-1 positive cells which make up a
small fraction of the total population of bone marrow cells.

Based on evidence that cholera toxin binds to GMl ganglioside
(148,149) and that it binds specifically to murine lymphocytes (161,
162), it was possible to predict that pretreating anti-GMl susceptible
lymphocytes with cholera toxin would abrogate the cytotoxicity of anti—
GM1 antisera. This proposal has been confirmed by incubating thymic
factor treated marrow stem cells with cholera toxin before addition
of anti-GMl serum (Table V). The cytotoxicity of anti—Thy-l.2 was also
blocked by pretreatment of Thy—1.2 positive cells with cholera toxin
or choleragenoid. The blocking effect of cholera toxin was also con-
firmed in experiments with thymocytes (Table VI). The finding that
cholera toxin and choleragenoid, which bind specifically to GM1 ganglio—
side, can block the cytotoxicity of anti—G serum as well as anti—

Ml

Thy-1.2 serum suggests that GMl and Thy«l.2 are immunologically related
and supports previous findings. It should be noted, however, that the
possibility that the blocking of anti—Thy—l.2 sera occurs through steric

hindrance of Thy—1.2 antigen cannot be eliminated with these experiments.

52

The properties of anti—GM1 antisera with respect to murine
lymphocytes parallel anti—Thy—l.2 activity in several ways. 1. Anti-
GMl and anti—Thy—l.2 antisera were only slightly cytotoxic for bone
marrow cells but became significantly cytotoxic for these cells after
the cells were incubated for brief periods with thymic factor or
neuraminidase. 2. These antisera were primarily cytotoxic for
neuraminidase or thymic factor—treated bone marrow cells derived from
the 19-21% and 21—23% BSA layer of a discontinuous gradient. 3. Anti-
GMl and anti—Thy-l.2 were not cytotoxic for nude mouse splenocytes,

unless these cells were pretreated with neuraminidase or thymic factor.

4. The cytotoxicity of anti-G or anti—Thy—l.2 could be blocked by

M1
pretreatment of thymic factor-treated marrow cells (or thymocytes) with
cholera toxin which binds GMl ganglioside.

It should be noted that GM1 ganglioside from lymphoid sources has
not been fully identified although it co—chromatographs with brain GMl'
The water soluble glycolipids (containing GMl) of lymphocytes are a
complex mixture of gangliosides and large oligoglycolipids and often
contain glucosamine and fucose rather than galactosamine found in brain
gangliosides. Barton and Rosenberg (74), have shown that Vibrio cholerae

neuraminidase does not hydrolyze the sialic acid of pure G ganglio—

Ml
side and that the gangliosides of the intact cell surfaces are deeply
localized and sterically inaccessible to VCN. They prOposed that
sialic acid released from intact cells is from surface sialoglyc0-
proteins rather than gangliosides.

The chemical nature of Thy—1.2 is not completely resolved.

Evidence presented here of blocking by cholera toxin supports the

53
glycolipid nature of the antigen, but reports have appeared presenting
evidence for protein nature of the antigen (10,23). It is possible
that Thy—1.2 antigenicity resides in oligosaccharides which may be
attached to membrane sphingolipid and protein. The cytotoxicity of
anti—Thy-l.2 and anti-GMl antisera after neuraminidase treatment may
be due to either glycolipid or glyc0protein antigens similar in
nature to Thy-1.2.

The similarity in the rate of appearance of anti—Thy—l.2 and
anti—GMl reactive determinants on either VCN or thymic factor-treated
marrow cells suggests that the thymic factor-induced appearance of
Thy-1.2 and GMl may proceed in a similar manner. Neuraminidase
activity has not, however, been detected in the thymic factor prepara—
tion used in this study. It is possible that there is endogenous
cell surface neuraminidase which is activated by thymic factor. Based
on the evidence presented herein, it is proposed that Thy-1.2 and GMl
antigen are present on marrow stem cells in a cryptic state where they
may be rapidly expressed through membrane rearrangement or by altera-

).

tion of a Thy-1.2 precursor (e.g. GDla

BIBLIOGRAPHY

10.

ll.

BIBLIOGRAPHY

Reif, A.E. and Allen, J.M.V. 1964. The AKR thymic antigen and
its distribution in leukemias and nervous tissues. J. Exp. Med.
120:413-433.

Douglas, T.C. 1972. Occurence of a theta—like antigen in rats.
J. Exp. Med. 136:1054-1062.

Stern, P. L. 1973. Theta alloantigen on mouse and rat fibro—
blasts. Nature (New Biol.) 247:76-78.

