.. .
0‘1”",‘1h‘hcl'fl'
-- Ni.

‘5 or. 3‘1;

_ v1.2" Mr“.
" :4

.\

«2-
_J

x1221:
71,2"
.F.-:‘2“

.._.C c

a ‘ .f.‘ .‘yr

PJ"

‘J W?“
33:22 1%} £313.22
.: 'F “'3‘

A.~V.-..
.0“

,.- - - :454‘ t‘

.3,

..
2.4'.‘-‘
.J J. ..J.

J.‘ ~vvui.

“finwzw't
' J“x:$:“‘ :‘f;v;;' "vvbitfvi

.,i 3;
23.2322

r"?

3‘. Vv" y“
u‘. “.33.
. - ' _ ._J
n u'l’..'-'."i
vv

 

u u". . .2
w "u; L. ‘2“
' s

H ’02: .11“.
2 .1 '4'“-

'2‘
.' n;

 

LIBRARY

MichiganS
II' 'ty ‘

 

This is to certify that the

thesis entitled
STUDIES ON THE g m INDUCTION
OF THE THY-1 DIFFERENTIATION
ANTIGEN ON MURINE PROTHYMOCYTES
presented by

Christine Ann Clark

has been accepted towards fulfillment
of the requirements for

 

 

 

M.S. degree inMicrobiology & Public
Health
7?
Midi/I /fi4€%¢fim\
Major plofessor
Date 7// / ff’, / ’./{

 

0-7639

 

'“‘

 

STUDIES ON THE IN_VITRO INDUCTION
OF THE THY-l DIFFERENTIATION

ANTIGEN ON MURINE PROTHYMOCYTES

BY

Christine Ann Clark

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

1978

67057213

ABSTRACT

STUDIES ON THE E y_I_T§9_ INDUCTION
OF THE THY-1 DIFFERENTIATION
ANTIGEN ON MURINE PROTHYMOCYTES
by

Christine Ann Clark

The induction of Thy-l, a T cell differentiation antigen, on murine
prothymocytes has been studied with an in vitro conversion assay. Sev-
eral parameters of the assay have been examined closely in an effort to
improve its reproducibility. These include methods used to isolate pro-
thymocytes and the sensitivity of the cytotoxicity test used to detect
the expressed antigen.

Glass and nylon wool columns as well as density centrifugation were
used to separate prothymocytes from bone marrow and splenic populations.
Nonadherent and low density fractions treated with thymic factor expressed
Thy-1. Two to three times more Thy-1 positive cells could be detected
with an improved counting method for the cytotoxicity test.

In addition, a new agent capable of inducing Thy-l expression has
been discovered. Sodium butyrate in low concentrations induces antigen
expression on both bone marrow and splenic precursor T cells in a

fashion similar to that of thymic factor induction.

ACKNOWLEDGEMENTS
I would like to thank my advisor Walter Esselman for his help,

understanding, and support. In addition, I would like to thank my

friends for their friendship and Bill just for being Bill.

ii

TABLE OF CONTENTS
Page
LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . 1
Thy—1 Antigen . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Definition and distribution - - - - - ' ° - - - - - - - ~ - - - 1
Chemical characterization of Thy-l . - - - - - - - - - - - ~ - - 2
Protein nature of Thy-1 - - - - - - - - - - - - - . - - - - - - 2
Glycoprotein nature of Thy-l ° - - - - - - ° - - - - - - - - - - 3
Glycolipid nature of Thy-l - - - - - - . - . . - - - . . . - ~ - 5

ThymicFaCtor..........................'7

Introduction 0 . . o o . . . o . . . . o . . . . . o . . o . o o 7
Assays for thymic factor ° ' - - - - - - - . . - - - - - - - - - 7
Thy—1 induction assay . . . . . . . . . . . . . . . . . . . . . 8
Hormonal action of thymic factor - . - . . . . . . . . . . . . . 9
Intracellular effects of thymic factor . . . . . . . . . . . . . 10
Specificity of thymic factor - . - - . . . . . . . . . . . . . . 11
Purified thymic hormones . . . . . . . . . . . . . . . . . . . . 13

Sodium Butyrate . . . . . . . . . . . . . . . . . . . . . . . . . l4

Butyrate treatment of cells . . . . . . . . . . . . . . . . . . l4
Butyrate and cAMP. . . . . . . . . . . . . . . . . . . . . . . . 15
Butyrate and glycolipid metabolism . . . . . . . . . . . . . . . 16
Molecular level of butyrate action . . . . . . . . . . . . . . . 18
Specificity of butyrate. . . . . . . . . . . . . . . . . . . . . 20

Target Cell for Inducing Agents 20

Stem cells . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Identification of CFU'S . . . . . . . . . . . . . . . . . . 21
Relationship between CFU'S and T lymphocytes - . . - - . . . . . 22
Thymic precurso cells - - - - - - . . . . . . . . . . . . . . . 24

Stem Cell Separation Techniques. 26

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Glass and nylon wool column separation . . . . . . . . . . . . . 27
Isopycnic density gradient separation. - . . . . . . . . . . - . 28
Velocity sedimentation . . . - . . . . . . . . . . . . . . . . . 29
Two dimensional cell separation. - - . . . . . . . . . . . . . . 30
Additional separation techniques - - - - . - . - . - - - - - ~ - 31

iii

INTRODUCTION TO EXPERIMENTAL REPORT . . .
MATERIALS AND METHODS . . . . . . . . . .

Cells. . . . . . . . . . . . . . . . .
Cell separation techniques . . . . . . .
In vitro Thy-1 conversion assay . . . .
Detection of expressed Thy-1 antigen .
Antisera and complement . . . . . . .
Cytotoxicity test . . . . . . . . . .
Absorption of anti-Thy-l sera . . . . .

RESULTS . . . . . . . . . . . . . . . . .

Sensitivity of the Cytotoxicity Test . .
Increased detection of Thy-l inducible

bone marrow

Detection of Thy-1 inducible splenocytes . . . . .
Sodium Butyrate Induction of Thy-l Expression . . .
Absorption of Anti-Thy-l Sera by TF or Butyrate Treated Marrow
.Prothymocyte Separation - The Use of Glass and Nylon Wool Columns .

Time Course of Increased Cell Lysis . .
DISCUSSION . . . . . . . . . . . . . . . .

BIBLIOGRAPHY . . . . . . . . . . . . .

iv

Cells

Page
32
34

34
34
36
36
36
36
37

39

39
39
41
43
43
48
51

54

65

LIST OF TABLES

Table Page
I Induction of Thy-1.1 and Thy-1.2 on murine bone marrow cells

by thymic factor 40

II Induction of Thy-1.2 on C3H splenocyted by thymic factor 42

III Induction of Thy-1.1 and Thy-1.2 on murine bone marrow cells
by sodium butyrate 44

IV Induction of Thy-l positive cells in murine bone marrow and
spleen cells fractionated by BSA density gradient centrifugation 45

V Induction of Thy-1.2 positive cells by thymic factor in C3H bone
marrow cells fractionated by glass and nylon wool columns 49

VI Detection of Thy-1.2 positive cells in C3H splenocytes fraction-
ated by glass and nylon wool columns and treated with thymic
factor. 50

LIST OF FIGURES

Figure Page
1 Absorption of anti-Thy-l.2 serum with thymic factor
and butyrate treated bone marrow cells 47
2 Time course of increased cell lysis during the induction
of Thy-l on glass and nylon wool separated bone marrow
cells 53

vi

LITERATURE REVIEW

Thy-l Antigen

 

Definition and distribution. In 1964, Reif and Allen (1) first

 

demonstrated the presence of a shared allogeneic antigenic determinant
in the thymocytes, peripheral T cells, and central nervous tissue of
mice. It is now known that this alloantigen, Thy-l (formerly known as
theta antigen), is expressed as one of two allelic forms, Thy-1.1
(en-AKR) or Thy-1.2 (OH-C3H). These are coded for by the Thy:l_locus
on chromosome 9 of the mouse (2) and all mouse strains carry one or
the other of these two antigenic specificities. It was originally
thought that this antigenic system was species specific. However, the
presence of a serologically indistinguishable antigen has also been
described in rat brain and thymocytes (3). In contrast to the murine
system, only one form of the antigen, Thy-1.1, has been demonstrated
in rat tissue (3).

In the mouse, the membrane bound Thy-l antigen has been used as
a T cell marker and as an example of a differentiation antigen due to
its distribution and the level of its appearance in murine tissue. It
is found in high quantities only on immature T cells (thymocytes) and
on brain tissue (1,4). This level appears to decline with T cell matur-
ation, as Thy-l is demonstrable only in small amounts on peripheral
thymus-derived cells in mice (1). In addition, Thy-l has been demon-
strated on murine fibroblastic cell lines (5), epidermal cells (6),
cells derived from mammary tissue (7), and cells in the peripheral and
central nervous system (4). It has not been found on murine bone mar-
row or marrow derived lymphocytes (1), plasma cells (8) or granulocytes,

macrophages, or red blood cells (9). In the rat, however;while

thymocyte and brain expression of Thy-l are comparable to that in mice
(10), it has been reported that as high as 30-40% of bone marrow cells
express Thy-1.1 (11), and that rat peripheral T lymphocytes express
little or no Thy-1 (10). Hunt et a1. (12) reported that approximately
25% of the Thy-l positive bone marrow cells were also positive for sur-
face immunoglobulin, a B cell marker. They conclude that, in the rat,
Thy-l is not exclusively associated with thymic-dependent lineage.

Chemical characterization of Thy-1. The membrane bound nature of
the Thy-l molecule has posed some problems in its isolation and subse-
quent purification. Consequently, the exact chemical characterization
of the antigen is still a matter of controversy. Thy-1 antigenicity
has been proposed to be carried by a protein moiety (13), by a glycopro-
tein (14,15), or by glycolipids (16,17). This controversy may be due
in part to the variety of methods used to isolate the Thy-1 molecule.
It may also be related to the various assay procedures used to monitor
Thy-l activity during the isolation procedure. Detergent solubiliza-
tion (18), immunoproecipitation (13), enzyme solubilization (19), and
organic solvent extraction (20) have been the most widely employed
isolation techniques.

Protein nature of Thy-1. Atwell et a1. (13) described the isolation

 

and partial characterization of a cell surface protein which carried
Thy-1 determinants. Thymocytes were cell surface radioiodinated by
lactoperoxidase; their proteins were extracted into organic solvents;
and the antigen was identified by specific co-precipitation of radio-
active cell surface proteins with anti-Thy-l sera and rabbit antiserum
to mouse immunoglobulin. Analysis of the immunoprecipitate by SDS-

acrylamide disc electrophoresis led to the conclusion that Thy-l was

a protein monomer of approximately 60,000 daltons. Other investigators
(l9), characterizing Thy-l as expressed on a murine lymphoblastoid cell
line, also reported that the Thy-l antigen was associated with protein.
Solubilization of the alloantigen was accomplished by limited papain
digestion of the cells. Increased exposure of the papain digest to
papain or to protease reduced the amount of antigenic activity as mea-
sured by a 51Cr release cytotoxicity assay, suggesting that the Thy-l
determinant was protein associated. The size of the material possess—
ing Thy-1 activity was estimated at 200,000 daltons by gel filtration

of the concentrated papain digest on Sephadex G-200. 0n the other hand,
Letarte-Muirhead et al. (18) demonstrated that Thy-l activity isolated
from rat tissue by detergent solubilization was associated with a glob—
ular protein of much smaller size, approximately 28,000 daltons. The
molecular weight determination was based on gel filtration and sucrose
gradient centrifugation of thymocyte membranes extracted with the deter-
gents deoxycholate of Lubrol-PX. Thy-l activity in the detergent ex-
tracts was monitored by inhibition of a radioactive binding assay speci-
fic for the alloantigen. Therefore, while all of these studies sug-
gested that Thy-l may be protein in nature, they did not agree on fur-
ther physical parameters. Nor did any of them eliminate the possibility
that the Thy-l activity was associated with a carbohydrate portion of a
glycoprotein or with glycolipids.

Glycoprotein nature of Thy-l. Trowbridge et al. (15) isolated a
glycoprotein of about 25,000 daltons from mouse thymocyte membranes by
immunoprecipitation. This molecule (T-25) was found on the surface of
cell lines that carried Thy-1 antigen, but it was absent from derivative

lines that lacked the antigen (21). Further studies by Letarte-Muirhead

et al. also described the Thy-1.1 alloantigen from rat thymocytes (14)
and rat brain (22) as a glycoprotein weighing approximately 25,000 dal-
tons. Their new purification procedure included not only detergent sol-
ubilization and gel filtration but also affinity chromatography on anti-
body or lentil lectin columns. Finally, Johnson et al. (23) reported
that the Thy-1.2 alloantigen expressed on the S49 lymphoblastoid cell
line was a glycoprotein.

Debate still exists, however, over which moiety of the Thy-l glyco-
protein is responsible for the antigenic activity of the molecule. The
amino acid and carbohydrate composition of the purified rat Thy—l anti-
gen have been determined (24). The brain and thymus glycoproteins con-
tained very similar amino acid compositions, but strikingly different
carbohydrate moieties which accounted for approximately 30% of the mol-
ecular weight of the antigen. The authors, therefore, suggested a pro-
tein basis for the determinants since they could not detect any anti-
genic differences between brain and thymus Thy-l (22). Furthermore,
heating and proteolysis by pronase resulted in the loss of Thy-l anti-
genic activity (24). However, it was also found that the antigenicity
was not affected by other proteolytic enzymes, including trypsin and
papain. No studies were done to alter the carbohydrate portion of the
molecule. Others (21,23) have proposed that the antigenic determinants
of the mouse Thy-l molecule are on the carbohydrate portion of the gly-
c0protein. Johnson et al. (23) based their proposal on the fact that
neuraminidase, which cleaves sialic acid from suitable substrates, was
shown to cause the Thy-l alloantigen from lymphoblastoid cell lines to
lose its antigenic activity. It appears likely that protein may be only

partially responsible for the molecule's antigenicity perhaps by serving

as a hapten carrier for the carbohydrate moiety.

Glycolipid nature of Thy-l.

 

In their original description of

Thy-l antigen, Reif and Allen (4) demonstrated that the antigen was

nondializable and sensitive to lipid solvents.

Vittetta et a1.

(17)

isolated a complex containing the Thy-l antigen from the surface of

murine thymocytes and T cells by cell surface radioiodination, lysis

by freeze-thawing, and immunoprecipitation. This

ily labeled with 3H-galactose but not with labeled amino acids.

also shown to have a lower density than protein.

entity could be read-
It was

These observations

prompted the authors to suggest that the antigenicity of the Thy-1 may

reside in a glycolipid.

In accord with these findings, Esselman and Miller (20) proposed

that mouse brain associated Thy-1 antigen was glycolipid in nature in

view of their findings that the ganglioside rich upper layer of a Folch

partition of brain or thymus tissue was capable of inhibiting the cyto-

toxicity of anti-brain associated Thy-1 antiserum

auxiliary lipids. The isolated gangliosides G

le
ilar capacities and were, therefore, purported to
city. Furthermore, pentasaccharides derived from
were able to specifically inhibit the appropriate
hapten inhibition assay (25). Conversely, others
immunofluoresence that, while purified antibodies
peripheral T cells and most thymocytes in several

reaction was independent of Thy-l phenotype.

gests that the Thy-1 glycolipid was only a small part of the G

GD1

when associated with

and G

M1 possessed Sim-

carry Thy-l antigeni-
these gangliosides
anti-Thy-l sera in a
(26) have shown by

to G did react with

M1

strains of mice, this

More recent evidence sug-

M1 and

b preparations used in earlier studies (27). These gangliosides

had appeared to be pure in two distinct thin layer chromatography

6

systems. However, further fractionation by a third thin layer system
resulted in the separation of a very minor component which contained
all of the Thy-l activity. An immune response assay was used to mon-
itor the improved purification of Thy-l. Isolated Thy-l glycoprotein
reacted in parallel to the Thy-l glycolipid in the assay. It was pro-
posed, therefore, that the carbohydrate moiety of Thy-l determined the
antigenicity of the molecule and was conjugated either to a lipid or a
protein carrier.

In contrast, Arndt et a1. (28) found that extraction of murine
thymocyte membranes with organic solvents yielded a considerable loss
of activity of the Thy-l antigen with some residual activity in the
protein fraction but none in the organic phase. The reduced antigenic
activity of the delipidated protein fraction could be restored by the
addition of lipids. The investigators proposed that the thymocyte-
brain antigen was protein in nature, but that lipid-protein interaction
was necessary for the antigenicity of the Thy-l molecule.

Further evidence that Thy-1 may be glycolipid in nature is the
recent report by Thiele et a1. (29) that choleragen and Thy-l show a
common ligand-induced redistribution when subjected to cocapping exper-

iments. Since the ganglioside G is the main receptor for cholera tox-

M1
in (30), it appears likely that the Thy-l antigen is either glycolipid
in nature or closely associated with glycolipids in the membrane.

