Characterization of Abelson murine leukemia virus-transformed midgestation embryonic cells and their normal homologues


Material Information

Characterization of Abelson murine leukemia virus-transformed midgestation embryonic cells and their normal homologues
Physical Description:
ix, 220 leaves : ill. ; 29 cm.
Siegel, Michael L., 1949-
Publication Date:


Subjects / Keywords:
Leukemia Virus, Murine   ( mesh )
Muridae -- embryology   ( mesh )
Immunology and Medical Microbiology thesis Ph.D   ( mesh )
Dissertations, Academic -- Immunology and Medical Microbiology -- UF   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1986.
Includes bibliographical references (leaves 189-218).
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Michael L. Siegel.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 022096813
oclc - 16955357
System ID:

Full Text








This dissertation and my accomplishments over the past

several years are the result of a team effort. I wish to

acknowledge the contributions of all of my "teammates," and

hope that I slight none by inadvertent oversight.

First, I would like to thank my friend, mentor, and

supervisory chairperson, Edward J. Siden. His profound

powers of perception and his ability to assimilate new

observations into an existing body of knowledge have set

ideals which I shall take with me. His friendship and

camaraderie will always be of value to me.

Second, I wish to acknowledge and thank my supervisory

committee, J. Bert Flanegan, Carlo Moscovici, Steve Russell,

and Roy Weiner, who collectively guided me through my

dissertation research, sometimes despite my reluctance and


My family has been a constant source of support,

cheer, and inspiration. My wife Jeremie has coped with

every crisis I brought home and has survived more mood

fluctuations in the last few years than most people endure

in a lifetime. David and Rebecca have also endured the

journey with few complaints and good humor, and it is to

them I owe the preservation of my humor. I must also thank

my father, Seymour Siegel, who inspired me to strive to be

the best at whatever I tried (even if I was a garbage man),

and my mother, Frances Heifer Siegel, who showed me the

value of patience and determination. My brother, Victor,

is also acknowledged for encouraging me to read and learn.

I owe much of my recent success to the support of my

fellow graduate students. In particular, I express my

thanks to Randy Horwitz, who has helped me stay young,

sharpened my cynical wit, shared my most profane moments,

and provided me with friendship which has endured almost

six years in Gainesville.

Finally, I wish to recognize the faculty and staff

(past and present) of the Department of Immunology and

Medical Microbiology who have made my experience more

fulfilling. In particular, I wish to thank Ken Berns, who

encouraged me to return to graduate studies after a seven

year hiatus. I am also indebted to Catherine and Richard

Crandall, George Gifford, and Michael Boyle, for a

seemingly endless supply of reagents and advice. Last, but

not least, I thank Muriel Reddish, Patrice Boyd, Ellen

Boukari and their superb staffs, without whom the work

would probably have taken six additional years.










Introduction: Hematopoietic Cell Differentiation
and Tumor Models 1
Mast Cell Function, Origin, Ontogeny, and
Heterogeneity 6
Connective Tissue Mast Cells 18
Mucosal Mast Cells 19
Basophils 25
Mastocytomas 26
Culture-Derived Mast Cells 28
Other In Vitro-Derived Metachromatic Cells 53
Relationship of In Vivo- and In Vitro-Derived
Mast Cells 56
Epilogue 59


Introduction 61
Materials and Methods 63
Results 77
Discussion 95


Introduction 105
Materials and Methods 106
Results 117
Discussion 135


Introduction 144
Materials and Methods 146
Results 154
Discussion 172






Table II-I Lineage-Specific Antibodies Used in Surface
Marker Analysis

Table 11-2 Histamine Content of Embryonic Tumor Cell
Lines and Control Tumor Cell Lines

Table 11-3 Analysis of Lineage-Specific Surface
Determinants on A-MuLV-Transformed
Embryonic and Control Tumor Cell Lines

Table 11-4 Analysis of Surface Membrane Receptors for
IgE and IgG on A-MuLV-Transformed Embryonic
Cell Lines and on Control Tumor Cell Lines

Table 11-5 Metachromatic Granules in A-MuLV-Transformed
Embryonic Cell Lines and Control Tumor Cell

Table 11-6 Interleukin 3 Content of Conditioned Media
and Cell Lysates of A-MuLV-Transformed
Embryonic Cells and Control Tumor Cells

Table III-1

Table IV-1

Effect of Cocultivation of Culture-Derived
Mast Cells with Adherent Cells and Their
Conditioned Media

Frequency of Mast Cell Precursors in Adult
Bone Marrow from Homozygous and
Heterozygous Mice

Table IV-2 Expression of Surface and Cytochemical
Markers on Colony-Derived Mast Cells

Table IV-3

Table IV-4

Sorting of Control Cells by Rosetting

Sorting of Bone Marrow Cells by Surface








Figure II-1

Figure 11-2

Detection of Cell Surface Determinants
on Abelson Murine Leukemia Virus-
Transformed Embryonic Cells

Detection of Surface Receptors for IgE
on A-MuLV-Transformed Embryonic Cells

Figure 11-3 Virus-Transformed Cells Contain Abelson
Murine Leukemia Virus-Specific DNA