Scheid, M., Boyse, E.A., Carswell, E.A., Old, L.J. 1972.
Seriologically demonstrable alloantigens of mouse epidermal cells.
J. Exp. Med. 135:938-955.

Reig, A.E. and Allen, J.M. 1966. Mouse thymic iso-antigens.
Nature 209:521.

Raff, M.D. 1969. Theta isoantigen as a marker of thymus derived
lymphocytes in mice. Nature 224:378-379.

Aoki, T. Hammerling, U., deHarven, E., Boyse, E.A. and Old, L.J.
1969. Antigenic structure of cell surfaces. J. Exp. Med. 130:
979.

Schwartz, B.D. and Nathenson, S.G. 1971. Isolation of H—2
alloantigens solubilized by the detergent NP—40. J. Immunol.
107:1363—1367.

Vitetta, E.S., Bianco, C. Nussenzwig, B. and Uhr, J.W. 1972.
Cell surface immunoglobulin IV. Distribution among thymocytes,
bone marrow cells, and their derived populations. J. Exp. Med.
_l§6:181-193.

Kuchich, U.N., Bennett, J.C., and Johnson, B.J. 1975. The protein
nature of the Thy—1.2 alloantigen as expressed by the murine
lymphoblastoid line S-49.1TB.2.3. J. Immunol. 115:626—630.

Letarte-Muirhead, M., Acton, R.T. and Williams, A.F. 1974. Pre—

liminary characterization of Thy-1.1 and Ag-B antigens from rat
tissues solubilized in detergents. Biochem. J. 143:51—61.

54

12.

13.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

55

Barclay, A.N., LetarteéMuirhead, M., and Williams, A.F. 1975.
Purification of the Thy—l molecule from rat brain. Biochem. J.
151:699-706.

Barclay, A.N., Letarte—Muirhead, M., Williams, A.F. and Faulkes,

R.A. 1976. Chemical characterization of the Thy-1 glyc0proteins
from the membranes of rat thymocytes and brain. Nature 263:563-

567.

Morris, R.J., Letarte-Muirhead, M. and Williams, A.F. 1975.
Analysis in deoxycholate of three antigenic specificities
associated with rat Thy—1 molecule. Eur. J. Immunol. 55282-285.

Esselman, W.J. and Miller, H.C. 1974. Brain and thymus lipid
inhibition of antibrain-associated theta-cytotoxicity. J. Exp.
Med. 139:445-450.

Miller, H.C. and Esselman, W.J. 1975. Modulation of the immune
response by antigen reactive lymphocytes following cultivation
with gangliosides. J. Immunol. 115:839-843.

Esselman, W.J. and Kato, K. 1976. Studies on the antigenic
nature of murine Thy—1 differentiation antigen. Fed. Proc.
(abstr.) .22‘1643-

Vitetta, E.S., Boyse, E.A. and Uhr, J.W. 1973. Isolation and

characterization of a molecular complex containing Thy-l antigen
from the surface of murine thymocytes and T cells. Eur. J. Immunol.
_3:446-453.

Vitetta, E.S., Uhr, J.W. and Boyse, E.A. 1974. Metabolism of
H—2 and Thy-l alloantigens in murine thymocytes. Eur. J. Immunol.
_4:276—282.

Hartmann, K.U. 1970. Induction of a hemolysin response in_vitro;
interaction of cells of bone marrow origin and thymic origin. J.
Exp. Med. 132:1267-1278.

Bloom, B.R. 1971. Ig_vitro approaches to the mechanism of cell-
mediated immune reactions. Adv. Immunol. .13:102.

Atwell, J.L., Cone, R.E. and Marchalonis, J.J. 1973. Isolation
of theta antigen from the surface of thymus lymphocytes. Nature
(New Biol.) 241:251-252.

Johnson, B.J., Kuchich, U.N. and Maurelli, A.T. 1976. Studies
on the antigenic determinants of the Thy—1.2. alloantigen as ex—
pressed by the murine lymphoblastoid line S—49.l TB.2.3. J.
Immunol. .116:1669-l672.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

56

Schlesinger, M. and Gottesfeld, S. 1971. The effect of
neuraminidase on the expression of cellular antigens. Trans.
Proc. 3:1151.

Trainin, N. and Small, M. 1970. Studies on some physiochemical
prOperties of a thymus humoral factor conferring immunocompetence
on lymphoid cells. J. Exp. Med. 132:885-891.

Small, M. and Trainin, N. 1967. Increase in antibody-forming

cells of neonatally thymectomized mice receiving thymus extract.
Nature 216:377-379.