To date the exact chemical nature of the Thy-l alloantigen is un-
known. Evidence for both glycoprotein and glycolipid moieties has been
shown. Its functional nature also is unknown. Thy-l is shed from thymo-

cytes (31) and lymphoblastoid cells in culture (32) suggesting that it

is a peripheral rather than an integral member of the cell surface.

Proposals have been made that this shed material may play a nonspecific
role in the modulation of the immune response (16). Purification of the
antigen will aid in the discovery of its true biological role.

Thymic Factor

 

Introduction. The essential role of the thymus in the immune sys-

 

tem was first demonstrated through the ability of thymic grafts to re-
store immunological competence to neonatally thymectomized (NTx) mice
(33). Partial recovery of cell-mediated and humoral immune reactions
was also observed when thymic grafts were enclosed in cell impermeable
chambers (34,35) implying that a thymic humoral factor participated in
the phenomenon. Subsequently, Trainin (36,37,38) reported that admin-
istration of thymus tissue extracts from different species to NTx mice
partially restored their ability to produce immune responses to sheep
red blood cells and to reject skin and tumor allografts. Preliminary
experiments suggested that lymphoid cells, probably of thymus origin,
were the site of action of these extracts (39). It was also suspected
that these factors were produced by thymus epithelium (40).

Assays for thymic factor. A wide variety of methods have been

 

used to isolate and assay a number of so-called thymic hormones. Not

all of these methods were necessarily based on the restoration of immuno-
competence. For example, G. Goldstein's thymopoietin is a purified poly-
peptide hormone isolated from bovine thymus and assayed by its effect on
neuromuscular transmission in a laboratory model of the disease myas-
thenia gravis (41). Only later was it shown to be active in the induc-
tion of T-cell differentiation (42). A. Goldstein and White first test-
ed the activity of their calf thymosin preparation with a "lymphopoietic"

assay which measured DNA synthesis by lymph nodes after in viva injection

of the extract (43). It was also discovered that thymosin was active
in a bioassay based on the induction of azathrioprine sensitivity in
rosette-forming spleen cells from adult thymectomized mice (44). Sub-
sequently, J.F. Bach et al. discovered the presence of a substance in
normal mouse serum which possessed thymosin—like activity in the rosette
assay (45,46). This "circulating" thymic hormone was absent in the se-
rum of thymectomized or nude mice suggesting that is was thymus specific.
It was postulated that the capacity of the lymphoid cells to form ro-
settes correlated with restoration of immunocompetence and the ability
of the cells to react with an antigen (47). Later, Komuro and Boyse
found that thymosin was able to induce the development of cells bearing
characteristic T lymphocyte surface antigens from populations of murine
spleen or bone marrow cells (48,49).

Thy-l induction assay. The study of T lymphocyte differentiation
in the mouse has been aided by the presence of several distinctive sur-
face antigens found on T lymphocytes in various differentiated states.

These can be summarized as follows:

Antigens Location
Thy-l (6) Thymocytes, peripheral T cells and

brain (1), epidermal cells (6),
fibroblasts (5).

TL Thymocytes only (50).
Ly l, Ly 2,3 Thymocytes, peripheral T cells (51).
G1x Thymocytes and sperm of certain mouse

strains (52).

MSLA Thymocytes, peripheral T cells,
brain, and epidermal cells (53).

Following a two hour incubation with thymosin, Komuro and Boyse (49)

demonstrated the appearance of TL and Thy-1 antigens on approximately

20-30% of those bone marrow and spleen cells from normal mice which
rested on the 10-23% interface of a discontinuous BSA gradient. They
were also able to show the expression of TL and Thy-l on embryonic liver
cells and on bone marrow and spleen cells from 14 day old nu/nu mice (48).
These findings demonstrated that the inducible cell had not undergone any
previous thymus-mediated process. However, the actual mechanism of action
of the relatively crude thymic extract which induced expression of the
antigens was still a matter for speculation. The findings of the in viva
Thy-l conversion assay implied that the inducible cell from bone marrow
and spleen was committed to T cell differentiation even though it dis-
played none of the recognizable phenotypic traits of T lymphocytes.

Hormonal action of thymic factor. Several groups of investigators

 

suggested that thymic factors may act in a hormonal fashion on the mem-
brane of the precursor cell, perhaps through the intermediary second
messenger cAMP. In support of this hypothesis Scheid et al. (54) dem-
onstrated that agents of nonthymic origin, which had previously been
shown to increase intracellular levels of cAMP in other systems, were
also capable of inducing T cell surface antigens. These included Poly
A:U, endotoxin, cAMP and DB-cAMP. In addition, insulin, a purported
inhibitor of adenyl cyclase, decreased the expression of TL or Thy-l
antigens and aminophylline, a phosphodiesterase inhibitor, enhanced
induction. It appeared, therefore, that any agent which increased cAMP
levels was capable of inducing a predetermined Thy-1 negative precursor
cell to become a Thy-l positive cell.

Kook and Trainin (55) also demonstrated that agents which affect
intracellular cAMP levels mimicked the activity of their thymus humoral

factor (THF) as measured in an in vitro (graft vs. host) GVH response.

10

They also documented an increase in adenyl cyclase activity in spleen
cells from neonatally thymectomized mice when they were incubated with
THF for five minutes (56). Finally, Bach et al. (57,58) reported that
low levels of cAMP could be substituted for their circulating TF in vitra
to produce rosette forming cells. The effect of cAMP and thymic factor
were found to be synergistic. They also found prostaglandins E1 and B2

as well as theophylline enhanced levels of rosette forming cells, while

and A and AMP had no effect (59).

prostaglandins A1 2

Intracellular effects of thymic factor. The mode of action of thy-

 

mic hormones, outside of a suspected involvement with the adenylate cy-
clase-cAMP system, remains undetermined. Initial experiments have shown
that metabolic activity was required for TF action since no antigen ex—
pression can be demonstrated at temperatures less than or equal to 4°C
(32,42,49). Komuro and Boyse (49) demonstrated that cycloheximide could
block the induction of Thy-l expression on spleen cells suggesting that
protein synthesis was required for hormone action. In experiments de-
signed to test the induction of immunocompetence in lymphoid cells by
THF, Kook and Trainin (56) also demonstrated the need for protein syn-
thesis after studying the effects of cycloheximide. Two hypotheses were
put forward by J.F. Bach to explain such evidence (59). First, thymic
extracts induce the neosynthesis of Thy-l antigen by activation of a
gene that was not expressed previously. Second, these hormones elicit

a membrane rearrangement which allows a sufficient amount of previously
present antigen to become available for detection. Either gene activation
or membrane rearrangement could be mediated by cAMP. The membrane rear-
rangement hypothesis was favored by Bach since Thy-l induction in viva

by both TF (60) and cAMP (61) in adult thymectomized mice had been

11

shown to be reversible. Boyse and Abbot (62) postulated that the ini-
tial consequence of induction was the reconstitution of the cell sur-
face by an earlier-expressed set of genes. Finally, Milewicz et al.
(63) suggested that the modification of existing compounds on the
cell's surface or the expression of a cryptic antigen accounted for
the rapid rate of Thy-l appearance (15 min) by TF rather than de nava
synthesis.

In further experiments, Storrie et al. (64) have demonstrated
that the induction of the surface phenotype requires both transcription
and translation but not DNA replication. Metabolic inhibitors of RNA
metabolism, actinomycin D, campothecin, and cordycepin, suppressed the
expression of TL on precursor cells incubated with thymopoietin. These
results indicated a need for newly synthesized RNA. It was suggested
that this RNA could either specify the actual molecules that appear as
cell surface components or that it could constitute or specify a regula-
tory molecule responsible for the display of preexisting components.
Inhibitors of DNA replication, cytosine arabinoside and hydroxyurea,
did not block induction of antigen expression. This finding was in
accord with the rapid onset of the induction process (§_two hours)
which does not appear to allow time for DNA replication. Definitive
work on intracellular effects of TF will be possible when purer popu-
lations of precursor cells are obtainable.

Specificity of thymic factors. Extracts from different tissues

 

prepared in parallel to thymic extracts were generally employed as con-
trols during testing of thymic hormone activity. Occasionally, these
supposedly inactive tissue extracts, for example, from calf muscle or

spleen (54), would also induce antigen expression or mimic another

12

suspected property of thymic factors. During isolation of thymopoietin,
G. Goldstein et al. (65) discovered another polypeptide of low molecular
weight which was capable of inducing Thy-l expression. This polypeptide
was found in nearly every tissue tested. Consequently, it was named
UBIP—-ubiquitous immunopoietic polypeptide. Its presence in a wide
variety of sources was thought to explain the occasional induction by
tissues of non-thymic origin. Unlike low levels of thymopoietin, how-
ever, UBIP was not shown to be specific for T cell differentiation.

UBIP also induced the appearance of complement receptors (CR) on precur-
sor cells (66), which are purported B cell markers. Thymopoietin did
not induce CR.

The specificity of thymopoietin was further shown by the effect of
propanolol, a beta-adrenergic blocking agent, on TP or UBIP induced anti-
gen expression (66). Precursor cells incubated with TP either in the
absence or presence of propanolol still expressed TL but no CR. Con-
versely, cells incubated with UBIP in the presence of the agonist ac-
quired neither TL or CR. These results suggested that TP and UBIP acted
via two different receptors to induce antigen expression and that TP is
the more likely physiological inducer. Since agents which elevate cAMP
levels not only induce expression of T cell antigens but also purported
B cell markers in a like fashion to UBIP, Scheid et al. (66) postulated
further that these nonspecific inducers acted via the same adrenergic
receptor as UBIP, not the supposed thymopoietin receptor.

In contrast to these findings, Hammerling et al. (67) have shown
that TP at higher concentrations also affects mouse B cell differentia-
tion by the induction of alloantigen Ia and surface Ig. Furthermore,

striking effects were also demonstrated on granulocyte membranes (68).

13
Recently, Kagan et al. reported that complement receptors can be used
as differentiation markers for granulocytes and that induction of this
marker on precursor cells was demonstrated by both TP and ubiquitin (69).
These authors have postulated that the specificity of TP resides not only
in its selective inductive capacity on cells which possess its receptor,
but also in the restricted distribution of the molecule in viva. There-
fore, TP may be a specific inducer of T cell differentiation in viva only
if it reaches an active concentration in the thymus.

Purified thymic hormones. Recently, much work has been done to

 

purify to homogeneity the various factors described in thymic extracts
and serum by different investigators. The amino acid sequence of three
of these purified factors has now been reported. Dardenne et al. (70)
sequentially purified and finally isolated the smallest of these factors,
a circulating thymic factor (FTS), from 1000 liters of pig serum by ultra
filtration, gel filtration, and ion exchange chromatography. The factor
is a nonapeptide of sequence <Glu-Ala-Lys—Ser-Gln-Gly-Gly-Ser-Asn with

a molecular weight of about 900 daltons (71). A synthetic peptide with
the FTS sequence was produced by the Merrifield technique which display-
ed full biological activity and identical behavior to FTS on TLC and
Sephadex G-25 (72). The properties possessed by FTS very closely re-
semble Trainin and Small's thymic humoral factor (THF) obtained from
calf thymus (73). The amino acid sequence of a second polypeptide thy-
mic hormone has been described by Schlesinger and G. Goldstein (74).
Thymopoietin II is a larger (i.e. 49 amino acids) molecule with a mole-
cular weight of 5562 daltons isolated from bovine thymus. These inves-
tigators were also able to synthesize an active peptide based on the

determined amino acid sequence, but they recovered only 3% of the

14

activity of the parent polypeptide (75). Since the fragment had been
based on an internal segment of the hormone, it was postulated that the
tertiary structure of TP II may be required for optimal biological acti-
vity. This peptide seems to bear no relationship to FTS or THF based on
amino acid sequencing. A. Goldstein et al. (76) have described a third

sequenced factor, thymosin a This was found to be one of approximate-

1.
1y 10-15 major components found in the relatively crude thymosin fraction
5 isolated from calf thymus. This polypeptide is intermediate in length
between FTS and TP II (i.e. 28 amino acid residues) and is heat stable

and highly acidic. No sequence homology was demonstrable between thy-

mosin a and TP II, UBIP, FTS or THF. It was found that thymosin a

1 1

was more potent than fraction 5 in some T cell assays but not in others
suggesting that more than one peptide component may be necessary to
elicit full immunologic responsiveness. Whatever the case, it is ap-
parent that a family of thymic factors exist, some of which appear to
have overlapping functions. Perhaps, as suggested by Each (71), more
crude preparations, such as thymosin fraction 5, contain precursors of
the smaller, more defined thymic factors. This hypothesis was suggested
by a rise in serum FTS levels detected by the rosette assay after thymo-
sin injection (77).

Sodium Butyrate

 

Butyrate treatment of cells. Short chain fatty acids and other

 

lipophilic acids, some of which are used as antimicrobial food addi-

tives, inhibit bacterial growth and the uptake of amino and keto acids
into bacterial membrane vesicles (78,79,80). Growth of certain mamma-
lian cells in tissue culture also were inhibited by millimolar concen-

trations of sodium butyrate, a four carbon, short chain fatty acid (81).

15

At higher concentrations (10 m§_and above), butyrate has been shown to
be toxic to human cells in culture (82). More detailed studies of the
effects of sodium butyrate were initiated after it was reported that the
morphological and growth rate changes induced in tissue culture by
dibutyrl-cAMP (DB-cAMP) could be mimicked by the control, sodium buty-
rate, alone (83).

Butyrate and cAMP. When DB-cAMP is added to the medium of certain

 

cultured mammalian cells, the growth rate of the cells decreases and the
morphology changes from a compact multi-layer to a monolayer of cells
that appear totxamore fibroblastic (83,85). Compounds which raise in-

tracellular levels of cAMP, such as prostaglandin E , also elicit these

1
growth changes (86). Wright (83), while studying the effect of DB-cAMP
on morphology and growth rate of cultured Chinese hamster ovary (CHO)
cells, discovered that 0.5 m§_sodium butyrate produced similar results
to those of the cyclic nucleotide. The changes were not identical since
much longer incubation was needed in the presence of sodium butyrate to
observe similar alterations in cellular morphology. The investigator
speculated that either 1) the lipophilic fatty acid was affecting mem-
brane bound adenyl cyclase thereby causing adjustments in cAMP levels;
or 2) cell morphology, being dependent on the lipid content of the cell
surface, was affected by the incorporation of sodium butyrate into the
lipid component of the membrane.

Attempts have been made to measure directly the effects of butyrate
on cAMP metabolism. In mouse neuroblastoma cells, addition of sodium
butyrate elicited reversible inhibition of cell division but did not

cause morphological differentiation as does DB-cAMP (87). It was shown

in these cells that incubation with 0.5 m§_sodium butyrate raised the

16

intracellular levels of cAMP two—fold (88,89). The basal level of
adenyl cyclase activity in the mouse cells was also increased approx-
imately two-fold (90). On the other hand, Tallman et al. (91) did not
find increased adenylate cyclase activity, basal or fluoride activated,
in sodium butyrate treated HeLa cells. They did report, however, in-
creases in l-isoproterenol stimulated cyclase activity and cAMP accum-
ulation, as well as increased synthesis of beta-adrenergic receptors.

Effects of sodium butyrate on other cell types showed a similar
array of results (90). Low concentrations of butyrate induced similar
morphological or growth rate changes in certain epithelial cell lines
to those elicited in CHO cells. However, these epithelial cells were
not inducible by exogenous cAMP (81,92). Butyrate, therefore, appears
to act on mammalian cell cultures in at least two fashions. One mode
mimics the action of increased intracellular levels of cAMP, and the
other is unique to sodium butyrate itself.

Butyrate and glycolipid metabolism. Butyrate and other short chain

 

fatty acids up to hexanoate induced filamentous protrusions in epithe-
lial-like cell lines such as HeLa, Chang liver, L-132 and Intestine

407 (81). These morphological effects were not mediated or enhanced
by DB-cAMP (81). In the course of examining the growth inhibition and
marked morphological alteration induced in HeLa cells by 5 m§_sodium
butyrate, Fishman et al. (92) discovered a correlation between the mor-
phology of HeLa cells in culture and their glycolipid content. In the
presence of butyrate, HeLa cells contained 3.5 to 5 times more of the
ganglioside GM3 than in its absence. This coincided with a 7-20 fold

increase of the specific activity of CMP-sialic acid:lactosylceramide

sialyltransferase within 24 hours of butyrate treatment (93,94). No

17

other glycosphingolipid enzyme or product was significantly altered by
butyrate exposure. In addition, these intracellular glycolipid meta-
bolic changes preceded any morphological alterations of the cell sur-
face, prompting the suggestion of a possible role for glycolipids in the
control of cell growth and structure (92).