Figure 11-4

Figure III-i

Figure III-2

Figure III-3

Figure III-4

Figure III-5

Figure III-6

Figure III-7

Figure IV-1

Tumors Isolated from Mice Injected with
Cell Lines 10P12 and 11PO Contain
A-MuLV-Specific DNA Sequences

Progression of Hematopoietic Lineage
Markers in Long-Term Mast Cell Cultures
Derived from Embryonic Tissues

Metachromatic Granules of Long-Term,
Culture-Derived Embryonic Mast Cells

Progression of Hematopoietic Lineage
Markers In Long-Term Mast Cell Cultures
Derived from Adult Bone Marrow

Population Dynamics of Bone Marrow-
Derived Mast Cells Infected with

Expression of Ly5 Antigen on A-MuLV-
Infected Mast Cells

Mixed Population of RA3-3Al-Positive
Lymphoid Cells and RA3-3Al-Negative
Cultured Mast Cells in Long-Term Bone
Marrow Cultures

Abelson Murine Leukemia Virus-Infected
Mast Cells Express v-abl Gene

Colonies in Long-Term Agar Cultures of
Embryonic Cells in Conditioned Media

Figure IV-2 Frequency of Mast Cell Precursors in
Midgestation Embryonic Tissues











Figure IV-3

Frequency of Placental Mast Cell
Precursors in the Third Trimester of



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



Michael L. Siegel

May 1986

Chairman: Edward J. Siden
Major Department: Immunology and Medical Microbiology

The embryonic origin and ontogeny of mast cells is

poorly understood, despite a growing body of literature

relevant to that area of study. We have systematically

investigated the development of mast cells in the embryonic

mouse, beginning our studies with the observation of mast

cell characteristics of midgestation embryonic placental

cells transformed with the defective retrovirus, Abelson

murine leukemia virus. Unlike previously reported Abelson

virus-transformed cells, the placental cell lines exhibited

many of the characteristics of culture-derived mast cells,

including differentiation antigens, high affinity receptors

for IgE, and metachromatic granules containing histamine

and sulfated proteoglycans. Some of the cell lines also

expressed the B220 marker previously reported to be

specific for cells of the B lymphoid lineage. We also


developed a simple, sensitive, nonfluorometric, nonisotopic

assay to detect membrane receptors for immunoglobulins.

The observation of the mast-like, Abelson virus-

transformed cell lines led us to investigate the presence

of mast cell precursors in normal midgestation embryonic

tissues. We found embryonic precursors to mast cells in

homologous, noninfected tissues and conducted a detailed,

systematic analysis of the differentiation of mast cells in

liquid cultures over the course of several weeks of

selection and enrichment. We also studied the effects of

Abelson virus infection and adherent cell cytokines on

lymphoid differentiation antigens in mast cell cultures.

Mast cell precursors in embryonic tissues of mid- and

late gestation were quantitated by a clonal assay. We

described the embryologically earliest reported mast cell

precursors in the mouse and report that the mouse embryo is

a rich reservoir of such precursors, containing

proportionately at least as many such cells as adult bone

marrow. We have observed that mast cells which

differentiate in agar culture, like some of the Abelson

virus-transformed cell lines, express the B220 determinant.

We have also described preliminary experiments in which we

selected mast cell precursors in bone marrow on the basis

of surface membrane determinants.


Introduction: Hematopoietic Cell Differentiation
and Tumor Models

The ontogeny of the hematopoietic system of the mouse

can be viewed as a progression of finite, genetically

programmed stages in the maturation of pluripotent stem

cells into the terminally differentiated state of each of

the various blood lineages. Pluripotent hematopoietic stem

cells, defined by their ability to reconstitute lethally

irradiated recipients (Till and McCulloch, 1961), are first

detected in the murine yolk sac between eight and twelve

days of gestation (Tyan, 1968). Beginning with day ten and

throughout the remainder of gestation, cells with the same

differentiative capacity are found in the fetal blood and

liver (Moore and Metcalf, 1970). In the adult, pluripotent

hematopoietic stem cells are found in the bone marrow (Till

and McCulloch, 1961) and spleen (Nakahata and Ogawa, 1982).

The mechanisms involved in the differentiation of

pluripotent hematopoietic stem cells into mature,

functional blood elements are, for the large part, unknown.

These pathways may involve the interaction of pluripotent

or committed progeny stem cells with other cells or

macromolecular products in their inductive environment

(Kincade et al., 1981a), resulting in the cell's commitment

to one of several genetically programmed, phenotypically

distinguishable chains of events; alternately, random

stochastic processes may play a role in the differentiation

of hematopoietic cells (Nakahata et al., 1982a; Suda et

al., 1984).

The differentiation pathways of hematopoietic cells

and the molecular events involved in normal hematopoiesis

are best understood and defined by delineating discrete

cellular intermediates. Observation and identification is

frequently hampered by hematopoietic tissue heterogeneity,

short life span, and low frequency of cells of interest.

These difficulties are overcome in part by virus-induced

transformation of such cells, resulting in relatively

homogeneous populations of adequate size and frequently

unlimited growth potential. Although transformed

homologues of normal cells are frozen at a particular stage

of development by the action of the transforming gene

product, virus infection may also induce phenotypic changes

which are unparalleled in the course of normal

differentiation. To avoid this pitfall, it is therefore

prudent to confirm that the characteristics of tumor cell

models of early blood progenitors mimic their naturally

occurring counterparts.