Trainin, N. and Linker—Israeli, M. 1967. Restoration of immuno-
logic reactivity of thymectomized mice by calf thymus extracts.
Cancer Res. 215309-313.

Bach, J.F., Papiernik, M., Levasseur, P., Dardenne, M., Barvis,
A. and LeBrigand, H. 1972. Evidence for a serum factor secreted
by the human thymus. Lancet _2:1056—1058.

Komuro, K. and Boyse, E.A. 1973. Infvitro demonstration of thymic
hormone in the mouse by conversion of precursor cells into
lymphocytes. Lancet 13740-743.

Boyse, E.A. and Old, L.J. 1969. Some aspects of normal and ab-
normal cell surface genetics. Ann. Rev. Genet. §;269.

Itakura, K., Hutton, J.J., Boyse, E.A. and Old, L.J. 1972.
Genetic linkage relationships of loci specifying differentiation
alloantigens in the mouse. Transplantation .13:239-243.

Shigeno, N., Hammerling, U., Arpels, C., Boyse, E.A. and 01d, L.J.
1968. Preparation of lymphocyte-specific antibody from anti—
lymphocyte serum. Lancet 2:320.

Stockert, E., 01d, L.J. and Boyse, E.A. 1971. The G system:
A cell surface alloantigen associated with murine leu emia virus;

implications regarding chromosomal integration of the viral genome.
J. Exp. Med. 133:1334.

Scheid, M.P., Hoffmann, M.K., Komuro, K., Hammerling, U., Abbott,
J., Boyse, E.A., Cohen, G.H., Hooper, J.A., Schulof, R.S. and
Goldstein, A.L. 1973. Differentiation of T cells induced by
preparations from thymus and by nonthymic agents. J. Exp. Med.
138:1027-1032.

Goldstein, A., Guha, A., Zatz, M., Hardy, M. and White, A. 1972.
Purification and biological activity of thymosin, a hormone of
the thymus gland. Proc. Natl. Acad. Sci. (USA) 6231800-1803.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

57

Kook, A.I., and Trainin, N. 1974. Hormone—like activity of a
thymus humoral factor on the induction of immune competence in
lymphoid cells. J. Exp. Med. 139:193—207.

Bach, M.A., Bach, and Bach J.F. 1973. Studies on thymus
products VI. The effect of cyclic necleotides and prostaglandins

on rosette—forming cells; interactions with thymic factor. Eur.
J. Immunol. .3:778-783.

Scheid, M.P., Goldstein, G., Hammerling, U. and Boyse, E.A. 1975.
Lymphocyte differentiation from precursor cells in_vitro. Ann.
N.Y. Acad. Sci. 249:530-540.

Bach, M.A., Fournier, C. and Bach, J.F. 1975. Regulation of
theta antigen expression by agents altering cyclic AMP levels and
by thymic factor. Ann. N.Y. Acad. Sci. 249:316-327.

Goldstein, G. 1974. Isolation of bovine thymin: a polypeptide
hormone of the thymus. Nature 247:11-14.

Goldstein, G., Scheid, M., Hammerling, U., Boyse, E.A., Schle—
singer, D.H. and Niall, H.D. 1975. Isolation of a polypeptide
that has lymphocyte—differentiating properties and is probably
represented universally in living cells. Proc. Natl. Acad. Sci.
(USA)_12:11—15.

Storrie, B., Goldstein, G., Boyse, E.A. and Hammerling, U. 1976.
Differentiation of thymocytes: evidence that induction of the

surface phenotype requires transcription and translation. J.
Immunol. 116:1358-1362.

Rotter, V. and Trainin, N. 1975. Increased mitogenic reactivity
of normal spleen cells to T lectins induced by thymic humoral
factor (THF). Cell. Immunol. 163413-421.

Basch, R. and Goldstein, G. 1975. Antigenic and functional
evidence for the in_vitro inductive activity of thymopoietin
(thymin) on thymocyte precursors. Ann. N.Y. Acad. Sci. 249:290-
299.

Basch, R.S. and Goldstein, G. 1975. ThymOpoietin—induced

acquisition of responsiveness to T cell mitogens. Cell. Immunol.
29:218-228.

Schlesinger, D.H., Goldstein, G., Scheid, M.P. and Boyse, E.D.
1975. Chemical synthesis of a peptide fragment of TPII that
induces selective T cell differentiation. Cell 53367-370.

Schlesinger, D.H. and Goldstein, G. 1975. The amino acid
sequence of thymopoietin II. Cell _5:361—365.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

58

Till, J.E., McCulloch, E.A. and Siminovitch, L. 1964. A
Stochastic model of stem cell proliferation based on growth of
spleen colony—forming cells. Proc. Natl. Acad. Sci. (USA) 51:29.