The effects of butyrate treatment on glycolipid metabolism have
been studied most extensively in HeLa cells. However, it has also been
shown that treatment of neuroblastoma cells with butyrate resulted in a
substantial change in ganglioside composition and an increase in two
sialyltransferase activities (30). Moskal et al. (95) reported a simi—
lar increase in sialyltransferase activity in these cells treated with
DB-cAMP. Furthermore, human epidermoid carcinoma (KB) cells have been
examined for changes in glycosphingolipid metabolism in the presence of
butyrate (96). KB cells exhibit characteristic process formation under
butyrate treatment. Likewise, CMP-sialic acid:lactosylceramide sialy-
transferase was stimulated nine-fold. Using cell surface labeling tech-
niques, the major glycolipid of the cell (GL-3) was increased two-fold
on the surface of butyrate treated versus control cells. In addition,
the activity of a specific galactosyl transferase was inhibited 50%
after exposure to 2-4 9E butyrate for 15-24 hours. On the other hand,
in Chinese hamster ovary cells, butyrate did not increase levels of GM3
ganglioside or its specific sialyltransferase (97). However, this may
be because both are present normally at levels which are comparable to
the induced levels measured in HeLa cells.

The manner in which butyrate-induced changes in ganglioside meta-
bolism are related to morphological differentiation of cultured mamma-

lian cells remains to be determined. Preliminary data from cell surface

18

labeling experiments indicated that approximately five-fold more GM3
was labeled in butyrate treated HeLa cells than in control cells (30).

Fishman and Brady postulate, therefore, that newly synthesized ganglio-
side may be incorporated into the plasma membrane where it functions in

a physical—chemical manner to promote extension of cell processes.

Molecular level of butyrate action. Butyrate has been shown to

 

affect cAMP and glycolipid metabolism in certain mammalian cell cultures.
Investigators have also examined the consequences of butyrate treatment
on protein metabolism. In HeLa cells, process formation induced by bu-
tyrate was sensitive to inhibition by actinomycin D and cychoheximide
(81,93). In addition, neurite extension was blocked by colcemid (93)
and the calcium ionophore A23187 (97). These findings implied that
butyrate required both protein synthesis and microtubule assembly to
exert its morphological effects.

Nucleic acid metabolism has also been studied in mouse neuroblas-
toma cells incubated in the presence of 0.5 mM_butyrate for three days.
The DNA content per cell decreased by about one half (98) and total RNA
and protein content increased 2-3 fold (99). In HeLa cells and chick
embryo fibroblasts, DNA synthesis, as measured by uptake of 3H-thymi-
dine, was also strongly inhibited after butyrate exposure (100). In
contrast to neuroblastoma cells, however, incorporation of labeled pro-
line or uridine in HeLa cells showed that butyrate had little or no
effect on overall RNA or protein content (100,101). It is not known
whether these differences are caused by different assay methods to de-
termine RNA and protein content, or by inherent differences in the cells
themselves in their responses to butyrate treatment. During closer

examination of neuroblastoma cells, it was found that, while the total

19

RNA content of butyrate treated cells increased, the concentration of
poly A containing cytoplasmic RNA (presumed to be messenger RNA) did
not change (99). Finally, in butyrate induced erythroid differentia-
tion in cultured erythroleukemic cells, increased synthesis of a speci-
fic mRNA coding for globin has been demonstrated (102,103). It appears
that careful examination of the different RNA types in these cells will
be necessary before an overall hypothesis on the effect of butyrate on
RNA metabolism can be stated.

An examination of the early time course of inhibition of DNA syn-
thesis in HeLa cells and chick embryo fibroblasts demonstrated that
after only one hour the rate of thymidine incorporation had fallen to
68% of the control (100). After 24 hours, DNA synthesis in butyrate
treated cells was only 2-10% of control cells. Histone modification
in both HeLa and Friend erythroleukemia cells has been shown to occur
(100) by a time course which closely resembles the one of DNA synthesis
inhibition. Acetylation of histone H4 can be demonstrated as early as
one hour after the addition of butyrate to the medium (104). This has
been shown to be due to an inhibition of histone deacetylase activity
(105) .

The thymidine incorporation experiments have led several investiga-
tors to postulate that butyrate treatment may led to cell synchrony.
Hagopian et al. (100) suggest from the shape of the time course of
inhibition of DNA synthesis and histone acetylation that cells in S
phase during initial butyrate treatment are halted there and that others
cannot enter it. Others (90,93,96) have found similar results and pro-
pose that butyrate may be synchronizing cells in the G1 phase of the

cell cycle. The fact that butyrate arrests cells in 61 or S phase does

20

not entirely account for its action, however, since cells blocked in S
phase by an alternate method do not display characteristic morphological
or enzymatic alterations until butyrate is added (93). The fact that
glycolipid biosynthesis has been shown to be maximal during G1 (106)

may explain why glycosphingolipid levels appear to be so sensitive to
butyrate treatment.

Specificity_of butyrate. In Wright's original comparison of morpho-

 

logical changes induced by butyrate and DB-cAMP, he reported that sodium
isobutyrate had no effect on the growth of CHO cells (83). Simmons et
al. (93) found that C3, C4 or C5 saturated fatty acids produced typical
process formation and caused enzyme induction in treated HeLa cells.
Analogs of butyrate also were shown to have no effect on morphology.

In addition, specificity has been demonstrated by Leder and Leder (102)
in induction of erythroid differentiation in cultured erythroleukemic
cells. They reported a requirement for a three carbon, straight chain,
aliphatic residue attached to a carboxyl group. This requirement was
based on the ineffectiveness of longer or shorter related fatty acids
in eliciting differentiation and on the ineffectiveness of analogues
such as isobutyramide, N-butyramine, butylacetate and 2-pentanone.
Similar observations on the specificity of sodium butyrate induction

of alkaline phosphatase activity in HeLa and other mammalian cells have
been reported (107).

Target Cell for Inducing Agents

 

Stem Cells. Maximow postulated the existence of a pluripotent

 

hemopoietic stem cell in 1909. Subsequently, controversy between
proponents of monophyletic, dualistic and polyphyletic theories of

blood cell differentiation existed for many years. It was not until

21

Till and McCulloch (108) described a technique for the detection of
primitive haemopoietic cells present in low frequency in mouse spleen
and bone marrow that the controversy abated. The cells were capable
of forming macroscopically visible colonies of differentiating gran-
ulocytic, erythroid and megakaryocytic cells in spleens of lethally
irradiated mice and were termed in viva colony forming units (CFU).
The clonal nature of these colonies was implied by the linearity of
the plot relating the number of marrow cells injected to the number
of colonies that developed in the spleen (100). More direct evidence
was obtained with the use of radiation induced chromosomal markers
(110,111). Wu et al. (111) reported that 91-100% of mitoses were
found to be of unique karyotype in any one colony and that 10 out of
12 colonies showed the same unique marker for both erythroid or gran-
ulocytic metaphases. In light of such experiments, most investigators
now agree the in viva CPU is a multipotential stem cell (112).

Identification of CFU. Many attempts have been made to identify

 

CFU's functionally and morphologically and an equal number of candidate
stem cells have been proposed. The small lymphocytes of the bone marrow,
as suggested by Cudkowicz et al. (113) have often been thought to ful-
fill this capacity, as have certain 'transitional lymphocytes' (114)
described as large basophillic blast cells. Combining these theories,
Yoffey (115) has postulated that no single specific stem cell exists,

but rather that there is a compartment of cells which is in steady state.
This lymphocyte transitional (LT) cell compartment is proposed to con-
sist of 80-90% pachychromatic small lymphocytes and 10-20% leptochroma-
tic transitional cells of various sizes and degrees of basophilia. Iden-

tification studies such as these have been hampered by the low numbers of

22

stem cells present in hemopoietic tissue. VanBekkum et al. (116) concen-
trated this population by repeated density centrifugation which was mon-
itored by an in vitra assay for CFU's. The stem cell enumerated in this
thin layer agar colony technique was termed a CFU-C and was thought to
represent a similar cell to CPU (117). Both light and electron micro-
scopy were used to describe the hemopoietic stem cell (HSC) in mice (116)
and in human and monkey (118) bone marrow. The candidate HSC was shown
to be quite distinct from small lymphocytes and close to Yoffey's lympho-
transitional cell. While these investigators felt that they had identi-
fied the HSC, others criticized a strict morphological approach for
identification. Experiments such as those of Worton et al. (119) sup-
ported the criticisms. In these studies, velocity sedimentation was
used to separate both CFU-S and CFU-C populations according to size.
Broad peaks were found for both groups indicating that the functionally
homogeneous populations were heterogeneous in respect to size.

Relationship between CFU's and T lymphocytes. Much work has been

 

done in an effort to delineate the relationship between CFU's and T
lymphocytes. Originally, it had been postulated by Auerbach (120), that
lymphoid cells in the thymus arose from the thymic epithelial compartment.
However, it has since been shown that thymocytes develop from stem cells
which migrate into the thymus from the yolk sac in embryonic chicks (121)
and from the fetal liver in mice (122). Furthermore, in adult animals,
chromosomal marker studies have demonstrated migration of stem cells
between various hemopoietic organs, including traffic between bone mar-
row and thymus (123). Owen and Raff (124) reported that these cells
appeared to be large blast-like cells which did not carry thymocyte

alloantigens such as Thy-1 or TL on their surface. This implied that

23

they were undifferentiated precursors, perhaps CFU's. Indeed, the
occurrence of spleen lymphoid colonies in the CPU assay was claimed by
Mekori et al. (125) following the injection of large numbers of thymic
or lymph node cells from phytohemagglutinin treated donors into irra-
diated recipient mice. However, others (126) could not duplicate these
results and the claim was subsequently retracted. Since then, it has
been postulated that lymphocytes do not form aggregates in viva due to
their migratory properties (127). This alone could account for the
lack of lymphoid colonies in CPU assays.

Despite the failure of lymphoid colonies to occur, precursors of
lymphoid populations have the same origin as CFU. Using radiation in-
duced chromosomal markers to identify the progeny of singly hemopoietic
stem cells, Wu et al. (128) found that cells with the same abnormal ka-
ryotype as CFU's in the spleen reached the thymus and lymph nodes of
recipients within 1 month. Whether all CFU's are capable of differen-
tiating into both myeloid and lymphoid cells, or whether CFU's and
lymphocytes have a common precursor was not established.

Attempts to distinguish between CFU's and thymic factor sensitive
cells have been made. Separation of bone marrow and spleen by density
centrifugation has led to the finding that both CPU and prothymocyte
concentrations are enriched in low density fractions of discontinuous
BSA density gradients (49,129). However, unlike precursor T cells, CFU's
have also been demonstrated in fractions of higher density (130). Den-
sity estimates for the low density CFU's isolated in such experiments
vary slightly. El-Arini et al. (131) have reported that cells in the
low density region of 1.050-l.059 gm/cm3 become immunocompetent within

15 days after injection into the spleen of irradiated isogeneic

24

recipients. The results of Turner et al. (129), who estimate the den-
sity of the CFU's at around 1.051 gm/cm3, corroborate that finding.
Dicke et al. (118) found their HSC to migrate between the slightly
higher density range of l.055-1.0682 gm/cm3. Some overlap was demon-
strated when density estimates of the putative thymic precursor were
reported. For example, Incefy et al. (132) found that the layer of
cells settling on the 21-23% interface of a BSA gradient (density
l.0552—1.065 gm/cm3) contained an enriched population of prothymocytes.
Others have reported densities of 1.064 gm/cm3 (131) and 1.068 gm/cm3
for progenitor T cells (133). Therefore, while it appears that the
precursor T cell may be slightly denser than a CFU, the minor differ-
ences in density distribution are not sufficient to distinguish be-
tween the two cell types.

Other evidence bearing on the relationship of CFU's to prothymo—
cytes has been reported. Thymic precursor cells have been found to be
capable of repopulating the thymus of a lethally irradiated mouse (134)
as do CFU's. However, these repopulating cells have a different density
from CFU's (131). Also, it has been shown that prothymocytes can be in-
duced to express thymocyte differentiation antigens without influencing
the number or distribution of CFU's (135). Taken together, the evidence
suggests that the two cell populations are not the same.

Thymic precursor cells. Some cell surface characteristics of the

 

suspected precursor T cell have been defined. They display H-2 antigen
and at least two antigens recognized by antisera to mouse brain (133,136).
Other investigators have reported that prothymocytes lack detectable
amounts of surface markers associated with either B or T cells (133)

placing them in the "null" cell population of the spleen (137). On the

25

other hand, Loor and Roelants (128) reported that, while a precursor T
cell has no membrane Ig, it does possess a low density of Thy-l antigen
which is detectable only by indirect immunofluorescence. Their finding
has been disputed in light of the multiple specificities of antisera
used to detect Thy-l (133).

Recently, the use of an internal cell marker for the identification
of precursor T cells has been suggested (139). Terminal deoxyribonucleo-
tidyl transferase (TdT) catalyzes the polymerization of deoxyribonucleo-
tides in the presence of a primer without the requirement for a template
(140). This enzyme is found uniquely associated with the thymus and bone
marrow of mice (141). It can also be demonstrated in a developing chick
embryo thymus (142). Pazmino et al. (143) found (TdT) only in the low
density (lo-23%) fraction of marrow separated by BSA density gradient.
Furthermore, treatment of bone marrow cells with thymopoietin can make
the TdT positive cells sensitive to anti-Thy-l sera and complement (139).
Both of these are properties which mimic those of prothymocytes.
Silverstone et al. (139) postulate that TdT is the earliest known marker
of cells that can undergo thymus dependent development. Others (144),
however, have shown that TdT can also be demonstrated in the leukemic
cells of a patient with B cell acute lymphocytic leukemia. These
authors feel that TdT is related more to the immaturity and prolifera-
tion of certain lymphoid cells than to their progress toward B or T
cell differentiation.

In erythropoiesis, it has been suggested that the erythropoietin
sensitive cell (ESC) is already permanently committed to a restricted
pathway of differentiation (145). Following irradiation, recovery rates

of colony forming units differed from erythropoietic responsiveness.

26

Indirect evidence for pre-determined differentiation also exists for
progenitor T cells. Scheid et al. (54) have demonstrated that a var-
iety of non—thymic agents were capable of inducing T cell differentia-
tion antigens on precursor cells. It has also been reported that thy-
mic factors can induce T antigen expression on cells isolated from nu/nu
mice which are genetically athymic and on cells from fetal liver (48,146).
However, the fact that the presence of a thymus is not needed to induce
differentiation is only a presumptive basis for predetermination. It
remains possible that these cells are "proimmunocytes," rather than
prothymocytes, capable of differentiation into either T or B cells de-
pending on the environment to which they migrate (133). This might
explain the existence of cells which express both B and T cell markers
(147).

Stem Cell Separation Techniques

 

Introduction. Stem cells and prothymocytes have been isolated from

 

two main sources in the mouse: the bone marrow and the spleen. These
two hemopoietic organs contain a highly heterogeneous population of
cells of both lymphoid and nonlymphoid origin. Furthermore, Dicke et
al. (118) have estimated that the pluripotent hemopoietic stem cell is
present at a concentration of only four to six per 1000 cells in bone
marrow. This fact has hampered the development of definitive studies
designed to morphologically and functionally identify stem cells and
committed progenitor cells. To alleviate this problem, a variety of
cell separation techniques have been applied to spleen and marrow sus-
pensions. These fall into four main catagories, l) filtration or ad-
hesion to glass and nylon wools (148-151) columns, 2) isopycnic density

gradient separation (49,63,115,152), 3) velocity sedimentation (153—155),

27

and 4) two dimensional cell separation (69,156,157).

Glass and nylon wool column separations. Glass wool or beads have
been used to partition cell suspensions since it was reported that it
was possible to purify peripheral blood leukocytes by allowing them to
adhere to glass bottles (158). Cudkowicz et al. (159) demonstrated the
partial purification of a suspected hemopoietic stem cell which resembled
a small lymphocyte after the passage of mouse bone marrow through columns
of glass wool. More recently, others (149) have used a similar technique
to concentrate stem cell populations and have obtained a 5-fold increase
in stem cells relative to unfiltered marrow. Progenitor T cells have
also been found to be glass wool nonadherent (133) when spleen cells
suspensions are passed over glass wool columns.

Further fractionation of lymphocytes and their precursors can be
obtained if filtration over a glass wool column is followed by passage
over a nylon wool column. Trizio and Cudkowicz (160) reported the sep-
aration of functional B and T cells from spleens with this double column
technique. Both cell populations passed through glass wool columns
while only T lymphocytes were found in the effluent fraction of nylon
wool columns. Conflicting reports exist, however, on the adherence of
precursor T cells to nylon wool. It was shown by Twomey et al. (151)
that precursor T cells from spleen cell suspensions were eluted from
nylon wool columns. Prothymocyte populations were demonstrated by the
induction of Thy-l antigen after incubation with thymopoietin and ubiq-
uitin. In contrast, Basch and Kadish (133) found thymocyte precursors
in both the adherent and nonadherent fractions after passage of spleno-
cytes through nylon fibers. In this case, a quantitative assay which

determined the number of donor type cells to repopulate the thymus of

28

an irradiated recipient was employed to monitor the presence of progen-
itor T cells (161). It is possible that the differences in assays sys-
tems employed may account for this apparent discrepancy or that subpop-
ulations of precursor cells may exist.