Abelson murine leukemia virus (A-MuLV) is a

replication-defective retrovirus which arose by

recombination of portions of the genome of the replication-

competent, thymotropic Moloney murine leukemia virus and a

cellular gene, c-abl (Shields et al., 1979; Goff et al.,

1980) in a prednisolone-treated BALB/cCR mouse (Abelson and

Rabstein, 1970). A-MuLV is capable of rapid transformation

of bone marrow-derived thymus-independent lymphoid cells in

vivo (Sklar et al., 1975; Premkumar et al., 1975) and in

vitro (Rosenberg et al., 1975; Baltimore et al., 1979;

Rosenberg and Baltimore, 1976a, 1980).

Although most of the in vivo transformants reportedly

have been of the B cell phenotype (Premkumar et al., 1975),

plasmacytomas have also been reported (Potter et al.,

1973). T cell lymphomas (Cook, 1982), myelomonocytic

leukemias (Raschke et al., 1978), and mast cell tumors

(Mendoza and Metzger, 1976; Risser et al., 1978; Pierce et

al., 1985) have also resulted from in vivo infections,

indicating that Abelson virus may affect the growth and

differentiation of multiple hematopoietic lineages.

Abelson murine Leukemia virus is capable of

transforming both hematopoietic and nonhematopoietic cells

in vitro. Rosenberg and Baltimore tRosenberg et al.,

1975; Rosenberg and Baltimore, 1976a, 1976b) have developed

an in vitro culture system for the transformation and

clonal proliferation of lymphoid cells from murine

hematopoietic tissues. Under well-defined conditions,

permanent cell lines with pre-B lymphoid characteristics

(Siden et al., 1979) have been generated. These cell lines

are believed to exhibit the earliest known differentiation

markers and immunoglobulin gene organizational structures

of pre-B cells (Boss et al., 1979; Siden et al., 1979; Alt

et al., 1981, 1984). Using identical conditions for

transformation of mouse placenta and fetal liver, Waneck

and Rosenberg (1981) described colonies of cells expressing

various differentiated erythroid characteristics,including

cessation of growth. Unlike its Moloney leukemia virus

ancestor, A-MuLV transforms the NIH/3T3 fibroblast cell

line (Scher and Siegler, 1975).

Modifications of the original Rosenberg and Baltimore

in vitro transformation protocol have resulted in the

transformation of phenotypically disparate lineages.

Whitlock and colleagues (1983) and Serunian and Rosenberg

(1986) have reported the transformation of more

differentiated B lineage cells from in vitro-derived

"normal" populations. More recently, permanent cell lines

expressing mast cell characteristics have been reported

following transformation of midgestation embryonic placenta

(Siegel et al., 1985) and third trimester fetal liver

targets (Pierce et al., 1985).

Recent reports from several laboratories indicate that

retroviruses may alter the growth factor requirement of

cells of several lineages. Rapp and colleagues (1985)

reported the development of factor-independent cell lines

following transformation of interleukin 2-dependent T

lineage and interleukin 3-dependent myeloid cell lines by

recombinant viruses bearing the v-myc oncogene. Abelson

virus was similarly capable of abrogating the interleukin

3-dependence of both myeloid (Greenberger et al., 1979;

Cook et al., 1985) and mast (Pierce et al., 1985; Chapter

III, this dissertation) lineages. Taken with the

observation of erythropoietin-independent evthroid cells

(Waneck and Rosenberg, 1981) and interleukin 3-independent

mast cells (Pierce et al., 1985; Siegel et al., 1985) from

A-MuLV-infected primary cell populations, the data suggest

that A-MuLV may alter the requirements of these cells for

growth factors while permitting the cells to differentiate.

Although the precise role of the v-abl oncogene in

maintaining factor-independent proliferation is not yet

known, it is interesting to note that the cellular

homologue, c-abl, is transcribed at its highest level in

the developing embryo at the same time that the number of

A-MuLV targets reaches its peak. At least two retroviral

oncogene products have recently been shown to have

analogous structures in normal cells. The epidermal growth

factor receptor exhibits striking similarity to the erb B

oncogene product of avian erythroblastosis virus (Downward

et al., 1984), while the v-sis oncogene of simian sarcoma

virus encodes a protein structurally and immunologically

related to platelet-derived growth factor (Doolittle et

al., 1983; Robbins et al., 1983; Waterfield et al., 1983).

Although the function of c-abl has not been established, it

is known that all of the hematopoietic lineages sensitive

to A-MuLV transformation are also sensitive to the

proliferative effects of interleukin 3 (reviewed in Iscove

and Roitsch, 1985; Rennick et al., 1985). Seminal studies

on the culture of mast cells revealed that cells of this

lineage would proliferate in the presence of media

conditioned by lectin- or antigen-stimulated T lymphocytes

or by the myelomonocytic leukemia cell line WEHI-3B

(Hasthorpe, 1980; Nabel et al., 1981; Nagao et al., 1981;

Schrader, 1981; Tertian et al., 1981). More recently,

interleukin 3, the proliferative factor in the conditioned

media, was purified to homogeneity (Ihle et al., 1983;

Razin et al., 1984a). Subsequent studies performed with

the glycoprotein product of cloned interleukin 3 gene have

substantiated the proliferative activity of the factor

(Yokota et al., 1984; Rennick et al., 1985).