Osmond, D.G. and Nossal, G.J.V. 1974. Differentiation of
lymphocytes in mouse bone marrow. Cell. Immunol. 135117.

Yoshida, Y., and Osmond, D.G. 1971. Graft vs. host activity of
rat bone marrow, marrow fractions and lymphoid tissue. Transplan-
tation £25121.

Osmond, D.G. and Yoshida, Y. 1971. Blastogenic transformation

in lymphocyte-rich fractions of guinea pig and rat bone marrow.

In_Proc. of the Fourth Annual Leukocyte Culture Conference. (Ed:
R. McIntyre) page 97. Appleton Century Crafts, N.Y.

El—Arini, M.O. and Osoba, D. 1973. Differentiation of Thymus-
derived cells from precursors in bone marrow. J. Exp. Med.
137:825.

Loor, F. and Roelants, G. 1975. Immunofluorescence studies of
a possible prethymic T-cell differentiation in congenitally athymic
(nude) mice. Ann. N.Y. Acad. Sci. 254:226-242.

Turner, R.W.A., Siminovitch, L., McCulloch, E.A. and Till, J.E.
1967. Density gradient centrifugation of hemopoietic colony-
forming cells. J. Cell Physiol. .92373'82-

Boyse E.A. and Bennett, D. 1974. In Cellular Selection and
Regulation in the Immune Response. (Ed: G.M. Edelman) Raven
Press, N.Y. page 164.

Bach, J.F. and Carnaud, C. 1976. Thymic factors. Progr. Allergy
I21:342-408.

Basch, R. and Kadish, J. 1977. Hematopoietic thymocyte pre-
cursors 11. Properties of precursors. J. Exp. Med. 145:405—419.

Rambourg, A. 1971. Morphologic and histochemical aspects of
g1ycoproteins at the surface of animal cells. Int. Rev. Cytol.
_3l:57—ll4.

Parsons, D.F. and Subjeck, J.R. 1972. The morphology of the
polysaccharide coat of mammalian cells. BBA 265:85-113.

Drzeniek, R. 1967. Differences in splitting capacity of virus
and V. cholerae neuraminidases on sialic acid type substrates.
Biochem. BiOphys. Res. Comm. .2§:631v638.

Ambrose, E.J. 1966. Alteration of cell charge after neuramini-
dase treatement. Progres. Biophys. 163241.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

59

Lee, A. 1968. Effect of neuraminidase on the phagocytosis of
heterologous red cells by mouse peritoneal macrophages. Proc.
Soc. Exp. Biol. Med. 128:891—894.

Weiss, L. Mayhew, E. and Ulrich, K. 1966. The effect of
neuraminidase on the phagocytic process in human monocytes. Lab
Invest. 12:304-309

Baylor, M.E. 1964. Neuraminidase alteration of viral hemagglutin-
ation. Trans. N.Y. Acad. Sci. .26:1103.

Woodruff, J.J. and Gesner, B.M. 1969. The effect of neuramini-

dase on the fate of transfused lymphocytes. J. Exp. Med.
129:551-567.

Yamakawa, T. and Suzuki, S. 1951. The chemistry of the lipids
of post-hemolytic residue or stroma of erythrocytes. 1. Con—
cerning the ether insoluble lipids of lyophilized horse blood
stroma. J. Biochem (Tokyo) '2§:l99-212.

Yamakawa T. and Suzuki, S. 1953. The chemistry of the lipids of
post-hemolytic residue or stroma of erythrocytes. IV. Distri-
bution of lipid-hexosamine and lipid—hemataminic animals. J.
Biochem. (Tokyo) 49:7-10.

Currie, G.H. and Bagshawe, K.D. 1968. The role of sialic acid
in antigenic expression. Brit. J. Cancer 22:843.

Aminoff, D., Bell, W.C., Fulton I. and Ingebrigtsen, N. 1976.
Effect of sialidase on the viability of erythrocytes in circula—
tion. Am. J. Hematol. 12:419—432.

Barth, R.E., and Singla, 0. 1974. Alterations in the immuno-
genicity and antigenicity of mammalian erythrocytes following
treatment with neuraminidase. Proc. Soc. Exp. Biol. Med.
145:168.

Gregoriadis, G., Putnam, D., Louis, L. and Neerunjun, D. 1974.
Comparative effect and fate of non-entrapped and liposome-en-
trapped neuraminidase injected into rats. Biochem. J. 140:323—
330.