The reason for the adherence of certain cell types to either glass
or nylon wool remains unknown. It has been suggested that an active
metabolic process is involved since cell separation values were low
when columns were maintained and operated at 4°C rather than at room
temperature or 37°C (149). In addition, siliconized glass bead columns
separated cells by a temperature independent selective trapping by phy-
sical adherence or by size (162).

Isopycnic densitngradient separation. The most commonly employed

 

method to fractionate thymic precursor cells from heterogeneous hemopo-
ietic populations has been isopycnic density gradient separation. In
this technique, cells centrifuged in continuous or discontinuous den-
sity gradients are banded in locations in the gradient that are equal
to their own densities. Previous to the description of a linear bovine
albumin gradient by Leif and Vinograd (163), density separation and
determination of cells had been done either by the packed cell method
in which a cell pellet was separated by layers or by the neutral den-
sity method in which cells were centrifuged in a dense liquid medium
and segregated into pelleted and floating fractions (164). Since the
introduction of gradient density separation, however, these tedious
methods have been abandoned.

Both stem cell (115) and prothymocytes (49) populations have been
concentrated with the aid of discontinuous bovine serum albumin density

gradients, and their suspected densities have been reported. These vary

29

from 1.051 gm/cm3 for the hemopoietic stem cell (129) to 1.068 gm/cm3
for a purported progenitor T cell (133).

A wide variety of gradient media have been examined and Ficoll and
bovine serum albumin have been used most often. These substances leave
the desired cell populations both viable and functional after separation.
Recently, a new gradient medium has been described, metrizamide, which is
a non-ionic, tri-iodinated benzamide derivative of glucose. This sub-
stance appears suitable for gradient separation of living, intact cells
since it is possible to prepare isotonic, buffered metrizamide solutions
of high density and low viscosity. It has been used successfully for
blood cell separation (165), fractionation of rat liver cells (166), and
separation of granulocyte-macrophage progenitor cells from mouse bone
marrow (167).

Velocity sedimentation. Hemopoietic cell populations have also

 

been fractionated by velocity sedimentation techniques. Miller and
Phillips (168) developed a procedure in which a thin band of cells was
allowed to sediment through a shallow gradient of 3-30% fetal calf
serum at unit gravity. Velocity sedimentation such as this partitions
cells on the basis of size, or volume, with only a small dependence on
cell density.

Both stem cells and progenitor cells have been separated by velo-
city sedimentation. Worton et al. utilized this technique to demon-
strate that stem cells differing in their self renewal capabilities
could be fractionated on the basis of size (169). In contrast, others
(118), have shown similar distribution profiles between CFU-S and CFU-C.
Reasons for this difference may lie in differences in procedures.

Velocity sedimentation also possesses the capability of partitioning

3O

progenitor cells of various classes. The thymocyte precursor cell has
recently been examined by a modification of this technique (133) and has
been shown to be similar in size to medium to large lymphocytes (i.e.
10-15 pm in diameter). Lafleur et al. (170) have found that B lympho-
cytes can be partitioned from their precursor cells by sedimentation at
unit gravity. Finally, McCool (171) has been able to isolate erythro-
poietin sensitive cells from rat bone marrow with this procedure. Frac-
tionation of cell suspensions by velocity sedimentation at unit gravity
appears, therefore, to be a useful separation method.

Two dimensional cell separation. Several groups of investigators

 

have combined two separation techniques based on distinct parameters in
an effort to obtain more purified stem or progenitor cell populations.
Haskill and Moore (157) employed both equilibrium density and velocity
sedimentation to resolve differences between murine embryonic and adult
hemopoietic stem cells. Also, Kagan et al. (69) separated human bone
marrow first by density on Ficol-Hypaque gradients and then by size by
velocity sedimentation to isolate a granulocyte progenitor population.
In a third study, immature myeloid cells from human bone marrow were
enriched by density and size as well as by hypotonic shock (to remove
erythroid precursors) (156). After the final step of velocity sedimen-
tation at unit gravity, fractions were collected that contained an aver-
age of 62.6% immature myeloid cells versus an average of 8.9% in the
original unprocessed marrow. Finally, Paterson et al. (172) reported
on the partial purification of lung mast cells after separations by
both density and size on metrizamide gradients. Further examination of
combination techniques should yield progressively purer populations of

cell types from heterogeneous mixtures such as bone marrow and spleen.

31

Additional separation techniques. Several other cell separation

 

methods have been employed which rely on cell surface properties direct-
ly. Immunoabsorbent column chromatography and rosette-depletion techni-
ques have been used to eliminate cells with known B and T cell markers
from peripheral blood mononuclear cells (173). This was done in an
effort to concentrate the "null" cell population which has been thought
to contain both stem and progenitor cells. Others (174) have identified
mouse lymphocyte subpopulations by their natural binding of eight strains
of bacteria. Differences in cell surface charges have been utilized to
fractionate cell populations by electrophoresis (175) and by counter
current distribution (176). Cells separated by such techniques remain
both viable and functional. Finally, Herzenberg et al. (177) have util-
ized fluorescent labeled antibody to specific membrane receptors to
partition cells with the aid of a fluorescence-activated cell sorter.
This technique allows greater specificity than is afforded by physical

criteria such as size and density alone.

INTRODUCTION TO EXPERIMENTAL REPORT

The ability of the thymus and thymic extracts to induce different
stages of T cell differentiation has been documented by a variety of
in viva (34,36,37) and in vitra (44,48,55) assay systems. One of these
assays, originally descirbed by Komuro and Boyse (48), monitors the in
vitra expression of T cell specific differentiation antigens (Thy-l and
TL) on the surface of prothymocytes which are found in low numbers in
the bone marrow and the spleen. Antigen appearance during incubation
of precursor cells with thymic factor occurs in two hours or less, and
a direct cytotoxicity test is employed to quantitate the percentage of
cells induced to express the antigens. This is the earliest known effect
of thymic factor on precursor T cells. Since the Thy-l conversion assay
appeared to be both rapid and precise, many investigators began to use
it to assess tissue extracts and other compounds for thymic hormone
activity (42,54,65). Others used it to examine the mode of action of
thymic factor (63,64). Milewicz et al. (63) suggested that the rapid
appearance of Thy-l by thymic factor treatment was the result of the
modification of existing membrane compounds or the expression of a cryp-
tic antigen on the cell's surface. They also utilized the system to
attempt to characterize the Thy-l antigen chemically by comparing the

expression of Thy-l and G , a ganglioside which carries the Thy-l allo-

Ml
antigen, on the surface of the precursor cell.

Recently, it has been reported that difficulties exist with the
reproduction of the Thy-l conversion assay. Fournier and Bach (178)
have found that spontaneous induction of Thy-l antigen occurred in the

absence of any inducing agent nearly 50% of the time after bovine serum

albumin density gradient separation of splenocytes. Twomey et al. (151)

32

33

reported that a two hour incubation of splenocytes with thymic factor
led to low or variable results. They found it necessary to employ an
eighteen hour incubation of precursor cells with thymic hormone in order
to obtain consistency.

Several parameters involved in the Thy-l conversion assay have been
examined in this report in an attempt to develop a more reproducible
assay. The sensitivity of the cytotoxicity test used to detect the
appearance of Thy-l antigen on precursor cells has been studied. Also,
this report examines methods used to separate the low numbers of precur-
sor T cells from the heterogeneous population of hemopoietic cells which
exist in the bone marrow and spleen. Finally, the discovery of a new
agent capable of inducing Thy-l antigen expression on precursor cells is
reported here. Sodium butyrate at low concentrations is shown to induce
antigen appearance in the conversion assay in a similar fashion to thymic

factor.

MATERIALS AND METHODS

Cells: Inbred C3H/HeJ and AKR/J female mice (Jackson Laboratories,
Bar Harbor, ME) and outbred ICR-Swiss female mice (Spartan Research
Farms, Haslett, MI) were used at ages 8 to 20 weeks as donors of splen-
ocytes, thymocytes, and bone marrow cells. Bone marrow was aspirated
from the femurs and tibias of sacrificed animals into Minimal Essential
Media (MEM-Grand Island Biological, Buffalo, NY) with the aid of a 25-
gauge needle and further disaggregated by passage through a 27-gauge
needle. After centrifugation, seven to ten mice depending upon their
age and strain, would yield from 2-5 x 108 nucleated cells of 85% via-
bility by the trypan blue exclusion method. Spleens or thymuses were
gently teased with forceps into MEM. Single cell suspensions of spleno-
cytes or thymocytes were obtained by passage through a 27-gauge needle.
After washing by centrifugation, one spleen would yield approximately
l—2 x 108 nucleated cells of 85% viability dependent upon age and strain.
Both the bone marrow cells and the splenocytes were then subjected to
the various separation procedures described below.

Cell separation techniques. Populations of bone marrow cells or

 

splenocytes were fractionated on the basis of two distinct parameters--
either by their densities or by their ability to adhere to glass or ny-
lon wool. Density fractionation was achieved by centrifugation of the
cells on a discontinuous gradient of bovine serum albumin (BSA) as des-
cribed by Milewicz et al. (63). BSA stock solutions were screened for
endotoxin contamination by the Limulus lysate assay (179). This assay
was kindly performed by Dr. Robert Moon of Michigan State University.
The low density prothymocytes were found in equilibrium with the 21% and

23% layers of a 17-27% step gradient. These were washed twice and then

34

35

pooled for direct use in the conversion assay described below. Upon
examination by light microscopy, cells from the pooled fractions were
found to be primarily mononuclear and to have a viability of greater
than 95%. A small degree of contamination by polymorphonuclear cells
and erythrocytes was present. The latter could be eliminated by sub-
jecting the cell suspension to cold hypotonic shock with a NH4Cl-EDTA—
HEPES buffer (0.155 M, 0.1 mM, 1.0 mg, pH 7.35) for five minutes follow-
ed by several washes with fresh medium (180). The pooled 21% and 23%
layers accounted for less than 10% of the total population of cells
applied to the gradient. When tested with the appropriate antiserum and
complement, these layers from bone marrow suspensions were found to be
only 1-2% Thy-l positive, and those from the spleen were less than 5%
positive.

Fractionation based on adherence properties essentially followed
the procedure developed for the separation of B and T lymphocytes (160).
The only major exception to the originally described procedure was the
substitution of BSA (2.5 mg/ml) for the 10% fetal calf serum in the incu-
bation and elution media. Cells were first passed over a loosely packed
glass wool (Corning Glass Works, Corning, NY) column, pelleted, and sub-
jected to cold hypotonic shock. The remaining cells were then incubated
on nylon wool (Fenwal Laboratories, Morton Grove, IL) columns for 45 min-
utes at 37°C. Nonadherent and adherent cells were collected. The efflu-
ent fraction, which accounted for 5-6% of the bone marrow cells and 10-
15% of the splenocytes applied to the nylon column, consisted of pri-
marily mononuclear cells with a few polymorphonuclear cells. The nylon

wool adherent fraction contained all major cell types of the original

suspension. Both fractions had viabilities of greater than 80%.

36

In vitra Thy-l conversion assay. Marrow and spleen cells

 

(106 cells/m1) separated by the above procedures were incubated for

two hours at 37°C, 5% CO2 in the presence of either thymic factor

(15-90 ug/ml) as previously described (181) or in the presence of
sodium butyrate (2.E§) to induce the expression of Thy-l antigen.

Thymic factor (TF), fraction 3, was prepared from fresh calf thymus
according to Hooper et al. (182) and was used at the concentration of
90 ug protein/ml. Commercially prepared TF, fraction 5 (Hoffman-
Laroche Inc., Nutley, NJ) was utilized at 15 ug protein/ml. Stock
sodium butyrate (100.Eflv pH 7.0) was prepared by the addition of

sodium hydroxide to butyric acid (Sigma, St. Louis, MO). BSA (20 pg/ml)
was used as a control. Following incubation, the cell suspensions were
pelleted by centrifugation at 1600 rpm for 10 minutes, washed once,

and resuspended in fresh medium at a concentration of 2.5 x 106 cells/ml

for use in the cytotoxicity test.

Detection of expressed Thy-1 antigen - Antisera and complement.

 

AKR anti-C3H (anti-Thy-l.2) and C3H anti-AKR (anti-Thy-l.1) antisera
were prepared according to Reif and Allen (1). Titers of these sera
were approximately 3200 and 1500 respectively. Normal AKR and C3H sera
were used as controls. Guinea pig complement (Grand Island Biological,
Buffalo, NY) was adsorbed with agarose and stored in aliquots at -70°C
dilutes 1:3 with MEM.

Cytotoxicity test. Treated bone marrow and spleen cells were test-

 

ed for the expression of Thy-1 antigen in a direct cytotoxicity test pat—
terned after that of Boyse (183) which used the uptake of trypan blue as
a measure of cell death. 106 washed cells were suspended in 360 pl of

MEM. To this was added 20 pl of the appropriate antisera (final dilution

37

l/2000) and 20 ul of complement (final dilution 1/60). Complement or
normal sera controls were adjusted with MEM to obtain the same final
volume. The samples were then incubated for 45 minutes at 37°C, 5% CO2
following which trypan blue was added to a final dilution of 0.02%. Both
live and dead (i.e. blue) cells in a constant volume were counted in an
eosinophil counter. The percent of conversion to Thy-l positive cells

was then calculated by two methods.

(1) % Thy-1 positive = % cells dead with - % cells dead with
cells anti-Thy-l sera control sera

(2) % Thy-1 positive # viable cells in - # viable cells in
cells control anti-Thy-l sample

# viable cells
in control

 

X 100

The first method was based on a comparison of the percent dead in the
control and anti-Thy-l treated samples. The second method utilized abso-
lute numbers of only viable cells to calculate percent conversion to
Thy—l positive cells. This approach accounted not only for dead cells
which have taken up the trypan blue dye but also for cells which had
undergone total lysis following antiserum and complement treatment (184).
Calculations based on the first method are referred to as % Thy-l posi-
tive cells by change in % dead. Those based on the second formula are
referred to as % Thy-1 positive cells by change in number viable. All
samples have been done in duplicate or triplicate and are presented as
means with standard errors.

Absorption of anti-Thyrl sera. Absorption of anti-Thy-l.2 sera

 

was patterned after Basch and Goldstein (42). A dilution of anti-Thy—l.2
sera that gave approximately 50% cytotoxicity with C3H thymocytes was ab-
sorbed overnite at 4°C with various dilutions of butyrate or TF treated

. 6 5
ICRrSwiss treated bone marrow cells (ranging from 2 x 10 to 2.5 x 10 )

38

pooled from 21% and 23% layers of BSA density gradient. The serum was

centrifuged free of cells and tested for residual cytotoxicity for C3H

thymocytes. Equivalent numbers of BSA treated bone marrow cells were

used as a control.

RESULTS

Sensitivity of the Cytotoxicity Test

 

Increased detection of Thy-l inducible bone marrow cells. Bone

 

marrow cells, pooled from the 21% and 23% layers of a discontinuous BSA
density gradient, were treated with TF for two hours to induce the ex-
pression of Thy-l. The newly expressed antigen was then detected by a
direct cytotoxicity test. The sensitivity of this test could be manipu-
lated by changing the procedure for counting viable and dead cells after
the addition of trypan blue. These results are shown in Table I. The
percentage of cells induced to express Thy-l was obtained formerly by
subtracting the percentage of cells killed by control sera from the per—
centage of cells killed by anti-Thy-l sera (formula 1 in Materials and
Methods—-% Thy-1 positive cells by change in % dead). If care was taken
to insure that each sample being analyzed contained the same concentra—
tion of cells at the start of the cytotoxicity test, the percentage of
cells induced to express antigen could also be calculated in a more pre-
cise manner. This method involved counting only viable cells in a set
number of chambers in an eosinophil counter. The number of dead cells
was obtained arithmetically by subtracting the number of viable cells in
the anti-Thy-l treated sample from the number of viable cells in the
control sample (formula 2 in Materials and Methods--% Thy-l positive
cells by change in number viable). As seen in Table I, the percentages
of Thy-l induced cells based on viable cell counts were found to be
consistently two to three times greater than those based on a comparison
of death percentages. The increased detection of Thy-l positive cells
by viability counting was probably due to the fact that some lysed cells,

which were previously lost from the counting because of their presence

39

4O

.mmamsmm aumoflamsp Eoum manna onmpcmum new oomua><a
.moonumz can mamflnoumz ca pauswmwum mmasfiuom Eoum oaumaooamo mommucaOHmmU

.umau xuaoflxououho ou Hoaum mowzu owsmm3 auoz maaoo .Houucoo m

mm 1H5\mn omc amm no .Hs\:amuouo m: omIch me nuns Nco am .oomm um muses m Mom emumnsocg mums Hs\maamo mono

.coflumnoocfl he on Howum mowzu pmsmmz

pom xoonm aflcouom>o paoo mo posoaaom pcaflpmuo wuflmcmo «mm mo muame wmm pom mam oaaoom scum vacwmuno mHHoUQ

.ucmflwmum suamaon «mm «0 mumsmH wmm new new omHooo scum emcflmuno maamom

 

 

mm + mam mm m.HImmeIHeza
mm I. m
6H + mum on I me
I. x~.HIsmeV
ma + mmm Hm m.HI»meIHez< ammo
o .I H
me + mmm om I «mm
em.“ mam am H.HI»meIHeza
aa a
ov.H Ham AH I my
I. xa.aImmeo
om + mmo he H.HIwmeIHeza mmxa
o I. H
H + 5mm ma I «mm
xmgm<H> mmmzaz <5 mmqm<H> inane w <1 name a zammmHeza oezmzeamma zHamem
6m>HeHmom HIwme a cm>HeHmom HIume a

 

m090¢m UHZMmB Nm MAAWU 30mm¢2 mzom
MZHMDZ ZO N.HINEB 92¢ H.lemB m0 ZOHBUDDZH
H mqmdfi

41

only as debris, were included by this method. This phenomenon was not
due to increased cell fragility caused by the hypotonic shock treatment
used to eliminate contaminating erythrocytes since it occured with both
shocked (C3H) and nonshocked (AKR) cells (Table I). To insure that the
observed differences were significant, large numbers of viable cells
(normally 300 to 600 cells for each sample) were counted. Only small
standard errors were found between duplicate samples (generally less
than 10%). In addition, death percentages for both anti-Thy-l treated
and control samples were always calculated to validate duplicates. Such
precautions were necessary to accurately quantitate the relatively small
(15-20%) loss of viable cells to death and total lysis.