Mast Cell Function, Origin, Ontogeny, and Heterogeneity

Contemporary knowledge and interest in the mast cell

has its roots in the midnineteenth century. The earliest

description of these cells is found in the work of von

Recklinghausen (1863), who observed and illustrated the

mast cell in the unstained mesentery of the frog. Credit

for the discovery of the mast cell, however, is generally

assigned to Paul Ehrlich, a young physician-scientist who

was then interested in the differential staining affinities

of certain tissue cells and their organelles. Ehrlich

(1877) first described mast cell-like cells in several

species as elements which stained atypically red-violet

with the blue basic analine dye dahlia. The term mast cell

is derived from Ehrlich's nomenclature (Mastzellen), which

he assigned to granular cells which were abundant in well-

nourished tissues of frogs (Ehrlich, 1879). In the same

work, Ehrlich first used the term "Metachromasie," or

metachromasia, to describe the anomalous staining of these

connective tissue cells. Aside from their role as

histochemical curiosities, however, few references to mast

cell derivation, functions, and heterogeneity were

published in the ensuing fifty years (Selye, 1965).

Mature mast cells have been attributed with a number

of physiologic functions, many of which have been reviewed

in the recent literature (Austen, 1984; Shanahan et al.,

1984; Katz et al., 1985a; Lagunoff, 1985). The most widely

known mast cell function is the anaphylactic response,

which was first described by Selye in 1937 (cited in Selye,

1965). The response was originally experimentally induced

in rats by intraperitoneal injection of egg white and

produced hyperemia and edema of the lips, ears, paws, and

genitalia which were aggravated by adrenalectomy and

ablated by stressors like formaldehyde, which induce

adrenocortical hyperplasia. The release of histamine

during anaphylactoid inflammation was hypothesized at this

time, although release of the mediator was not

experimentally associated with mast cells until 1954

(Benditt et al., 1954). Other substances shown to be

released during the anaphylactic response include heparin

and related proteoglycans, leukotrienes, prostaglandins, 5-

hydroxytryptamine and other amines, and neutral proteases

(reviewed in detail in Austen, 1984).

Mast cells have been directly or indirectly implicated

in a number of other physiologic roles. Histamine has been

associated with modulation and regulation of the immune

system (Askenase et al., 1981) including reduction of T

cell effector function (Plaut et al., 1973; Schwartz et

al., 1980) and decreased lymphokine production (Rocklin,

1976). Mast cells have also been implicated in delayed-

type hypersensitivity reactions (Askenase, 1977), immune

complex formation (Benveniste et al., 1972), natural

cytotoxicity (Farram and Nelson, 1980; Henderson et al.,

1981), and parasite resistance (Capron et al., 1978), as

well as elaboration of a factor akin to interleukin 1 which

potentiates inflammation and collagenase activity in

synovial cells (Van Den Hoof and Tichelar-Guttar, 1983;

Yoffee et al., 1984, 1985).

Nonimmune functions have also been ascribed to mast

cells. "Microenvironmental hormones" (Lewis and Austen,

1981) produced by mast cells have been implicated in tissue

growth and repair (Kahlson and Rosengren, 1968) and

thyroxine secretion by the thyroid (Melander, 1977).

Histamine was shown to be required for timely blastocyst

implantation (Nalbandov, 1971; Dey and Johnson, 1980a,

1980b; Dey, 1981), as were prostaglandins (Kennedy, 1977),

but the number of detectable mast cells in the gravid

uterus was shown to decrease after implantation

(Shelesnyak, 1960; Brandon and Bibby, 1979). Mast cell

association with nerve tissue was first noted in 1878

(cited in Selye, 1965) and intimate contact between nerve

endings and mast cell plasma membranes was documented by

Weisner-Menzel and colleagues (1981) and Newson and

colleagues (1983). Histamine release has been attributed

to stimulation of cutaneous nerves (Kiernan, 1972), thus

establishing a neuroendocrine-mast cell axis. Mast cells

in alimentary tract mucosa have been implicated in

promotion of gut mobility (Erjavek et al., 1981; Fjellner

and Hagermark, 1981); histamine-induced gastric secretion

was shown to be blocked by histamine H2 receptor agonists

(Soll et ai., 1981) and enhanced by glucocorticoids

(Sathiamoorthy et al., 1976). The latter observation is

contradictory to published reports of glucocorticoid-

induced suppression of intestinal anaphylaxis (King et al.,

1985), and may indicate different modes of action of

steroids on sensitized and nonsensitized mast cells. Thus,

Daeron and colleagues (1982) noted that glucocorticoids

inhibited antigen-induced, but not calcium ionophore

A23187-induced, histamine release from mast cells.

The derivation of mast ceils in mammalian tissues has

been the subject of considerable investigation, and

sometimes heated debate, since the early nineteen sixties.