Schlesinger, M. and Galili, U. 1974. The effect of neuramini-
dase on the formation of E-rosettes by human peripheral lympho-
cytes and sheep red blood cells. Behring Inst. Mitt. 52:272.

Schlesinger, M. and Amos, B.D. 1971. Effect of neuraminidase on

serologic properties of murine lymphoid cells. Transp. Proc.
3:895.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

60

Barton, N.W. and Rosenberg, A. 1973. Action of Vibrio cholerae
neuraminidase (sialidase) upon the surface of intact cells and
their isolated sialolipid components. J. Biol. Chem. 248:7353-
7358.

 

Reisner, E.G., Flye, M.W., Chung, K.S.S. and Amos, D.B. 1974.
The cytotoxic reactivity and sialic acid content of human lymphoid
cells. Tissue Antigens _4:7-20.

Rosenberg, S. and Rogentine, G.N. 1972. Natural human antibodies
to "hidden membrane components." Nature 239:203.

Rogentine, G.N. and Plocinik, B.A. 1974. Carbohydrate inhibition
studies of the naturally occurring antibody to neuraminidase-
treated human lymphocytes. J. Immunol. 113:848.

Grothaus, E.A., Flyde, M. W., Yunis, E.J. and Amos, D.B. 1971.
Human lymphocyte antigen reactivity modified by neuraminidase.
Science 173:542.

Schlesinger, M. and Gottesfeld, S. 1971. The effect of neuramini-
dase on expression of cellular antigens. Transp. Proc. 2:1151—1154.

Rosenburg, S.A., Plocinik, B.A. and Rogentine, G.N. 1972. Unmask-
ing of human lymphoid cell heteroantigens by neuraminidase treat—
ment. J. Natl. Cancer Inst. 42:1271-1276.

Reisner, E.G. and Amos, D.B. 1972. The complement-binding and
absorptive capacity of human white blood cells treated with
neuraminidase. Transplantation 23:455—461.

Burnet, F.M. and Anderson, S.G. 1947. The "T" antigen of guinea
pig and human red blood cells. Aust. J. Exp. Biol. Med. Sci.
25:213.

Mollison, P.L. 1970. The role of complement in antibody-mediated
red-cell destruction. Brit. J. Hematol. 22:249-255.

Ray, P.K., Gewurz, H. and Simmons, R.L. 1970. The mechanism of
increased sensitivity of neuraminidase-treated cells to antibody
induced cytolysis. Fed. Proc. (abstr.) 22:573.

Ray, P.K. and Simmons, R.L. 1971. Failure of neuraminidase to
unmask allogeneic antigens on cell surfaces. Proc. Soc. Exp. Med.
138:600-604.

Gray, B.N. 1976. Naturally occurring human antibody to neuramini-
dase—treated human lymphocytes. J. Natl. Cancer Inst. 56:211.

Seiler, F.R. and Sedlacek, H.H. 1974. Alterations of Immunological
phenomena by neuraminidase; marked rise in the number of lympho-
cytes forming rosettes or bearing immunoglobulin receptors. Behring
Inst. Mitt. 55:258-271.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

61

Kirby, D.R.S., Billington, W.D., Bradbury, S. and Goldstein, D.J.

1964. Antigen barrier of the mouse placenta. Nature (London)
204:548-549.

Bradbury, S., Billington, W.D. and Kirby, D.R.S. 1965. A histo—
chemical and electron microscopal study of the fibrinoid of the

mouse placenta. J.R. Microsc. Soc. (J. Microscopy, Oxford)
333199-211.

Warren, G.H., Williams, E.C., Alburn, H.E. and Seifter, J. 1949.
Rous chicken sarcoma as a source for hyaluronic acid. Archs.
Biochem. ‘29:300-304.

Girshman, E. 1952. Histochemical analysis of muc0polysaccharides
occurring in mucus—producing tumors. Cancer 2:700—707.

Franks, L.M., Oshea, J.D. and Thomson, A.E.R. 1964. Mucin in
the prostate: A histochemical study in normal glands, latent,
clinical and colloid cancer. Cancer 21:983-991.

Defendi, V. and Gasic, G. 1963. Surface mucopolysaccharides of
polyoma virus transformed cells. J. Cell. Physiol. 62:23—26.

Hughes, R.C., Palmer, P.D. and Sanford, B.H. 1973. Factors in-
volved in the cytotoxicity of normal guinea pig serum for cells
of murine tumor TA3 sublines treated with neuraminidase. J.
Immunol. .222:1071-1080.