Detection of Thyel inducible splenogytes. In addition to their pre-

 

sence in bone marrow, precursor T cells have also been found in the adult
mouse's other major hemopoietic organ, the spleen. As with bone marrow
cells, incubation with TF induced the expression of Thy-1 antigen on ap-
proximately 15-20% of the low density fraction of BSA density gradient
separated splenocytes (Table II). However, unlike the case with bone
marrow percentages, the splenic percentage of Thy-1 inducible cells was
demonstrable by either of the two counting methods employed regardless
of the level of sensitivity. Since the assay systems used to induce
Thy-1 on the two cell types were identical, it is probable that some
difference may exist within the cells themselves which makes the precur-
sor splenocytes less susceptible to total lysis following anti-Thy-l and
complement treatment. A basis for this difference remains to be deter—

mined.

42

panflnomap mm UoEHOMHam :ofluooocfl m8

.monEMm gum mo momma pumpcmum pom ammno>mo

.mponuoz new mamflwoumz ca voucommum mmaashow scum caumHaOHMU ommucaoumm

.H magma ca

.ucoEmHoEoo one whom N.Hlxnalwucm nuw3 powwow non3 o>fluwmom N.HI>£B am
one» mmaa on on pcoom one ucowpmum wuflmcac dmm mo muowma wmm pom mam poaoom scum omcflmuno ouoz manhoocoflmmm

a

 

 

 

a.“ mom om +
he m.H mmm he mm I he
AmnmmH> mmmzoz <c omumaH> Leann m <o came a ~.HIw=eIHeza mezmzeamma
m>HeHmom HIwme w m>HeHmom HIwme w

a

Q

 

mOBU€m UHEwEB Mm mMBMUOZMAmm mmo
ZO mAAmU N.lem9 m0 ZOHBUDQZH

HH mflmdfi

43

Sodium Butyrate Induction of Thyfl Expression

 

Sodium butyrate at the concentration of 2.Efl induced the expression
of Thy-l antigen in the low density fractions of BSA gradient separated
bone marrow cells of mice (Table III). Butyrate treatment induced anti-
gen appearance only in cells of the 21% and 23% layers of the gradient
(results not shown) as does TF treatment (63). Both Thy-1.1 and Thy-1.2
were shown to be inducible by butyrate. When the two different counting
methods described above were employed to detect the number of Thy-l in-
ducible cells, the same increase in sensitivity was observed with the
use of viability counting as had been observed with TF treatment. Furth-
er similarities between TF and butyrate induction of antigen expression
are shown in Table IV. Approximately the same percentage (15-20%) of
bone marrow cells became Thy-l positive after butyrate as after TF treat-
ment. In addition, butyrate was capable of inducing Thy-l appearance in
low density splenocytes as well as low density marrow cells.

Absorption of Anti-Thy-l Sera by TF or Butyrate Treated Marrow Cells

 

It is possible that TF or butyrate treatment of marrow cells may
cause an artificial increase in the number of cells thought to be Thy-l
positive due to the exposure of additional complement receptors or to
the uncovering of autologous antigens or other nonspecific determinants.
Both TF and butyrate treated cells, however, were shown to have an abso-
lute increase in the amount of Thy-l present on their surfaces by their
ability to absorb the activity of anti-Thy-l sera (Figure l). Increas-
ing numbers of TF or butyrate treated marrow cells absorbed increasing
quantities of anti-Thy-l.2 antibodies from the sera, while control cells
treated with BSA had no capacity for absorption at any concentration.

In addition, the fact that TF and butyrate treated cells absorbed

44

m mm He}: 08 own no a? mo

.monEmm aumoHHmsp mo Houua photomum pom ammwm>¢

c

.mponumz pom maneuauwz ca owucamoum mmHSEHom scum woumanoamo mammucaouamo

.umou muwoflxou09>o ou moanm ooa3u wanna? 0Ha3 mHHoO
06 mm .oonm um muse: oz» now emumnsocH mums HE\mHHmo oH

oumu>uon Esflpom sues N

.Houucoo

m n

.ucmHemua suHmcmc «mm H0 mumsmH wmm can me emHooo scum omchpno mHHoom

 

 

mH + mow mm ~.HIwmeIHeza
mm I. HH
mm + mam «H I meamweam
I. . H~.HImeV
o m + mHm m mH m H was Haze mmHsmImoH
mH.H mum NH I «mm
mH.H oma am H.HI»maIHez<
mH .I 5
mm + mHm pH I meamwsom
I. .H.HI»mav
om + mmw pH H.HIwmaIHeza use
0 .I H
H + nmo 6H I «mm
qum¢H> amazoz <V omHm4H> Hoama w <v name a zcmmmHeza nezmze<mme mzHamem

Om>HBHmOm HINEB w

OW>HBHm0m HIMSE w

 

madmwabm EDHQOw Nm mQAmU 30mm¢2 mzom

mZHmDZ ZO N.HIMZB OZfl H.HIHSB m0 ZOHBUDDZH

HHH mnmdfi

45

.oGOG uozp

.mponuoz pom mHoHnoumz ca noncommnm mmHDEnom Baum nonmaaoamo mommucaonwmo

.HHH can H mmdnma CH confiuumww mm ucmfiummua

n

.ncmnnmnm anmcmo «mm no mnmmmH wmm I mHm eonn emHoom mHHoom

 

 

m ma ma oumnausm

v OH mH me Hm.HV

a Hv a «mm mmo :mmHom

m Hm HH wnmnsnsm

m mH 0H me HH.HV

m o o amm med

oz oz woz oumn>usm

m pH a ma H~.HV

m o Hv «mm mmo

e mH a onmnxnsm

0H mm a me H~.Hv

n m Hv amm mmHsmImoH gonna: mcom
mezmszmmxm HmHm<H> mmmzoz <o Hoamo a <1 nezmzeamme znamem mnemDOm Home

no mmmzaz

Om>HEHmOm HINEB m

Om>HBHmOm HINSB &

 

ZOHBflUDhHMBZmU BZMHDGmO MBHmZMD dmm Mm

DMB<ZOHBU<mm qumu ZWWAmm 02¢ 30mm¢2 mzom

MZHmDS ZH mAAmU m>HBHmOm HINmB ho ZOHBUDDZH

>H mqmdfi

46

FIGURE I

Absorption of anti-Thy-l.2 serum with TF and butyrate treated bone
marrow cells. ICR—Swiss bone marrow cells used for absorption were ob-
tained from pooled 21% and 23% layers of BSA density gradients. Anti-
Thy-l.2 sera of 50% cytotoxicity against C3H thymocytes was absorbed
overnite with BSA treated (O ) , TF treated (O ) , or butyrate treated
()() marrow cells at varying concentrations and tested for residual
cytotoxicity. Each point is an average of two experiments. Calcula-
tions are based on method 1 described in Materials and Methods. Further

details in Materials and Methods.

47

 

 

 

40
°/°\ *.
X
3% /::<.
§
Q J
320
Q)
U I
It
0.
l0-
0 . . .
20 Lo 05 025

Number Cells ((0'6)

48

anti-Thy-l sera in nearly identical patterns is further evidence that
butyrate induction of antigen expression is similar to that of TF.

Prothymocyte Separation-The use of Glass and Nylon Wool Columns

 

It has been suggested that inconsistent results in the in vitra
conversion assay may be due to increased cell fragility caused by BSA
fractionation and prolonged incubation with inducing agents (178). We
have documented an increase in the amount of cell lysis with TF and bu-
trate treated bone marrow cells. To test the hypothesis that this is
due to BSA fractionation, an alternative method of cell separation was
chosen--adherence to glass and nylon wool columns. The standard sepa-
ration technique was modified to use only low levels of BSA (2.5 mg/ml)
in place of the 10% fetal calf serum. It was also hoped that this me-
thod might improve on the purification of stem cells obtained by density
centrifugation. Results are shown in Table V. Both the nylon wool ad-
herent and nonadherent fractions were found to contain all Thy—l nega-
tive cells prior to TF treatment (results not shown). After treatment,
the effluent fraction had 20% Thy-l positive cells while the retained
fraction still had none. Enrichment of stem cells in the nylon wool
eluted fraction was, therefore, comparable to, but no greater than,
the low density fractions of a BSA gradient. It appeared that sub-
stituting low levels of BSA had no effect on the degree of cell fragil-
ity. Conversion percentages obtained with viability counting were still
twice those calculated with death percentages indicating that the same
ratio of cells expressing Thy-1 were being subjected to lysis and dis-
integration in a two hour period.

Similar results were obtained when splenocytes were separated by

glass and nylon wool columns (Table VI). The inducible cell population

49

TABLE V

INDUCTION OF THY-1.2 POSITIVE CELLS BY

THYMIC FACTOR IN C3H BONE MARROW FRACTIONATED

BY GLASS AND NYLON WOOL COLUMNS

 

% THY-1 POSITIVEC

% THY-l POSITIVEC

 

CELLSa TREATMENTb (A % DEAD) (A NUMBER VIABLE)
Nylon Wool TF 9 20

Eluted
Nylon Wool TF 0 0
Retained

 

aSee Materials and Methods for details.

bTreatment as described in Table I.

cAverage six experiments - Percentages calculated from formulas in
Materials and Methods.

dAverage Two experiments - Percentages calculated from formulas in
Materials and Methods.

50

.moonuaz pom mamflnaumz
cH poucamonm mMHSEHOM Eonm woumasoamo mommucoonam I monEMm oumowamflnu mo omnnabgo

.H manna ca omnflnomap mm ucaspmana

A

.mHflmuwmv HON mwonumz USN mHMflkuwz wwmm

 

 

o a me
ov v cmchnmm
m o amm Hooz :onz
he mm me
am HH omnsHm
cm mH «mm Hooz :onz
mH eH me
o o
mH oH amm cmmHom wHonz
mom¢H> w < name w < onHm4H> we oxoamo my nezmzecmme meHmo
onmmm>zoo onmmm>zoo m>HeHmom HIwme n m>HeHmom HIwme w

 

MOBU¢m UHENEB mBHS Qm8¢mm9 92¢
mZEDAOU A003 ZOAMZ Qz< mmflAU Mm OHSflZOHBOdmh
WMBMUOZEAmm EMU ZH mAAmU m>HBHmOm N.HINmB m0 ZOHBUMBWO
H> mqm<fi

51

was found in the nylon wool nonadherent fraction (nearly 30%). This
population was discovered over a background of Thy-l positive cells
(approximately 20%) that normally reside in the spleen and which were
also nonadherent to nylon wool. The cells which were retained by the
column were neither Thy-l positive before nor after TF treatment. As
with bone marrow cells separated by the columns, inducible splenocytes
showed the characteristic increase in detection of Thy-l positive cells
when the more sensitive counting procedure was used. This finding, in
conjunction with the results shown in Table II, suggest that, instead
of increasing fragility, BSA fractionation may stabilize splenocytes in
some cases.

Time Course of Increased Cell Lysis

 

The suggestion that prolonged incubation times for the in vitra
conversion to Thy-1 positive cells may cause increased cell fragility
(178) was also examined. Bone marrow cells separated by glass and
nylon wool columns were incubated with TF for various times before being
tested for the presence of Thy-1. It was found that after one hour of
incubation the percentages of cells found to be Thy-l positive were
essentially the same regardless of the method used to calculate them
(Figure 2). Past one hour, an increase in the number of Thy-l positive
cells was not detectable unless calculations were being made on viable
cell counts. It is possible, therefore, that the length of the induc-
tion assay may have an effect on the number of Thy-l positive cells
which can be quantitated by the traditional method of counting dead

cells in a trypan blue exclusion cytotoxicity test.

52

FIGURE II

Time course of increased cell lysis during the induction of Thy-J.
on glass and nylon wool separated bone marrow cells. Nylon wool elutexi
and retained fractions were treated with TF for various lengths of tinm:
and tested for the presence of Thy-l. Percentages are based on foruuxlas
presented in Materials and Methods. Open symbols (0, A) refer to percent
Thy-l positive cells by change in number viable. Closed symbols (0, A)
refer to percent Thy—l positive cells by change in percent dead. Circlefis
(O, 0) refer to nylon wool eluted fractions. Triangles (A, A ) refer to

nylon wool retained fractions.

53

4
I

 

 

I
I

   

Time ( hours)

 

 

 

a
(If)

0
O
<3 .39
(\l
00/5194 003 103.9190!

.0
O

54

DISCUSSION

A rapid, precise, and reproducible assay for thymic hormones would
greatly aid the study of T cell differentiation. The Komuro and Boyse
Thy-l conversion assay (48), which measures the appearance of T cell
specific antigens on precursor T cells, potentially is such an assay
but, practically, often falls short of the description. Some of the
inherent difficulties of the system have been examined in this report.
These include the sensitivity and reproducibility of the cytotoxicity
test used to detect the expressed antigen, the heterogeneity in the
agents used to induce antigen expression, and the small number of pre-
cursor cells present in the highly heterogeneous hemopoietic organs and
the methods used to isolate them.

Others have attempted to improve the sensitivity of the cytotoxi-
city test. Twomney et al. (151), for example, added a proteolytic en-
zyme to the final incubation mixture in an attempt to detect Thy-1 posi-
tive cells which had suffered only a small degree of membrane damage
from antibody and complement treatment. They still found it necessary,
however, to extend their TF incubation time considerably beyond two hours
to obtain consistent results. On the other hand, it has been shown here
that, instead of adding another variable to an already complex system, a
simple change in counting procedures could increase the sensitivity of
the test. In some instances, as much as a three-fold increase in the
detection of Thy-1 inducible cells could be demonstrated (Tables I and
III). It is probable that this increase is due to the fact that the
viability counting method accounts for all cells killed by anti-Thy-l
sera and complement regardless of the stage of cellular damage and dis-

integration. This suspected increase in the number of lysed cells has

55

not been previously documented in the Thy-1 conversion system, but
other investigators (178) have reported that inducing agents occa-
sionally lead to spurious results. Perhaps these inconsistent find-
ings have been due in part to the loss of some Thy-l positive cells to
total lysis, and viability counting would have improved consistency.
Attempts have been made to identify morphologically the cell popula-
tion which is lysed (results not shown). Light microscopy revealed few
differences. Wright stained smears of treated samples from the 21% and
23% layers of the BSA gradient appeared morphologically homogeneous even
though it is apparent that these cells are functionally heterogeneous.
The actual reason for the increase in the number of Thy-1 positive
cells susceptible to disintegration remains to be determined. One pos-
sibility is that in vitra handling of the cells may change their fragil-
ity. It has been suggested that BSA gradient separation of prothymocytes
might significantly alter their fragility rendering the cells sensitive
to anti-Thy-l sera even though they have only minute quantities of the
antigen on their surfaces (185). Two different methods of cell separa-
tion have been used here to examine this possibility, one of which
(column separation) uses only low levels of BSA. Virtually no differ-
ence was found between the two techniques in the ratio of bone marrow
cells susceptible to lysis (Tables IV and V). In the case of inducible
splenocytes. it was shown that BSA gradient separation may actually
stabilize the cells rather than increase their fragility (Tables II and
VI). It appears, therefore, that high levels of BSA do not alone ac-
count for altered fragility of precursor cells. The significance of the
difference between bone marrow and spleen inducible cells is unknown.

In theory, the in vitra Thy-l conversion assay tests for the same

56

precursor cell regardless of its location in the body. However, until
this cell is isolated, or at least separated into a more homogeneous pop-
ulation, it will be difficult to accurately assess any other than gross
morphological or functional differences between candidate prothymocytes
of various sources. Perhaps precursors of different subpopulations of

T cells exist in different organs.