Because the developmentally earliest described mast cells

were observed in connective tissue, and a gradation of

"immature" to "mature" forms of mast cells could be

isolated from this source, several investigators proposed

that mast cells were derived directly from connective

tissue precursors (Burton, 1963; Michels, 1963). Combs and

colleagues (.1965) observed the development of mast cells in

embryonic rats between fifteen and twenty-three days post

coitum. The cells appeared to arise in undifferentiated

mesenchymal tissue and progressed through a gradation of

intermediates to characteristic mast cells. Asboe-Hanson

(1971) further noted that mast ceils in the skin appeared

to differentiate locally from mesenchymal elements.

Based on observations of profound mastocytosis

associated with immune and neoplastic lymphocyte

proliferation, a second faction proposed the hematopoietic

origin of mast ceils. Accordingly, Ginsburg (1963) removed

thymuses from mice and cultured them with embryonic skin

monolayers. The mast cells so derived led the author

(Ginsburg, 1963; Ginsburg and Lagunoff, 1967) to propose

that mast cells were derived from thymocytes, an

observation subsequently confirmed in the rat (Ishizaka et

al., 1976). Burnet (1965, 1975, 1977), observing that mast

cells exhibited evidence of thymic origin or dependence,

participated in immunologic reactions, were similarly

activated by lectins and immune stimuli, and were

relatively amitotic, speculated that the mast cells were
"post-mitotic" T lineage cells.

Other hematopoietic cell types have been postulated to

be mast cell by precursors in vivo and in vitro cultivation

techniques. Desaga and colleagues (1971) performed

repeated peritoneal lavages on rats to deplete them of

mature serosal mast ceils. Mast cell-deficient

peritoneal exudate cells, cytochemically identified as

monocytes at the beginning of in vitro growth, developed

into metachromatic, granulated mast cells within two days

of harvest. Similarly, Czarnetzki and Behrendt (1981)

reported that peritoneal exudate cells from mast cell

depleted rats (injected with sterile water

intraperitoneally) resembled mononuclear phagocytes both

morphologically and cytochemically before and shortly after

culture in L-cell conditioned media. The in vitro-

propagated population, however, differentiated into mast

cells which were identified morphologically and which

contained granules with histamine and alpha-naphthol

acetate esterase.

The hematopoietic origin of some, and perhaps all,

mast cells was defined in great detail by Kitamura and

colleagues in a series of reports beginning less than ten

years ago. Initially, they demonstrated that irradiated,

mast cell-deficient mice could be reconstituted by

injection of bone marrow from untreated donors (Kitamura et

al., 1977). Since irradiation did not eradicate all

recipient mast cells prior to reconstitution, a donor

strain with phenotypically distinct mast cell granules

(beige) was used to definitively demonstrate the origin of

the cells which were detected. Similar experiments, using

unirradiated, genetically mast cell-deficient W/Wv mice,

demonstrated that adult bone marrow (Kitamura et al., 1978;

Hatanaka et al., 1979), blood (Kitamura et al., 1979a) and

spleen (Kitamura et al., 1979d) were rich reservoirs of

mast cell precursors which were also found in smaller

numbers in thymus, lymph node, and Peyer's patches

(Kitamura et al., 1979d). Mast cells were also detected in

fetal liver populations despite the apparent lack of mature

mast cells in that tissue (Kitamura et al., 1979c).

The hematopoietic nature of the in vivo mast cell

precursor was further defined by the same group. fn

preliminary studies, Kitamura and colleagues (1981) showed

that genetically mast cell-deficient W/Wv mice could be

reconstituted with cells from individual spleen colonies of

normal (C57BL/6) mice which had been irradiated and

subsequently reconstituted with bone marrow cells. Having

assured themselves of the clonality of each donor spleen

colony (by injecting mixed phenotypically distinct beige

and wild type bone marrow cells into the primary,

irradiated recipients and screening the colonies for cells

with only one type of granule), the authors were able to

conclude that the colony forming unit of the spleen (Till

and McCulloch, 1961) was the ultimate progenitor cell for

mast cells in the spleen, stomach, caecum, and skin as well

as for peripheral blood granulocytes and erythrocytes.

These results were corroborated by Sonada and colleagues

(1983), who demonstrated that late (twelve days post

injection) spleen colonies, which included both erythroid

and myeloid elements, contained mast cell precursors which

differentiated in secondary recipient skin.

The ontogeny of embryonic mast ceils in mice is

meagerly defined in the literature. As previously

mentioned, in vivo mast cell precursors are present,

despite the absence of more mature forms, in the mouse

fetal liver thirteen and more days post coitum (Kitamura

et al., 1979c). Embryonic mast cell ontogeny was better

defined in the rat. Csaba and Kapa (1960) demonstrated the

presence of mast cells, which incorporated exogenous

heparin, in the thymus, spleen, lymph nodes, myocardium,

and kidney of day sixteen to seventeen rat embryos. By

studying sections through rat embryos between fifteen and

twenty-three days of gestation, Combs and colleagues (1965)

placed mast cells into four stages of differentiation based

upon their staining characteristics with alcian blue and

safranin, dyes which preferentially bind to the

glycosaminoglycan components of mast cell granules. By

this technique, alcian blue binds to poorly sulfated

glycosaminoglycans like chondroitin sulfate, while safranin

binds to highly sulfated molecules like heparin. A

gradation of cells, beginning with large, lymphocyte-like

elements with few, aician blue-stained granules (Stage I),

and progressing through characteristic mature mast cells

with small nuclei and large numbers of safranin-staining

granules (Stage IV). was documented. Mast cell stages were

also differentiated by nuclear characteristics (mitotic

figures), granule heparin (periodic acid-Schiff staining),

granule glycosaminoglycan synthesis (sodium sulfate uptake),

granule histamine (diazotized parabromoaniline reaction),

and granule protease (phenylproprionyl naphthol AS

reaction) content. The first recognizable mast cells

(Stage I) were found in the head mesenchyme at fifteen days

of gestation. Mast cell numbers rapidly increased during

the sixteenth day of gestation, with Stage II mast cells

found in the connective tissue of the dorsal vertebrae.