Lundgren, G., Jeitz, L., Lundin, L. and Simmons, R.L. 1971.
Increased stimulation by neuraminidase-treated cells in mixed
lymphocyte cultures. Fed. Proc. (abstr.) .2Q:395.

Flye, M.W., Reisner, E.G. and Amos, D.B. 1973. The }n_vitro
effect of neuraminidase on humanlymphocytes. J. Surg. Res.
15:96.

Sanford, B. 1967. An alteration in tumor histocompatability
induced by neuraminidase. Transplantation 5:1273—1279.

Currie, G.A. and Bagshawe, K.D. 1969. Tumor specific immuno-
genicity of methylcholanthrene-induced sarcoma cells after in-
cubation with neuraminidase. Brit. J. Cancer I22:141-149.

Simmons, R.L., Rios, A. and Ray, P.K. 1970. Mechanism of
neuraminidase induced antigen "unmasking." Surg. Forum
‘2l:265-267.

Simmons, R.L. 1971. Immunotherapy of cancer; immunospecific
rejection of tumors in recipients of neuraminidase—treated tumor
cells plus BCG. Science 258:62.

 

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

62

Sanford, B.H. 1974. Neuraminidase and cancer therapy:
possible mechanisms of action. Behring. Inst. Mitt. 22:306-308.

Carubelli, R., Trucco, R.E. and Caputto, R. 1976. Neuraminidase
activity in mammalian organs. Biochim. Biophys. Acta 69:196-197.

Horvat, A. and Touster, O. 1968. On the lysosomal occurrence
and the properties of the neuraminidase of rat liver and of Erlich
ascites tumor cells. J. Biol. Chem. 243:4380-4390.

Yeh, A.K., Tulsiana, D.R.P. and Carubelli, R. 1971. Neuraminidase
activity in human leukocytes. J. Lab. Clin. Med. 12:771-778.

Taha, B.H. and Carubelli, R. 1967. Mammalian neuraminidase:
intracellular distribution and changes of enzyme activity during
lactation. Arch. Biochem. Biophys. 119:55-61.

Tulsiani, D.R.P. and Carubelli, R. 1970. Studies on the soluble
and lysosomal neuraminidasesof rat liver. J. Biol. Chem. 245:
1821-1827.

W. Gielin. 1974. Neuraminidase in higher organisms. Behring Inst.
Mitt. .§§:85-88.

Schengrund, C.L., Jensen, D.S. and Rosenburg, A. 1972. Localiza-
tion of sialidase in the plasma membrane of rat liver cells. J.
Biol. Chem. 247:2742—2746.

Tettamanti, G., Preti, A., Lombardo, A., Suman, T. and Zambotti, V.
1975. Membrane-bound neuraminidase in the brain of different
animals: Behavior of the enzymes on endogenous sialo derivatives
and rationale for its assay. J. Neurochem. 22:451-456.

Yohe, H.C. and Rosenburg, A. 1977. Action of intrinsic sialidase
of rat brain synaptic membranes on membrane sialolipid and sialo-
protein components ln_situ. J. Biol. Chem. 252:2412-2418.

Finkelstein, R.A. and LoSpalluto, J.J. 1969. Pathogenesis of
experimental cholera. Preparation and isolation of choleragen and
choleragenoid. J. Exp. Med. 130:185-202.

LoSpalluto, J.J. and Finkelstein, R.A. 1972. Chemical and
physical properties of cholera exo-enterotoxin (choleragen) and
its spontaneously formed toxoid (choleragenoid) Biochim. Biophys.
Acta .252:158-166.

Finkelstein, R.A., Noms, H.T. and Dutta., N.K. 1964. Pathogene-
sis of experimental cholera in infant rabbits. J. Inf. Dis.
114:203.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

63

vanHeyningen, W.E., Carpenter, C.C.J., Pierce, N.F. and Greenough,
W.B. III. 1971. Deactivation of cholera toxin by ganglioside.
J. Inf. Dis. 124:415:418.

Baker, A.L., Kaplan, M.M., Kimberg, D.V. and Pierce, N.F. 1971.
Stimulation of rat liver enzyme activity by intravenous cholera
toxin. Gastroenterology .60:739.

Franks, D.J. 1976. Cholera toxin and adenylate cyclase: proper-
ties of activated enzyme in liver plasma membranes. Can. J. Bio-
Chem. 24:981-987.

Donta, S.T., King, M. and SlOper, K. 1973. Induction of
steroidogenesis in tissue culture by cholera enterotoxin. Nature
(New Biol)243:246-247.

Hollenberg, M.D. and Cuatrecasas, P. 1973. Epidermal growth
factor: receptors in human fibroblasts and modulation of action
by cholera toxin. Proc. Natl. Acad. Sci. (USA) 29:2964-2968.