Another in vitra condition which may affect cell fragility is the
actual length of the induction assay itself. The length of incubation
time with various inducing agents has been suggested by Bach (178) to
affect the ability of a given method to detect antigen expression. Bach
felt that his rosette-forming assay may have been more reliable than the
Thy-1 induction assay partially due to the shorter time span needed to
quantitate the antigen. It has been shown here that the prothymocyte's
susceptibility to total lysis does appear to increase under lengthened
assay conditions (Figure 2). Therefore, if standard counting methods
were employed, the reliability of the in vitra conversion assay would
drop after a sixty minute incubation period. Viability counting has
been shown here to restore lost sensitivity. On the other hand, some
feel that more consistent results may be attained with longer (i.e. 18
hour) rather than shorter incubations (151). It has been suggested that
stem cells are present in the bone marrow and spleen in various trans-
itional states correlating with their stage in the cell cycle (115).

If this were true, additional precursors would continually enter the
inducible cell cycle stage during a prolonged assay. The experiments
documented in this report have not been extended past two hours. This
was done to avoid further complications brought on by long incubations

since these experiments were originally designed to study only the

57

initial consequences of TF treatment. Therefore, it is not known
whether the total number of Thy-1 converted cells plateaued at 20—30%
of the low density or nonadherent fractions, or whether it was actually
higher.

Finally, increased disintegration of Thy-l positive bone marrow
cells may not be due to increased cell fragility but rather to the
abundant amount of Thy-l antigen which is induced on these cells. TF
treated bone marrow cells (in an 18 hour assay) have been shown to have
6-12 times more Thy-l expressed on their surfaces than the average
amount of Thy-l on normal thymocytes (42). A highly cytotoxic anti-Thy-l
sera such as the ones used here, therefore, would be capable of sensitiz-
ing a cell sufficiently to cause extensive damage upon the addition of
complement.

Other factors may influence the reproducibility of the Thy—1 induc-
tion assay. The presence of contaminating endotoxin in the BSA gradient
used for cell separation is another possible reason for inconsistent re-
sults. Endotoxin has been shown to be a good inducer of Thy-l antigen
expression (54). Its presence in a BSA gradient could artifically raise
the background level of Thy-l positive cells abrogating any chance of
monitoring low levels of conversion. BSA stock solutions used were
screened for the presence of endotoxin by the highly sensitive Limulus
lysate test (179). BSA solutions which were negative were employed to
construct the BSA gradients for cell separation or as controls for TF
and butyrate induction.

Biological reagents such as antiserum and thymic factors may in-
fluence the consistency of results obtained with the in vitra conversion

assay. Though prepared similarly, different lots of antiserum will have

58

different titers and affinities. Therefore, they may behave quite dif-
ferently from one assay system to another as well as behaving different-
ly within one assay system (Letarte--personal communication, 186,187).
In the various lots of antiserum tested here, all of which have been
prepared by the same method, few differences have been observed (results
not shown).

Finally, an even greater threat of variation exists in the use of
thymic hormones. The wide variety of isolation procedures currently in
use lead to the purification of factors of various sizes and properties
(71,74,76). Crude fractions also harbor a large spectrum of contami-
nants which may affect the results of the assay.

Many substances induce the expression of Thy-1 antigen on precursor
T cells in addition to thymic factors. These include cAMP, dibutyrl-

cAMP (DB-cAMP), Poly A:U, endotoxin (54), prostaglandin E (57), ubiqui-

2
tin (65), neuraminidase (63), and thymus RNA (188). The common denom—
inator which most of these agents appear to possess is the ability to
raise intracellular levels of cAMP. DB-cAMP was shown to be effective

at lower levels than cAMP itself presumably due to the cell's increased
permeability to the organic acid linked cyclic nucleotide (54). However,
Wright (83), in an examination of the morphological and growth rate
changes induced by DB-cAMP in Chinese hamster ovary cells, found that
his control, sodium butyrate alone, was also effective in eliciting
change. Subsequently, others have shown that sodium butyrate produces
reversible changes in morphology, growth rate, and enzyme activities of
several mammalian cell types in culture (90). Some changes mimic the

action of DB-cAMP, while others are unique to the fatty acid. Since the

induction of Thy-l expression it thought in some manner to be mediated

59

by cAMP, sodium butyrate was employed as a putative inducing agent in
the in vitra conversion assay. The results described here indicate that
butyrate at low concentrations does mimic the action of cAMP and TF.
Butyrate treated bone marrow and spleen cells of the low density frac-
tions of BSA gradients express Thy—l antigen in approximately the same
percentage of cells as those induced by TF treatment (Table IV). In
addition, anti-Thy-l sera can be absorbed with butyrate treated marrow
cells to the same degree as with TF treated marrow cells (Figure 1).

Other correlations can be made between butyrate induction of Thy-l
and butyrate elicited changes in certain mammalian cell culture systems.
These include induction of differentiation in another hemopoietic sys-
tem, induced morphological alterations, changes in ganglioside levels,
and inhibition of induction by certain transcriptional and translational
inhibitors.

It has been shown that butyric acid is a potent inducer of erythro—
poiesis in cultures erythroleukemic cells (102). Erythroid differentia-
tion, as measured by the percentage of cells containing hemoglobin be-
fore and after treatment, was induced by 1 EM butyrate, a concentration
which is comparable to the one used to induce Thy-l expression in the
experiments reported here. These investigators have postulated that
butyrate may act directly on the cell membrane in light of its lipo-
philic nature. Others have actually demonstrated morphological altera-
tions of the cell membrane in HeLa cell cultures upon treatment with
millimolar concentrations of butyrate (93,97). Normally round, or poly-
gonal HeLa cells extended long neurite-like processes and assumed a more
fibroblastoid shape when incubated in the presence of 2.5-5.0 EM buty—

rate. Similarly, it has been postulated that TF induced Thy-1

60
expression may be the result of a membrane rearrangement (63).
Ganglioside levels have also been shown to change in response to
butyrate treatment (92,96). In HeLa cells, the levels of the glyco-

sphingolipid GM increase three to five fold, while the levels of the

3
other major gangliosides remain constant (92). This was due to the in-
duction of a specific sialyltransferase, the activity of which was raised
twenty-fold (94). It has been proposed that the antigenic determinant of
Thy-1 is carried by a ganglioside (27). It is possible, therefore, that
an alternate mechanism of butyrate (and TF) induction is to raise the
levels of a particular ganglioside, in this case Thy-l glycolipid or

one of its precursors, and that this increase leads to the expression

of Thy-l on the cell's surface. An attempt has been made to document a
similar ganglioside increase in the in vitra conversion assay using ra-
dioactively labeled galactose and glucosamine to monitor newly synthe-
sized ganglioside (results not shown). However, equivocal results have
been obtained primarily due to the small number of cells available and
to the small percentage of stem cells found even in enriched fractions
from BSA gradients or nylon wool columns. A more homogeneous popula-
tion of cells will be needed before changes in ganglioside levels can

be detected.

Further evidence that TF and butyrate may act in a similar fashion
to induce Thy-1 is the fact that both TF and butyrate induction of dif-
ferentiation have been found to be sensitive to treatment with inhibi-
tors of transcription and translation. Actinomycin D and cycloheximide
were shown to inhibit the appearance of the alloantigens TL and Thy-l
which were normally expressed upon TF treatment (64). Butyrate induced

morphological alterations and increased in ganglioside levels were

61

inhibited by similar levels of these two inhibitors in HeLa cell cul-
tures (93,189).

In most systems where butyrate has been shown to induce some type
of differentiation, the effect has been found to be quite specific for
short, unmodified, straight chain fatty acids. HeLa cell modifications
occurred only with C3, C4, or C5 saturated fatty acids (30,93), and
erythroid differentiation was found almost exclusively with butyrate
(102). Further specificity was demonstrated by the inability of iso-
butyrate to elicit characteristic changes (83,100,102). Possible addi-
tional fatty acid inducers in the in vitra conversion system have not
been extensively examined. Acetate (2 mM) was utilized as a control in
some experiments and was found to occasionally induce low levels of
Thy-l expression. Further experimentation is necessary to clarify
these results.

The molecular mode of action of butyrate or TF is unknown. Cur-
rently, it is thought that TF induction is mediated by the second mes-
senger cAMP. This is also thought to be true of some, but not of all,
cases of butyrate induction. In addition, butyrate treatment has been
shown to cause a decrease in DNA synthesis and a rapid acetylation of
specific histones in both HeLa (100) and erythroleukemic (104) cell
lines. It has also been postulated that prolonged butyrate treatment
may synchronize cell cultures in the S (100) or G1 (93) phase of the
cell cycle. Which, if any, of these documented effects of butyrate
treatment induces the expression of Thy-l on a precursor cell remains
to be determined.

While butyrate is not the physiological inducer of Thy-l expres-

sion on precursor cells migrating to the thymus, there are advantages

62

to studying this system with butyrate rather than TF as an inducing
agent. It is chemically defined and easily obtainable in large quan-
tities. Though the chemical definition of various thymic factors is
now known (71,74,76), their availibility is still limited. Further-
more, butyrate has been shown to exert its effects on defined systems
such as on differentiation of erythroleukemic cell lines. It may be
possible to draw preliminary parallels between such systems, where cell
quantities and homogeneity do not pose additional problems, and Thy-l
induction.

A third area to consider when examining the reproducibility of
the Thy-l induction assay is the target cell of the inducing agent, the
precursor T cell. Identification of this cell in bone marrow or spleen
has not been made due to inadequate separation of the cell from highly
heterogeneous populations and its presence in such low quantities. To
obtain a larger, more homogeneous population, one of two approaches
can be taken. The first would be to obtain a cleaner separation of
precursor T cells than is possible with currently used procedures. A
second approach would be to artificially raise the number of prothymo-
cytes present prior to isolation.

It has been attempted here to improve precursor T cell separation
with the use of a parameter other than density to fractionate the cell
population. Adherence to glass and nylon wool is a functional rather
than a physical property which has been found effective in the enrich-
ment of both B and T lymphocytes from spleen (160). The results report-
ed here show that adherence to columns can also be useful in the separ-
ation of prothymocytes from bone marrow and spleen (Tables V and VI).

Precursor T cells are found in the nylon wool nonadherent fraction as

63

are mature T cells. Others have documented similar findings (133).
Still this procedure yielded a population of approximately 20% Thy-l
inducible cells, the same percentage obtained with density centrifuga-
tion. Preliminary experiments with bone marrow suspensions, however,
have shown that, if two separation techniques are combined, with den-
sity centrifugation followed by incubation of the 21% and 23% layers

on nylon wool, approximately twice the concentration of precursors are
found in the nylon wool eluted fraction (results not shown). It appears,
therefore, that separation on the basis of two distinct parameters may
increase the percentage of precursor cells found in a particular popu-
lation. Investigators in myelopoiesis have successfully explored such
an approach using two physical parameters (density and volume) for sep-
aration (69,156).

It has been proposed that stem cells are viable replicating cells
which exist in transitional states of various sizes (115). This would
make separation by any one physical parameter very difficult. Moore
et al. (130), using fetal liver and adult bone marrow cells separated
by BSA gradients, obtained an even distribution of in viva colony form-
ing units in each of five different fractions. This finding suggests
that the pluripotential stem cell may be found in various transitional
states. Preliminary evidence suggests that this may also hold true for
precursor T cells. When the normally noninducible cells of the 25%
layer of a BSA gradient were further separated on a nylon wool column,
it was found that ten percent of the cells in the nylon wool eluted
fractions were induced to express Thy-l (results not shown). Perhaps
it will be necessary to find another more specific parameter in order

to separate all progenitor T cells from a given population.

64

The number of stem cells committed to a certain line of hemopoiesis
is manipulable by artificial means. For example, a simulated altitude
increase produced hypoxia in mice which subsequently increased the de-
mand for erythrocytes and the level of erythropoietin sensitive cells
(190). An analogous situation exists in lymphopoiesis which may be use-
ful in raising the numbers of precursor T cells in bone marrow or spleen
prior to their isolation. It has been found that bone marrow from AKR
mice with spontaneous thymomas have a 10-15 fold increase in the num-
ber of cells which equilibrate with the low density (lo-23%) fraction
of BSA density gradients (143). These cells have been shown to contain
terminal deoxynucleotidyl transferase (TdT), an enzyme which has been
found mainly in the thymus and bone marrow (141) and which has been pro-
posed as a marker for precursor T cells (139,143). Treatment with thy-
mopoietin made this increased population of TdT positive cells sensitive
to anti-Thy-l sera and complement (139), suggesting that they are pro-
thymocytes. It appears that leukemic mice may be a source of larger
quantities of precursor T cells than normal mice and that they might
be useful for large scale prothymocyte isolation.

Most biological assay systems are subject to a certain degree of
variation due to the large number of known and unknown parameters in-
volved. The Thy-l conversion assay is no exception. However, it has
been shown here that this assay can be reproducible if a more sensitive
method is used to quantitate the numbers of cells included to express
Thy-l than has been employed previously. If a more homogeneous popu—
lation of precursor cells can be obtained either by manipulation or by
improved separation and a chemically defined inducing agent such as
butyrate employed, the Thy-l conversion assay may be quite useful in

elucidating some of the steps involved in cellular differentiation.

 

10.

11.

12.

13.

14.

BIBLIOGRAPHY

Reif, A.E. and Allen, J.M.V. 1964. The AKR thymic antigen and its

distribution in leukemias and nervous tissues.

413-433.

Blankenhorn, E.P.

Douglas, T.C. 1972.

J. Exp. Med. 136:

and Douglas, T.C. 1972.

1054-1062.

J. Exp. Med. 120:

Location of the gene
for theta antigen in the mouse. J. Hered. 635259-263.

Reif, A.E. and Allen, J.M.V. 1966. Mouse thymic iso-antigens.
Nature 209:521-523.

Stern, P.L. 1973.

Nature (New Biol.

) 246:76-78.

G-alloantigen on mouse and rat fibroblasts.

Occurrence of a theta-like antigen in rats.

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

Gillette, R.W. 1977.
neoplastic mammary cells of mice.

1633.

Raff, M.C. 1969.

J.

Expression of Thy-1 antigen in normal and
J. Natl. Cancer Inst. 5831629-

Theta isoantigen as a marker of thymus derived
lymphocytes in mice. Nature 224:378-379.

Aoki, T., Hammerling, 0., de Haven, E., Boyse, E.A. and Old, L.J.

1969. Antigenic structure of cell surfaces.

979-1001.

Acton, R.T., Morris, R.J. and Williams, A.F.
the amount and tissue distribution of rat Thy-1.1 antigen.

Immunol. 2:598-602.

Williams, A.F. 1976.
face Thy-l antigen.

Hunt, S.V., Mason, D.W. and Williams, A.F.
marrow Thy-1 antigen is present on cells with membrane immunoglob-
ulin and on precursors of peripheral B lymphocytes. Eur. J. Im-

munol. 23817-823.

J. Exp. Med. 130:

1974. Estimation of

1977. In rat bone

Eur. J.

Many cells in rat bone marrow have cell sur-
Eur. J. Immunol. 65526-528.

Atwell, J.L., Cone, R.E. and Marchalonis, J.J. 1973. Isolation of
O-antigen from the surface of thymus lymphocytes. Nature (New

241:251-252.

Letarte-Muirhead, M., Barclay, A.N. and Williams, A.F. 1975.
fication of the Thy-1 molecule, a major cell surface glycoprotein of

rat thymocytes.

Biochem. J. 151:685-697.

65

Biol.)

Puri-

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

66

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 256:654.

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

Vitteta, E.S., Boyse, E.A. and Uhr, J.W. 1973. Isolation and char-
acterization of a molecular complex containing Thy-l antigen from
the surface of murine thymocytes and T cells. Eur. J. Immunol. 3;
446-453.

Letarte-Muirhead, M., Acton, R.T. and Williams, A.F. 1974. Prelim-
inary characterization of Thy—1.1 and Ag-B antigens from rat tissues
solubilized in detergents. Biochem. J. 143:51-61.

Kucich, U.N., Bennett, J.C. and Johnson, B.J. 1975. The protein
nature of the Thy-1.2 alloantigen as expressed by the murine lym-
phoblastoid line S-49.l TB.2.3. J. Immunol. 115:626-630.

Esselman, W.J. and Miller, H.C. 1974. Brain and thymus lipid inhi—
bition of anti-brain associated 8-- cytotoxicity. J. Exp. Med. 139:
445-450.

Trowbridge, 1.8. and Hyman, R. 1975. Thy—l variants of mouse lym-
phomas:Biochemical characterization of the genetic defect. Cell 6;
279-287.

Barclay, A.N., Letarte-Muirhead, M. and Williams, A.F. 1975. Puri-
fication of the Thy-l molecule from rat brain. Biochem. J. 151:
699-706.

Johnson, B.J., Kucich, U.N. and Maurelli, A.T. 1976. Studies on
the Thy-1.2 alloantigen as expressed by the murine lymphoblastoid
line S-49.1 TB.2.3. J. Immunol. 116:1669-1672.