Subcutaneous (Stage III) mast cells were identified in

tissues of embryos eighteen and nineteen days post coitum.

Mature Stage IV mast cells were first present shortly

before birth and were indistinguishable from adult

connective tissue mast ceils.

The study of mast cells, which has accelerated

dramatically in the past quarter century, has revealed that

Ehrlich's "Mastzellen" are heterogeneous both

evolutionarily interspeciess) and functionally

(intraspecies). Mast cell heterogeneity has been the topic

of numerous general reviews in recent years (Enerback,1981;

Bienenstock et al., 1982, 1983; Pearce, 1982, 1983;

Shanahan et al., 1984; Jarrett and Haig, 1984; Austen,

1984: Katz et al., 1985a; Lagunoff, 1985; Pearce et al.,

1985). From this sea of literature, the differences

between mast ceils of various species are apparent. The

multipotent biogenic compound of mammalian mast cells,

histamine, has not been detectable in fishes and

amphibians, while serotonin is apparently unique to rodent

and dopamine to bovine mast cells. Heparin proteoglycan

from porcine, bovine, rat, and human sources is

heterogeneous in molecular weight and charge (Stevens and

Austen, L981). The proteolytic enzymes of mast cells are

also phylogeneticallv disparate; rodent mast cells contain

an alpha-chymotrypsin-like activity, while dog, human, and

turtle mast cells have trypsin-like activity and bird and

fish mast cells have no esteroprotease activity (Woodbury

and Neurath, 1980; Lagunoff, 1985). Ultrastructurally, the

granules of human mast cells appear to be organized in

crystalline scrolls, while rat mast cell granules are

dense, homogeneous spheres which dissociate into fibrillar

structures in hypertonic salt solution (Lagunoff, 1972).

The initial observation of intraspecies mast cell

heterogeneity is generally attributed to Maximow (1906),

who reported that the rat intestine was replete with mast

cells which differed from other rat mast cells in

morphology and stain affinity. These differences were

reinvestigated by Enerback, who, sixty years after

Maximow's observations, published a series of reports which

described in detail the differences in morphology, stain

affinities (Enerback, 1966a, 1966b), and sensitivity to

degranulating agents (Enerback, 1966c, 1966d) between

dermal mast cells (representative of the connective tissue

or serosal subset) and intestinal mast cells

(representative of mucosal or atypical mast cells). Thus,

mucosal mast ceils were shown to be smaller, possess uni-

or bilobed nuclei, and be less granulated than serosal mast

cells, and the granules of the former population were far

more heterogeneous in size than those of the latter.

Enerback also observed that the mucosal mast cells stained

red with acidic toluidine blue, while serosal mast cells

stained purple. He noted that standard formaldehyde

fixatives used to preserve serosal mast cell granules were

ineffective on mucosal mast cells, and selected and

adapted several fixatives (such as Carnoy's and Mota's

preparations) to more adequately preserve the more labile

mucosal mast cell granules. Mucosal mast cells also

required higher concentrations of thiazine dyes, like

toluidine blue, and azure A dyes, as well as prolonged

staining times, when compared to serosal mast cells, while

the granules of the former cells had higher affinity to

copper phthalocyanine dyes such as Astra blue at ph 0.3.

Enerback therefore concluded that the mucosal mast and

connective tissue mast cells differed not only

morphologically, but biochemically as well, and offered

that the mucosal mast cells contained less highly sulfated

mucopolysaccharides than the dermal cells. Finally,

Enerback noted that in rats systemically exposed to the

histamine releasing agent 48/80, serosal mast cells in the

mesentery, tongue, and skin were degranulated and therefore

undetectable, while mast cells in the duodenal mucosa were

unaffected and perhaps greater in number. While being

unable to explain the latter observation, Enerback was able

to conclude that mucosal mast cells differed from their

serosal counterparts functionally as well.

The differences between mucosal and serosal mast

cells, first noted by Maximow and then Enerback, have

since been appended by the observations of numerous

investigators and extend beyond those mentioned to surface

markers, histamine content, IgE receptors and

internalization of bound IgE, proteoglycan composition,

proteases, sensitivity to histamine secretagogues, effects

of neuropeptides and endorphins, thymus dependence, and

life span. These characteristics have been surveyed in

detail in reviews previously cited. We will therefore only

briefly survey the literature which is cogent to the

ultimate topic of this discussion, the in vitro, culture-

derived mast cell.