Cuatrecasas, P. 1973. Cholera toxin-fat cell interaction and
the mechanism of activation of the lipolytic response. Biochem.
22:3567-77.

Vaughn, M., Pierce, N. and Greenough, W.B. 1970. Stimulation of

glycerol production in fat cells by cholera toxin. Nature
226:658-659.

Moritz, M., Iber, F.L. and Moore, E.W. 1974. Rabbit cholera:
effects of cycloheximide on net water and ion fluxes and transmural
electrical potentials. Gastroenterology _§2:76-82.

Cuatrecasas, P. 1973. Interaction of Vibrio cholerae entero-
toxin with cell membranes. Biochem. I22:3547-3558.

 

Pierce, N.F. 1973. Differential inhibitory effects of cholera
toxoids and ganglioside on the enterotoxins of Vibrio cholerae
and Escherichia coli. J. Exp. Med. 137:1009-1023.

 

 

Holmgren, J., Lonnroth, I. and Svennerholm, L. 1973. Fixation
and inactivation of cholera toxin by GM1 ganglioside. Scand. J.
Inf. Dis. _§:77-78.

Hollenberg, M.D., Fishman, P.H., Bennett, V. and Cuatrecasas, P.
1974. Cholera toxin and cell growth: role of membrane ganglio-
sides. Proc. Natl. Acad. Sci. 12:4224-4228.

Cuatrecasas, P. 1973. Vibrio cholera choleragenoid: Mechanism
of inhibition of cholera toxin action. Biochem. 22:3577-3581.

Lonnroth, I. and Holmgren, J.J. 1973. Subunit structure of
cholera toxin. J. Gen. Microb. .26:4l7-427.

128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

64

Bennett, V., O'Keefe, E. and Cuatrecasas, P. 1975. Mechanism of

action of cholera toxin and the mobile receptor theory of hormone-
receptor adenylate cyclase activity. Proc. Natl. Acad. Sci. (USA)
72:33-37.

Bennett, V. and Cuatrecasas, P. 1975. Mechanism of activation
of adenylate cyclase by y, Cholerae enterotoxin. J. Membr. Biol.
22:29-52.

Craig, S.W. and Cuatrecasas, P. 1975. Mobility of cholera toxin
receptors on rat lymphocyte membranes. Proc. Natl. Acad. Sci.
(USA) IZ2:3844-3848.

Cuatrecasas, P. 1973. Isolation and characterization of hormone
receptors. Neurosci. Res. Program. Bull. _22:215-219.

Guerrant, R.L., Chen, L.C. and Sharp, G.W.G. 1972. Intestinal
adenyl-cyclase activity in canine cholera: correlation with
fluid accumulation. J. Inf. Dis. 125:377-381.

Singer, S.J. and Nicolson, G.L. 1972. The fluid mosaic model of
the structure of cell membranes. Science 175:720-731.

vanHeyningen, S. 1974. Cholera toxin: interaction of subunits
with ganglioside GMl' Science 183:656-657.

Gill, D.M. and King, G.A. 1975. The mechanism of action of
cholera toxin in pigeon erythrocytes lysates. J. Biol. Chem.
250:6424-6432.

Kandel, J., Collier, R.J. and Chung, D.W. 1974. Interaction of
fragment A from diptheria toxin with nicotinamide adenine dinucleo-
tide. J. Biol. Chem. 249:2088-2097.

Moss, J. 1977. Mechanism of action of choleragen: Evidence for
ADP-ribosyltransferase activity with arginie as an acceptor.
J. Biol. Chem. Z:2455(comm.).

Flores, J., Witkum, P. and Sharp, G.W.G. 1976. Activation of
adenylate cyclase by cholera toxin in rat liver homogenates. J.
Clin. Invest. 21:450-458.

Sultzer, B.M. and Craig, J.P. 1973. Cholera toxin inhibits
macromolecular synthesis in mouse spleen cells. Nature (New
Biol.) 244:179-180.

Honjo, T., Nishizuka, Y., Kato, I. and Hayaishi, O. 1971.
Adenosine diphOSphate ribosylation of aminoacyl transferase II
and inhibition of protein synthesis by diptheria toxin. J. Biol.
Chem. 246:4251-4260.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

151.

152.

65

Moss, J., Richards, R.L., Alving, C. and Fishman, P.R. 1977.
Effect of the A and B protomers of choleragen on release of trapped

glucose from liposomes containing or lacking ganglioside GMl' J.
Biol. Chem. 252:797-798 (comm.).