Barclay, A.N., Letarte-Muirhead, M., Williams, A.F. and Faulkes, R.A.
1976. Chemical characterization of the Thy-l glycoproteins from the
membranes of rat thymocytes and brain. Nature 263:563-567.

Esselman, W.J. and Kato, K. 1976. Studies on the antigenic nature
of murine Thy-1 differentiation antigen. Fed. Proc. (abstr.) 35:
1643.

Stein-Douglas, R.E., Schwarting, G.A., Naiki, M. and Marcus, D.M.
1976. Gangliosides as markers for murine lymphocyte subpopulations.
J. Exp. Med. 143:822-832.

Wang, T.J., Freimuth, W.W., Miller, H.C. and Esselman, W.J. 1978.
Thy-l antigenicity is associated with glycolipids of brain and thy-
mocytes. J. Immunol. (in press).

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

67

Arndt, R., Stark, R., Klein, P., Muller, A. and Thiele, H.G. 1976.
Solubilization and molecular characterization of membrane bound an-
tigens shared by thymocytes and brain. Eur. J. Immunol. §;333-340.

Thiele, H.G., Arndt, R. and Stark, R. 1977. Evidence for the pre-
sence of choleragen receptor on the thymocyte-brain antigen molecule
of mice. Immunol. 32:767-770.

Fishman, P.H. and Brady, R.O. 1976. Biosynthesis and function of
gangliosides - gangliosides appear to participate in the transmis-
sion of membrane mediated information. Science 194:906-915.

Vitteta, E.S., Uhr, J.W. and Boyse, E.A. 1974. Metabolism of H-2
and Thy-l (6) alloantigens in murine thymocytes. Eur. J. Immunol.
43276-282.

Freimuth, W.W., Esselman, W.J. and Miller, H.C. 1978. Release of
Thy-1.2 and Thy-1.1 from lymphoblastoid cells: Partial characteriza-
tion and antigenicity of shed material. J. Immunol. 120:1651-1658.

Miller, M.F.A.P. and Osoba, D. 1963. Role of the thymus in the
origin of immunological competence. In_Nature and Origin of Immun-
ologically Competent Cells, Ciba Foundation Study Group. No. 16
(Ed. Wolstenholme and Knight), Churchill, London, p. 62.

Osoba, D. 1965. The effects of thymus and other lymphoid organs
enclosed in Millipore chambers on neonatally thymectomized mice.
J. Exp. Med. 122:633-650.

Stutman, O., Yunis, B.J. and Good, R.A. 1969. Carcinogen induced
tumors of the thymus. IV:Humoral influences of normal thymus and
functional thymomas and influence of post thymectomy period of re-
storation. J. Exp. Med. £30:809-819.

Hand, T.L., Geglowski, W.S., Damrongsal, D. and Freidman, H. 1970.
Development of antibody forming cells in neonatal mice: Stimula-
tion and inhibition of calf thymus extracts. J. Immunol. 105:442-
450.

Small, M. and Trainin, N. 1967. Increase in antibody forming cells
of neonatally thymectomized mice receiving calf thymus extract.
Nature 216:377-379.

Trainin, N. and Linker-Israli, M. 1967. Restoration of immunologic
reactivity of thymectomized mice by calf thymus extracts. Cancer
Res. 21;309-313.

Rotter, V., Globerson, A., Nakamura, I. and Trainin, N. 1973.
Studies on the characterization of the lymphoid target cell for
activity of a thymus humoral factor. J. Exp. Med. 138:130-142.

Papiernik, M., Nabarra, B. and Bach, J.F. 1975. In vitra culture
of functional human thymic epithelium. Clin. Exp. Immunol. 12:281-287.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

68

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

Basch, R.S. and Goldstein, G. 1974. Induction of T-cell differen-
tiation in vitra by thymin, a purified polypeptide hormone of the
thymus. Proc. Natl. Acad. Sci. (USA) 3331474-1478.

Globerson, A.L., Slater, F.D. and White, A. 1966. Preparation,
assay, and partial purification of a thymic lymphopoietic factor
(thymosin). Proc. Natl. Acad. Sci. (USA) 3331010-1017.

Bach, J.F., Dardenne, M., Goldstein, A., Guha, A. and White, A. 1971.
Appearance of T cell markers in bone marrow after incubation with pur-
ified thymosin, a thymic hormone. Proc. Natl. Acad. Sci. (USA) 333
2734-2738.

Bach, J.F. and Dardenne, M. 1972. Thymus dependency of rosette
forming cells, evidence for circulating thymic hormone. Trans.
Proc. 33345-356.

Bach, J.F. and Dardenne, M. 1973. Studies on thymus products. II.
Demonstration and characterization of a circulating thymic hormone.
Immunol. 333353.

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 331056-1058.

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

Komuro, K. and Boyse, E.A. 1973. In vitra demonstration of thymic
hormone in the mouse by conversion of precursor cells into lympho-
cytes. Lancet 33740-743.

Boyse, E.A. and Old, L.J. 1969. Some aspects of normal and abnormal
cell surface genetics. Ann. Rev. Genet. 33269-290.

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 333239-243.

Stockert, E., Old, L.J. and Boyse, E.A. 1971. The G X system: A cell
surface alloantigen associated with murine leukemia Virus; implica-
tions regarding chromosomal integration of the viral genome. J. Exp.
Med. 33331334-1355.

Shigeno, N., Arpels, C. Hammerling, V., Boyse, E.A. and Old, L.J.
1968. Preparation of lymphocyte-specifying antibody from anti-lymph-
ocyte serum. Lancet 33320-323.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

69

Scheid, M.P., Hoffman, M.K., Komuro, K., Hammerling, V., 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. 33331027-1032.

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.

Kook, A.I. and Trainin, N. 1975. Intracellular events involved in
the induction of immune competence in lymphoid cells by a thymus
humoral factor. J. Immunol. 115:151-157.

Bach, M.A. and Bach, J.F. 1973. Studies on thymus products VI:

The effect of cyclic nucleotides and prostaglandins on rosette-form-
ing cells; interactions with thymic factor. Eur. J. Immunol. 33
778-783.

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

Bach, J.F. 1976. The mode of action of thymic hormones and its
relevance to T cell differentiation. Trans. Proc. 33243-248.

Dardenne, M. and Bach, J.F. 1973. Modification of rosette forming
cells by thymic extracts. Determination of the target RFC subpop-
ulation. Immunol. 333343-352.

Bach, M.A. 1975. Disparition transitoire de l'antigene theta sous
l'influence de l'AMP cyclique et de l'Indomethaine. Annls. Immunol.
126C:84.

Boyse, E.A. and Abbott, J. 1975. Surface reorganization as an ini-
tial inductive event in the differentiation of prothymocytes to thy-
mocytes. Fed. Proc. 32324-27.

Milewicz, C., Miller, H.C. and Esselman, W.J. 1976. Membrane ex-
pression of Thy-1.2 and G ganglioside on differentiating T lympho-
cytes. J. Immunol. 117:19 4-1780.

Storrie, B., Goldstein, G., Boyse, E.A. and Hammerling, B. 1976.
Differentiation of thymocytes: Evidence that induction of the sur-
face phenotype requires transcription and translation. J. Immunol.
33331358-1362.

Goldstein, G., Scheid, M., Hammerling, V., Boyse, E.A., Schlesinger,
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) 13311-15.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

7O

Scheid, M.P., Goldstein, G., Hammerling, V. and Boyse, E.A. 1975
Lymphocyte differentiation from precursor cells in vitra. Ann. N.Y.
Acad. Sci. 249:531-538.

Hammerling, V., Chin, A., Abbott, J. and Scheid, M.P. 1975. The
ontogeny of murine B lymphocytes. I. Induction of phenotypic conver-
sion of Ia- to Ia+ lymphocytes. J. Immunol. 115:1425-1431.

Kagan, W.A., Incefy, G.S., Gupta, S., Segal, F.P., Goldstein, G. and
Good, R.A. 1976. Induction of surface receptors on granulocytes, T,
and B cell precursors in human bone marrow by thymic and nonthymic
agants. 32 Leucocyte membrane determinants regulating immune reacti-
vity. N.Y. Acad. Press (Eds. Ejovoogel, V.P. et al.) p. 719.

Kagan, W.A., O'Neill, G.J., Incefy, G.S. Goldstein, G. and Good, R.A.
1977. Induction of human granulocyte differentiation in vitra by
ubiquitin and thymopoietin. Blood 333275-287.

Dardenne, M., Pleau, J.M., Man, N.K. and Bach, J.F. 1977. Struc-
tural study of a circulating thymic factor: A peptide isolated from
pig serum. I. Isolation and purification. J. Biol. Chem. 25238040-
8044.

Pleau, J.M., Dardenne, M., Bloquit, Y. and Bach, J.F. 1977. Struc-
tural study of a circulating thymic factor: A peptide isolated from
pig serum. II. Amino acid sequence. J. Biol. Chem. 252:8045-8047.

Bricas, E., Martinez, J., Blanst, D., Auger, M., Dardenne, M., Pleau,
J.M. and Bach, J.F. 1977. Proc of the 5th Amer. Peptide Symp. June,
1977. San Diego, La Jolla, CA.

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

Schlesinger, D.H. and Goldstein, G. 1975. The amino acid sequence
of thymopoietin II. Cell 33365-370.

Schlesinger, D.H., Goldstein, G., Scheid, M.P. and Boyse, E.A. 1975.
Chemical synthesis of a peptide fragment of thymopoietin II that in-
duces selective T cell differentiation. Cell 33365-370.

Goldstein, A., Low, T.L., McAdoo, M., McClure, J., Thurman, G.B.,
Rossio, J., Lai, C-Y., Chaing, D., Wang, S-S., Harvey, G., Ramel, A.E.
and Meinhofer, J. 1977. Thymosin a : Isolation and sequence anal-
ysis of an immunologically active thymic polypeptide. Proc. Natl.
Acad. Sci. (USA) 233725-729.

Dardenne, M. and Bach, J.F. 1974. The sheep cell rosette assay for
the evaluation of thymic hormones. 33_The Biological Activity of
Thymic Hormones. (Ed. van Bekkum), Kooyker Scientific Publish.,
Rotterdam. p. 235-243.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

71

Shew, C.W. and Freese, E. 1972. Effects of fatty acids on growth
and envelope proteins of Bacillus subtilis. J. Bac. 111:516-524.

 

Shew, C.W., Konings, W.N. and Freese, E. 1972. Effects of acetate
and other short-chain fatty acids on sugar and amino acid uptake of
Bacillus subtilis. J. Bac. 111:525-530.

 

Freese, E., Shew, C.W. and Galliers, E. 1973. Function of lipophil-
lic acids as antimicrobial food additives. Nature 241:321-325.

Ginsburg, E., Salomon, D., Sreevalsan, T. and Freese, E. 1973.
Growth inhibition and morphological changes caused by lipophillic
acids in mammalian cells. Proc. Natl. Acad. Sci. (USA) 1332457-2461.

Pace, D.M., Aftonomas, B.T., Elliot, A. and Sommer, S. 1967. Obser-
vations on some effects of the sodium salts of certain monocarboxy-
lic acids on established cell lines. Can. J. Biochem. 53381-88.

Wright, J.A. 1973. Morphological and growth rate changes in Chinese
hamster ovary cells cultures in presence of sodium butyrate. Exptl.
Cell Res. 233456-460.

Johnson, G.S., Friedman, R.M. and Pastan, I. 1971. Restoration of
several morphological characteristics of normal fibroblasts in sar-
coma cells treated with adenosine 3'5'-cyclic monophosphate and its
derivatives. Proc. Natl. Acad. Sci. (USA) 333425-429.

Hsie, A.W. and Puck, T.T. 1971. Morphological transformation of
Chinese hamster ovary cells by dibutyrl adenosine 3'5'-monophosphate
and testosterone. Proc. Natl. Acad. Sci. (USA) 333358-361.

Johnson, G.S. and Pastan, I. 1971. Change in growth and morphology
of fibroblasts by prostaglandins. J. Natl. Cancer. Inst. 2131357-
1364.

Prasad, K.N. and Hsie, A.W. 1971. Morphological differentiation of
mouse neuroblastoma cells induced in vitra by dibutyrl-cAMP. Nature
(New Biol) 233:141-142.

Prasad, K.N., Gilmer, K.N. and Kumer, S. 1973. Morphological "dif-
ferentiation" in mouse neuroblastoma cells induced by noncyclic AMP
agents: Levels of cAMP, nucleic acids and protein. Proc. Soc. Exp.
Biol. Med. 33331168-1171.

Sheppard, J.R. and Prasad, K.N. 1973. Cyclic AMP levels and the
morphological differentiation of mouse neuroblastoma cells. Life
Sci. Part II 333431-439.

Prasad, K.N. and Sinha, P.K. 1976. Effect of sodium butyrate on
mammalian cells in culture: a review. In vitra 333125-132.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

72

Tallman, J.F., Smith, C.C. and Henneberry, R.C. 1977. Induction
of functional B-adrenergic receptors in HeLa cells. Proc. Natl.
Acad. Sci. (USA) 353873-877.

Fishman, P.H., Simmons, J.L., Brady, R.O. and Freese, E. 1974.
Induction of glycolipid biosynthesis by sodium butyrate in HeLa
cells. Biochem. Biophys. Res. Comm. 333292-299.

Simmons, J.L., Fishman, P.H., Freese, E. and Brady, R.O. 1975.
Morphological alteration and ganglioside sialyltransferase activity
induced by small fatty acids in HeLa cells. J. Cell Biol. 333414-
424.

Fishman, P.H., Bradley, R.M. and Henneberry, R.C. 1976. Butyrate-
induced glycolipid biosynthesis in HeLa cells: Properties of the
induced sialyltransferase. Arch. Biochem. and Biophys. 172:618-626.

Moskal, J.R., Gardner, D.A. and Basu, S. 1974. Changes in glyco-
lipid glycosyltransferases and glutamate decarboxylase and their
relationship to differentiation in neuroblastoma cells. Biochem.
Biophys. Res. Comm. 333751-758.

Lockney, M., Moskal, J.R., Fung, Y.K. and Macher, E.A. 1978. The
effects of butyrate on glycosphingolipid metabolism in human epider-
moid (KB) cells. Fed. Proc. (abstr.) 3331766.

Henneberry, R.C., Fishman, P.H. and Freese, E. 1975. Morphologi-
cal changes in cultured mammalian cells: Prevention by the calcium
ionophore A23l87. Cell 331-9.

Prasad, K.N., Kumar, S., Gilmer, K.N. and Vernadakes, V.A. 1973.
Cyclic AMP induced differentiation in neuroblastoma cells. Changes
in total nucleic acid and protein contents. Biochem. Biophys. Res.
Comm. 393973-977.

Prasad, K.N., Bondy, S.C. and Purdy, J.L. 1975. Changes in poly A
containing cytoplasmic RNA in cAMP induced "differentiation" in neuro-
blastoma cells in culture. Exp. Cell Res. 33388-94.

Hagopian, H.K., Riggs, M.S., Swartz, L.A. and Ingram, V.M. 1977.
Effect of n-butyrate on DNA synthesis in chick fibroblasts and HeLa
cells. Cell 333855-860.

Chow, J.Y., Robinson, J.C. and Wang, S-S. 1977. Effects of sodium
butyrate on synthesis of human chorionic gonadotrophin in trophoblas-
tic and nontrophoblastic tumors. Nature 268:543-544.

Leder, A. and Leder, P. 1975. Butyric acid, a potent inducer of
erythroid differentiation in cultured erythroleukemic cells. Cell
5:319-322.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

73

Nudel, U., Salmon, J.E., Terada, M., Bank, A., Rifkind, R.A. and
Marko, P.A. 1977. Differentiation effects of chemical inducers
on expression of B globin genes in murine erythroleukemic cells.
Proc. Natl. Acad. Sci. (USA) 3331100-1104.

Riggs, M.G., Whittaker, R.G., Neumann, J.F. and Ingram, V.M. 1977.
N-butyrate causes histone modification in HeLa and Friend erythro-
leukemic cells. Nature 268:462-464.

Boffa, L.C., Vidali, G., Mann, R.S. and Allfrey, V.G. 1978. Sup-
pression of histone deacetylation in viva and in vitra by sodium
butyrate. J. Biol. Chem. 253:3365-3366.

Chatterjees, S., Sweeley, C.C. and Velicer, L.F. 1973. Biosynthe-
sis of proteins, nucleic acids and glycosphingolipids by synchron-
ized KB cells. Biochem. Biophys. Res. Comm. 333585-592.

Griffin, M.H., Price, G.H., Bozzell, K.L., Cox, R.P. and Ghosh, N.K.
1974. A study of adenosine 3',5' cyclic monophosate, sodium buty-
rate, and cortisol as inducers of HeLa alkaline phosphatase. Arch.
Biochem. Biophys. 3333619-623.

Till, J.E. and McCulloch, E.A. 1961. A direct measurement of the
radiation sensitivity of normal mouse bone marrow cells. Radiation
Res. 333213-222.