Connective Tissue Mast Cells

Although typical mast cells have been isolated from a

variety of connective tissue sources throughout the rodent

body, the most frequently studied member of this subset is

that which is isolated, free of extraneous tissues, from

the serosal surfaces of the peritoneal cavity. As

previously discussed, by injecting bone marrow into mast

cell-deficient hosts, Kitamura and colleagues (1977) were

able to demonstrate the relationship of hematopoietic

precursors to the serosal mast cells. The ultimate

precursor cell in the bone marrow was shown to be the

colony-forming unit of the spleen (Kitamura et al., 1981;

Sonoda et al., 1983). From the bone marrow, mast cell

precursors migrate through the blood (Kitamura et al.,

1979a; Zucker-Franklin et al., 1981; Sonoda et al., 1983)

and subsequently proliferate and differentiate in

connective tissue (Hatanaka et al., 1979; Kitamura et al.,

1979b,1979d). Connective tissue mast cells in rodents have

been found to proliferate and differentiate independently of

thymic influences. Thus, the athymic nude mouse has mast

cells in its connective tissues (Wlodarski, 1976; Reed et

al., 1982). Aldenborg and Enerback (1985) recently reported

that congenitally athymic rnu/rnu rats have at least as many

(or more) peritoneal mast cells as normal controls for the

first fourteen weeks of life; adult rnu/rnu rats, however,

have fewer peritoneal mast cells than their wild type

counterparts. These results may indicate that peritoneal

mast cell populations may be subject to thymic influences

later in life or may simply reflect a separate, thymus-

independent defect inherent in the athymic rat. Further

studies will be necessary to elucidate the apparent


The proteoglycan composition of serosal mast cell

granules has been the subject of study since the initial

discovery of heparin in the canine liver by Jorpes in 1937

(cited in Selye, 1965). Subsequently, heparin has been

identified in rat peritoneal mast cells (Tas and

Berndsen, 1977; Yurt et al., 1977; Stevens and Austen,

1982), human lung mast cells (Metcalfe et al., 1979), and

mouse peritoneal mast cells (Razin et al., 1982c).

Mucosal Mast Cells

Following the development of improved methods for

their fixation and staining by Enerback (1966a, 1966b), the

study of the atypical, or mucosal, mast cell accelerated

significantly. Early reports of similarities between

cultured, thymus-derived mast cells of the mouse (Ginsburg,

1963; Ginsburg and Lagunoff, 1967) and the rat (Ishizaka et

al., 1976) and Enerback's atypical mast cells of the lamina

propria spurred erroneous theories that the mucosal mast

cell was derived from the thymus (Burnet, 1965, 1975,

1977). Based upon observations of mucosal mastocytosis

following experimental infection of rats and mice with the

nematodes Nippostrongylus brasiliensis and Trichinella

spiralis, a lymphoid origin of these cells was also

proposed by other investigators (Rose et al., 1976; Befus

and Bienenstock, 1979; Mayrhofer, 1979a,1979b; Nawa and

Miller, 1979).

The derivation of mucosal mast cells was ultimately

resolved by Crowle (1982), who reconstituted mucosal mast

cell-deficient mice with cells derived from a variety of

hematopoietic tissues. Crowle observed that W/Wv mice

could be reconstituted by bone marrow and spleen cells but

not by thymocytes or thymus grafts, while athymic mice

could be reconstituted by thymus grafts, thymocytes or

splenocytes. Crowle proposed that the W/Wv mice were

defective in mucosal mast cell precursors which were

present in bone marrow (and spleen) of normal mice, while

athymic mice possessed the precursor population and needed

a thymus-related component to effect differentiation.

Crowle concluded that mucosal mast cells were derived from

bone marrow and required a thymic influence for

accumulation in mucosal surfaces. These relationships

were further reinforced by the same investigator (Crowle

and Reed, 1984). Reconstitution of athymic mice was

ablated by pretreatment of wild type mouse bone marrow

cells or splenocytes with anti-Thy 1 and complement, while

similar treatment of beige bone marrow or spleen cells

still resulted in the detection of some mast cells, albeit

fewer, in the mucosa of thymus-intact W/Wv mice.

The thymic dependence of mucosal mast cells, in

contrast to the thymus-independent growth and development

of serosal mast cells, has been documented by a number of

other investigators. Prior to the reports of Crowle

(1982,1984), Ruitenberg and Elgersma (1976) observed that

nude mice infected with Trichinella spiralis experienced no

intestinal mucosal mast cell response unless reconstituted

by thymus or parasite-immune thoracic duct cell grafts,

concluding that thymus-derived T-lineage cells were required

for mucosal mastocytosis. These results were substantiated

by a number of other researchers (Olson and Levy, 1976;

Mayrhofer and Bazin, 1981; Reed et al., 1982). Similar

studies were performed in the rat. Mayrhofer (1979a) noted

that the number of mucosal mast cells in Nippostrongylus

brasiliensis-infected rats increased in a pattern similar to

primary and secondary immune responses. Adult thymectomy or

chronic thoracic duct drainage several months prior to

nematode challenge (to deplete mature T cells) resulted in

significantly depressed intestinal mastocytosis, while

thymectomy shortly before challenge was ineffective in

ablating the mast cell response (Mayrhofer, 1979b). These

last observations indicated that the thymus, per se, is not

the immediate source of mast cells or mast cell growth

factors. Similar depression of mucosal mast cell response

and poor clearance of intestinal parasites were observed in

B rats, thymectomized, irradiated animals which were

reconstituted with bone marrow of T cell-depleted (chronic

thoracic duct drainage) donors (Mayrhofer and Fisher, 1979).