Stein-Douglas, K.E., Schwarting, G.A., Naiki, M. and Marcus, D.M.
1976. Ganglioside as a marker for murine lymphocyte subpopulations.
J. Exp. Med. 143:822-832.

Winchester, R.J., ShuManFu, Winfield, J.B. and Kunkel, H.G. 1975.
Immunofluroescent studies on antibodies directed to a buried mem-

brane structure present in lymphocytes and erythrocytes. J.
Immunol. 114:410-414.

Ray, P.K. and Simmons, R.L. 1973. Serological studies of enzyme-
treated murine lymphoid cells. Proc. Soc. Exp. Biol. Med. 142:
846-852.

Rosenberg, S.A., Schwartz, S. and Baker, A. 1974. Natural anti-
bodies to "cryptic" membrane antigens exposed by treatment with
neuraminidase. Behring Inst. Mitt. 26:204-208.

Hakormori, S. 1969. Differential reactivities of fetal and adult
human erythrocytes to antisera directed against glycolipids of
human erythrocytes. Vox. Sang. 26:478-485.

Gahmberg, C.T. and Hakomori, S. 1973. External labeling of cell
surface galactose and galactosamine in glycolipid and glyco-
protein of human erythrocytes. J. Biol. Chem. 248:4311-4317.

Homlgren,J., Lonnroth, I. and Svennerholm, L. 1973. Tissue
receptor for cholera exotoxin: postulated structure from studies
with GMl ganglioside and related glycolipids. Infect. Immun.
_§:208- 214.

King, C.A. and vanHeyningen, W.E. 1973. Deactivation of cholera
toxin by a sialidase-resistant monosialosyl-ganglioside. J. Inf.
Dis. 127:639-647.

Finkelstein, R.A., Boesman, M., SimHeeNeoh, LaRue, M.K. and
Delaney, R. 1974. Dissociation and recombination of the subunits
of the cholera enterotoxin (choleragen). J. Immunol. 113:145-150.

Miller, H.C. and Esselman, W.J. 1975. Identification of theta-
bearing T-cells derived from bone marrow cells treated with thymic
factor. Ann. N.Y. Acad. Sci. 245:54-60.

Esselman, W.J. and Miller, H.C. 1974. The ganglioside nature of
theta antigens. Fed. Proc. (abstr.) .22:771.

153.

154.

155.

156.

157.

158.

159.

160.

161.

162.

66

Dicke, K.A., Tridente, G. and vanBekkum, D.W. 1969. The
selective elimination of immunologically competent cells from bone
marrow and lymphocyte cell mixtures. Transplantation 6:422-434.

Naiki, M., Marcus, D. and Ledeen, R. 1974. PrOperties of anti-
sera to ganglioside GM1 and asialo GMl' J. Immunol. 113:84-93.
Finkelstein, R.A. and LoSpalluto, J.J. 1970. Production of highly
purified choleragen and choleragenoid. J. Inf. Dis. 121 (Supple-
ment):63-72.

Komuro, K. and Boyse, E.A. 1973. Induction of T lymphocytes from
precursor cells 22 vitro by a product of the thymus. J. Exp. Med.
138:479-482.

Haksar, A., Maudsley, D.V. and Peron, F.G. 1974. Neuraminidase
treatmentof adrenal cells increases their response to cholera entero-
toxin. Nature 251:514-515.

Sedlacek, H.H. and Seiler, F.R. 1974. Demonstration of Vibrio-
cholerae neuraminidase (VCN) on the surface of VCN—treated cells
Behring Inst. Mitt. 66:254-257.

Ray, P.K. and Simmons, R.L. 1972. Comparative effect of viral
and bacterial neuraminidase on the complement sensitivity of lymphoid
cells. Clin. Exp. Immunol. _;g:139-150.

Roelants, G.E., Loor, F., vonBoehmer, H., Sprent, J., Hagg, L.,
Mayor, K.S. and Ryden, A. 1975. Five types of lymphocytes
characterized by double immunofluroescence and electrophoretic
mobility. Organ distribution in normal and nude mice. Eur. J.
Immunol. 6:127-131.

Holmgren, J. 1973. Comparison of the tissue receptors for Vibrio
cholerae and Escherichia coli enterotoxins by means of ganglio-
sides and natural cholera toxoid. Infect. Immun. 6:851-859.

 

Trowbridge, I.S., Weissman, I.L. and Bevan, M.J. 1975. Mouse
T-cell surface glycoprotein recognized by heterologous anti-
thymocyte sera and its relationship to Thy-l antigen. Nature

256:652-654.