McCulloch, E.A. and Till, J.E. 1962. The sensitivity of cells from
normal mouse bone marrow to gamma radiation in vitra and in viva.
Radiation Res. 333822-832.

Fowler, J.H., WU, A.M., Till, J.E., McCulloch, E.A. and Siminovitch,
L. 1967. The cellular composition of hemopoietic spleen colonies.
J. Cell Physiol. 33365-72.

Wu, A.M., Till, J.E., Siminovitch, L. and McCulloch, E.A. 1967. A
cytological study of the capacity for differentiation of normal hemo-
poietic colony-forming cells. J. Cell Physiol. 333177-184.

Metcalf, D. and Moore, M.A.S. 1971. 33_Frontiers of Biology v. 24.
Haemopoietic Cells. North Holland Publish. Co., Amsterdam. p. 73.

Cudkowicz, G., Upton, A.C., Smith, L.H., Goslee, D.G. and Hughes,
W.L. 1964. An approach to the characterization of stem cells in
mouse bone marrow. Ann. N.Y. Acad. Sci. 114:571-585.

Moore, M.A.S., Williams, N. and Metcalf, D. 1972. Purification and
characterization of the in vitra colony forming cell in monkey hemo-
poietic tissue. J. Cell Physiol. 333283-292.

Yoffey, J.M. 1973. Stem cell role of the lymphocyte-transitional
(LT) cell compartment. 33_Hemopoietic Stem Cells, Ciba Foundation
Symp., Assoc. Scientific Publish., Amsterdam. p. 5-45.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

74

Van Bekkum, D.W., van Noord, M.J., Maat, B. and Dicke, R.A. 1971.

Attempts at identification of hemopoietic stem cells in mice. Blood.
J. Hematol. 333547-558.

Dicke, K.A., Platenburg, M.G.C. and van Bekkum, D.W. 1971. Colony
formation in agar: in vitra assay for haemopoietic stem cells. Cell
Tissue Kinet. 33463-477.

Dicke, K.A., van Noord, M.J., Maat, B., Schaeffer, V.W. and van
Bekkum, D.W. 1973. Attempts at morphological identification of the
haemopoietic stem cell in primates and rodents. 33_Hemopoietic Stem
Cells, Ciba Foundation Symp., Assoc. Scientific Publish., Amsterdam.
p. 47-69.

Worton, R.G., McCulloch, E.Q. and Till, J.E. 1969. Physical separ-
ation of hemopoietic stem cells from cells forming colonies in cul-
ture. J. Cell Physiol. 333171-182.

Auerbach, R. 1961. Experimental analysis of the origin of cell
types in the development of the mouse thymus. Devel. Biol. 33
336-354.

Moore, M.A.S. and Owen, J.J. 1967. Experimental studies on the
development of the thymus. J. Exp. Med. 126:715-726.

Owen, J.J. and Ritter, M.A. 1969. Tissue interaction in the devel-
opment of thymus lymphocytes. J. Exp. Med. 129:431-442.

Harvis, J.E., Ford, C.E., Barnes, D.W. and Evans, E.P. 1964. Cel-
lular traffic of the thymus: Experiments with chromosomal markers.
Nature 201:884.

Owen, J.J. and Raff, M.C. 1970. Studies on the differentiation of
thymus derived lymphocytes. J. Exp. Med. 132:1216-1232.

Mekori, L., Chieco-Bianci, L. and Feldman, M. 1965. Production of
clones of lymphoid cell populations. Nature 206:367-368.

Curry, J.L., Trentin, J.J. and Cheng, V. 1967. Hemopoietic spleen
colony study. III. Hemopoietic nature of spleen colonies induced by
lymph node or thymus cells, with or without phytohemagglutinin. J.
Immunol. 333907-916.

General discussion - Colonialism. 1973. 33_Haemopoietic Stem Cells.
Ciba Foundation Symp. v.13., Assoc. Scientific Publish., Amsterdam.
p. 328-330.

Wu, A.M., Till, J.E., Siminovitch, L. and McCulloch, E.A. 1968.
Cytological evidence for a relationship between normal haemopoietic
colony forming cells and cells of the lymphoid system. J. Exp. Med.
3333455-464.

129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

75

Turner, R.W., Siminovitch, L., McCulloch, E.A. and Till, J.E. 1967.
Density gradient centrifugation of hemopoietic colony-forming cells.
J. Cell Physiol. 33373-82.

Moore, M.A.S., Mdfleill, T.A. and Haskill, J.S. 1970. Density dis-
tribution analysis of in viva and in vitra colony forming cells in
developing fetal liver. J. Cell Physiol. 333181-192.

El-Arini, M.O. and Osoba, D. 1973. Differentiation of thymus-de-
rived cells from precursors in bone marrow. J. Exp. Med. 137:831-
837.

Incefy, G.S., L'Esperance, P. and Good, R.A. 1975. In vitra dif-
ferentiation of human marrow cells into T lymphocytes by thymic fac-
tors using the rosette technique. Clin. Exp. Immunol. 333475-483.

Basch, R.S. and Kadish, J.L. 1977. Hemopoietic thymocyte precursors.
II. Properties of the precursors. J. Exp. Med. 145:405-419.

Komuro, K., Goldstein, G. and Boyse, E.A. 1975. Thymus-repopulat-
ing capacity of cells that can be induced to differentiate to T
cells in vitra. J. Immunol. 115:195-198.

Basch, R.S. and Goldstein, G. 1975. Antigenic and functional evi-
dence for the in vitra inductive activity of TP (thymin) on thymo-
cyte precursors. Ann. N.Y. Acad. Sci. 249:290-299.

Golub, E.S. 1972. Brain associated stem cell antigen: An antigen
shared by brain and hemopoietic stem cells. J. Exp. Med. 136:369-
374.

Chess, L., Levine, H., MacDermott, R.P. and Scholossman, S.R. 1975.
Immunologic functions of isolated human lymphocyte subpopulations.
VI. Further characterization of the surface 19 negative, E-rosette
negative, (null cell) subset. J. Immunol. 33331483-1487.

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

Silverstone, A.E., Canton, A., Goldstein, G. and Baltimore, D.
1976. Terminal deoxyribonucleotidyl transferase is found in prothy-
mocytes. J. Exp. Med. 144:534-548.

Bollwan, F.J. 1974. Terminal deoxynucleotidyl transferase. 33_
The Enzymes, Vol. 10., (Ed. Boyer, P.D.), Academic Press Inc., N.Y.
p. 145.

Kung, P.C., Silverstone, A.E., McCaffrey, R.P. and Baltimore, D.
1974. Murine terminal deoxyribonucleotidyl transferase: Cellular
distribution and response to cortisone. J. Exp. Med. 141:855-865.

142.

143.

144.

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

76

Sugimoto, M., Chang, L.M.S. and Bollum, F.J. 1978. Terminal trans-
ferase in developing chick embryo cells. Fed. Proc. (abstr.) 3331305.

Pazimo, N.H., McEvan, R.N. and Ihle, J.N. 1977. Distribution of ter-
minal deoxyribonucleotidyl transferase in BSA gradient fractionated
thymocytes and bone marrow cells of normal and leukemic mice. J. Im-
munol. 3333494-499.

Shaw, M.T., Dwyer, J.M., Slaudeen, H.S. and Weitzman, H.A. 1978.
Terminal deoxyribonucleotidyl transferase activity in B-cell acute
lymphocytic leukemia. Blood 333181-187.

Bennett, M., Cudkowicz, G., Foster, R.S. and Metcalf, D. 1968. He-
mopoietic progenitor cells of 3 anemic mice studied in viva and in
vitra. J. Cell. Physiol. 33:211-226.

Loor, F. and Kindred, B. 1973. Differentiation of T cell precursors
in nude mice demonstrated by immunofluoresence of T cell membrane
markers. J. Exp. Med. 138:1044-1055.

Greenburg, R.S. and Katz, M.M. 1975. Spontaneous AKR lymphomas
with T and B cell characteristics. Nature 257:314-316.

Ly, T.A. and Mishell, R.I. 1974. Separation of mouse spleen cells
by passage through columns of Sephadex G-10. J. Immunol. Methods
5:239-247.

Davis, M.L., WOodard, D.A., Pesto, M. and Toya, R.E. 1977. Attempts
to purify hemopoietic stem cell enrichment in bone marrow by use of
glass wool filtration. Exp. Hematol. 33310-318.

Gambill, M.R., Ledney, G.D. and Macvittie, T.J. 1976. Mitigation
of graft-vs-host disease in lethally irradiated mice grafted with
spleen cells adherent to glass beads. Transplantation 333247-254.

Twomney, J.J., Goldstein, G., Lewis, V.M., Bealmear, P.M. and Good,
R.A. 1977. Bioassay determinations of thymopoietin and thymic hor-
mone levels in human plasma. Proc. Natl. Acad. Sci. (USA) 3332541-
2545.

Metcalf, D., Moore, M.A.S. and Shortman, D. 1971. Adherence col-
umn and bouyant density separation of bone marrow stem cells and
more differentiated cells. J. Cell Physiol. 333441-450.

Pretlow, T.G., Williams, E.E., Davis, M.L. and Zettergren, J.G.
1973. Separation of spleen colony forming units (CPU-S) from mouse
bone marrow cells using velocity sedimentation in an isokinetic
gradient of Ficoll in tissue culture medium. Amer. J. Path. 333
201—220.

Worton, R.G., McCulloch, E.A. and Till, J.E. 1969. Physical sep-
aration of hemopoietic stem cells differing in their capacity for
self-renewal. J. Exp. Med. 130:91-103.

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

77

Amato, D., Cowan, D.H. and McCulloch, E.A. 1972. Separation of
immunocompetent cells from human and mouse hemopoietic cell suspen-
sions by velocity sedimentation. Blood 333472-480.

Burghouts, J., Plas, A.M., Wessels, J., Hillen, H.C., Steenbergen, J.
and Haanen, C. 1978. Method for enrichment of proliferating myeloid
cells from normal and leukemic human bone marrow. Blood 3339-20.

Haskill, J.S. and Moore, M.A.S. 1970. Two dimensional cell sep-
aration: comparison of embryonic and adult hemopoietic stem cells.
Nature 226:853-854.

Wildy, P. and Ridley, M. 1958. Separation of human leucocytes from
blood. Nature 182:1801-1803.

Cudkowicz, G., Bennett, M. and Shearer, G.M. 1964. Pluripotent stem
cell function of the mouse marrow '1ymphocyte'. Science 144:866-868.

Trizio, D. and Cudkowicz, G. 1974. Separation of T and B lympho-
cytes by nylon wool columns: Evaluation of efficacy by functional
assays in viva. J. Immunol. 113:1093-1097.

Kadish, J.L. and Basch, R.S. 1976. Hemopoietic thymocyte precur-
sors. I. Assay and kinetics of the appearance of progeny. J. Exp.
Med. 143:1082-1099.

Metcalf, D. and Moore, M.A.S. 1971. 33_Frontiers of Biology v.24,
Haemopoietic Cells, North Holland Publish. Co., Amsterdam. p. 66.

Leif, R.C. and Vinograd, J. 1964. The distribution of bouyant
density of human erythrocytes in bovine serum albumin solutions.
Proc. Natl. Acad. Sci. (USA) 333520-528.

Pretlow, T.G., Weir, E.E. and Zettergren, J.G. 1975. Problems con-
nected with the separation of different kinds of cells. Int. Rev.
Exp. Path. 33391-204.

Loos, J.A. and Roos, D. 1976. Density Analysis as a tool for blood
cell separation. 33_Biological Separations in Iodinated Density Gra-
dient Media, (Ed. D. Rickwood), Information Retrieval Limited, London
and Washington, D.C. p. 97-105.

Seglen, P.O. 1976. The use of metrizamide for the separation of
rat liver cells. 33_Biological Separations in Iodinated Density
Gradient Media, (Ed. D. Rickwood), Information Retrieval Limited,
London and Washington, D.C. p. 107-121.

Byrne, P. and Heit, W. 1976. Separation of granulocyte-macrophage
progenitor cells by continuous density gradients. 33_Biological Sep-
arations in Iodinated Density Gradient Media. (Ed. D. Rickwood),
Information Retrieval Limited, London and Washington, D.C. p. 137-
143.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

179.

78

Miller, R.G. and Phillips, R.A. 1969. Separation of cells by vel-
ocity sedimentation. J. Cell Physiol. 333191-202.

Worton, R.G., McCulloch, E.A. and Till, J.E. 1969. Physical separ-
ation of hemopoietic stem cells differing in their capacity for self-
renewal. J. Exp. Med. 130:91-103.

Lafleur, L., Miller, R.G. and Phillips, R.A. 1973. Restriction of
specificity in the precursorSImfbone marrow-associated lymphocytes.
J. Exp. Med. 137:954-966.

McCool, D., Miller, R.J., Painter, R.H. and Bruce, W.R. 1970. Ery-
thropoietic sensitivity of rat bone marrow cells separated by velo-
city sedimentation. Cell Tiss. Kinet. 3355-65.

Paterson, N.A., Leid, R.S., Said, J.W., Wasserman, 8.1. and Austen,
K.F. 1976. Release of chemical mediators from dispersed and par-
tially purified human and rat lung mast cells. 33_Lungs in Disease,
(Ed. A. Bouhuys), Elsevier-North Holland Biomedical Press. p. 223—
238.

Richman, G.M., Chess, L. and Yankee, R.A. 1978. Purification and
characterization of granulocytic progenitor cells (CFU-C) from

human peripheral blood using immunologic surface markers. Blood
51:1-8.

Mayer, E.P., Chen, W—Y., Dray, S. and Teodorescu, M. 1978. The
identification of six mouse lymphocyte subpopulations by their
natural binding of bacteria. J. Immunol. 120:167-173.

Hayry, P., Anderson, L.C. and Nordling, S. 1973. Electrophoretic
fractionation of mouse T and B lymphocytes. Efficiency of the me-
thod and purity of the cells. Trans. Proc. 3387-90.

Brunette, D.M., McCulloch, E.A. and Till J.E. 1968. Fractionation
of mouse spleen cells by counter-current distribution. Cell Tiss.
Kinet. 33319-327.

Herzenberg, L.A., Sweet, R.G. and Herzenberg, L.A. 1976. Fluores-
ence-activated cell sorting. Scientific American 234:108-117.

Fournier, C. and Bach, J.F. 1975. Induction of G-positive cells
by thymic hormones. A critical evaluation. 33_Biological Activity
of Thymic Hormones, (Ed. D.W. van Bekkum), Kooyker Scientific Pub-
lications, Rotterdam, The Netherlands. p. 251-256.

Yin, B.T., Galanos, C., Kinsky, S., Bradshaw, R.A., Wessler, S.,
Luderitz, O. and Sarmiento, M.E. 1972. Picogram-sensitive assay
for endotoxin: Gelation of Limulus polyphemus blood cell lysate
induced by purified lipopolysaccharides and lipid A from Gram-neg-
ative bacteria. Biochim. Biophys. Acta 3333284-289.

 

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

190.

79

Shortman, K., Williams, N. and Adams, P. 1972. The separation of
different cell classes from lymphoid organs. V. Simple procedures
for the removal of cell debris, damaged cells, and erythroid cells
from lymphoid cell suspensions. J. Immunol. Meth. 33273-287.

Miller, H.C. and Esselman, W.J. 1975. Identification of 8-- bearing
T cells derived from bone marrow cells treated with thymic factor.
Ann. N.Y. Acad. Sci. 149:125-144.

Hooper, J.A., McDaniel, M.C., Thurman, G.B., Cohen, G.H., Schulof,
R.S. and Goldstein, A.L. 1975. Purification and properties of
bovine thymosin. Ann. N.Y. Acad. Sci. 149:125-144.

Boyse, E.A., Old, L.J. and Chouroulinkov, I. 1964. Cytotoxic test
for demonstration of mouse antibody. Meth. Med. Res. 33339-47.

Kierszenbaum, F. and Budzko, D.B. 1977. Cytotoxic effects of nor-
mal sera on lymphoid cells. Cell Immunol. 333137-146.

Bach, J.F. and Carnaud, C. 1976. Thymic factors. Prog. Allergy
333341-408.

Basch, R.C. 1974. Effects of antigen density and noncomplement
fixing antibody on cytolysis by alloantisera. J. Immunol. 113:554-
562.

Kamarck, M.E. and Gottlieb, P.D. 1977. Expression of mouse thymo-
cyte surface alloantigens in the fetal mouse thymus in viva and in
organ culture. J. Immunol. 119:407-415.

Archer, S.J. 1978. Induction of a T cell specific antigen on bone
marrow lymphocytes with thymus RNA. Immunol. 333123-129.

Henneberry, R.C. and Fishman, P.H. 1976. Morphological and biochem-
ical differentiation in HeLa cells. Effects of cycloheximide on bu-
tyrate-induced process formation and ganglioside metabolism. Exp.
Cell Res. 333:55-62.

Chevernick, P.A. and Boggs, D.R. 1968. Decreased neutrophils and
megakaryocytes in anemic mice of genotype W/W . J. Cell Physiol.
73:25-30.