Interestingly, contrary to the reports of Crowle (Crowle,

1982; Crowle and Reed, 1984), unchallenged B rats, as well

as athymic nu/nu mice, were reported to have normal numbers

of mucosal mast cells when compared to appropriate controls

(Mayrhofer and Bazin, 1981).

Three other independent lines of evidence fortified

the hypothesis that mucosal mast cells were dependent on a

T cell-derived proliferation-differentiation factor.

First, the primary mucosal mast cell response to

Nippostrongylus brasiliensis in the rat was enhanced by

adoptive transfer of immune T cells (Nawa and Miller,

1979). Second, mucosal mastocytosis was demonstrated in a

variety of other immune scenarios, including the

inflammatory reactions of ulcerative colitis, Crohn's

disease, and pulmonary fibrosis (Askenase, 1980). Third,

Guy-Grand and colleagues (1984) recently reported the

direct stimulation of intestinal mucosal mast cell

precursors in BALB/c mice bearing the myelomonocytic

leukemia WEHI-3, a constitutive producer of the mast ceil

growth factor interleukin 3, which is identical to the mast

cell growth factor produced by activated T lymphocytes

(Yung et al., 1981).

A variety of other distinguishing characteristics have

been ascribed to the cells alternately called mucosal mast

cells, atypical mast cells, and histaminocytes (Code,

L977). Early studies of mucosal mast cells were performed

on tissue sections or on heterogeneous populations of cells

isolated from the gut mucosa. The previously cited

methodologies of Enerback (1966a, 1966b) were later

optimized by the inclusion of techniques which further

stabilized the granules (neutral formalin fixation) and

enzymatically (with trypsin) stripped stain-retarding

proteins from glutaraldehyde-treated preparations (Wingren

and Enerback, 1983). The recent development of methods for

isolating such cells from the small intestine, which

exploited the mucosal mastocytosis induced by parasitic

infection, was reported by Befus and colleagues (1982a),

and made it possible to analyze mucosal mast cells in the

absence of extraneous elements and to confirm some previous

observations. In contrast to peritoneal mast cells,

isolated mucosal mast cells are smaller and have a shorter

lifespan. Mucosal mast cells have fewer granules which

contain nonheparin, lower sulfated proteoglycans (as vet

biochemically undefined), less histamine, and serotonin

(Bienenstock et al., 1983).

The response of mucosal mast cells to a variety of

secretagogues has been the subject of a number of studies.

Similar to serosal mast cells, mucosal mast cells are

responsive to the degranulation effects of IgE and antigen,

IgE and anti-IgE, concanavalin A, ionomycin, and compounds

23187 and Br-X537A, albeit with the release of less of

their total histamine content (Befus et al., 1982a, 1982b;

Pearce et al., 1982). Enerback's early observations on the

insensitivity of rat mucosal mast cells to degranulation by

compound 48/80 in vivo were confirmed by other

investigators in canine (Lorenz et al., 1969; Rees et al.,

1981) and murine (Enerback, 1981) models. This

unresponsiveness was confirmed in isolated mucosal mast

cells and extended to the secretagogue Bee Venom Peptide

401 (Befus et al., 1982a, 1982b), for which membrane

receptors were found to be absent on mucosal, but not

serosal, mast cells (Pearce et al., 1982). The same group

also reported that, unlike serosal mast cells, mucosal mast

cells were unresponsive to enhanced, antigen-induced

secretion of histamine mediated by phosphatidyl serine.

Mucosal mast cells were also shown to be distinct from

serosal mast cells in their responsiveness to secretary

antagonists disodium chromoglycate, theophylline, and

AH9679, while both subsets were equally sensitive to the

degranulation inhibitory effects of Doxantrazole (Befus et

al., 1982b; Pearce et al., 1982, 1985).


Cells with membrane receptors for IgE are not limited

to the mast cell lineage. Both lymphocytes (Gonzales-

Molina and Spiegelberg, 1978) and macrophages (Melewicz et

al., 1982) have been reported to bind IgE; the affinity of

the membrane receptors of the non-mast cells, however, was

ten to one-hundred times less than that of mast cells

(Ogawa et al., 1983). The best known of the mast-like

cells, however, are the basophils (basophilic

granulocytes), polymorphonuclear leukocytes which are

present in the blood of several mammalian species (Lagunoff

and Chi, 1980). Like mast cells, basophils contain

metachromatic granules and express surface membrane

receptors for IgE. Histamine is released by antigenic

challenge of IgE-bearing basophils (Lagunoff and Chi,

1980). The granules of basophils of several species have

been reported to contain chondroitin sulfate proteoglycan

(Olsson et al., 1970; Orenstein et al., 1978; Metcalfe et

al., 1980b) similar in its degree of sulfation to the

proteoglycan of mucosal mast cells (Tas and Berndsen,

1977). Basophilic granulocytes, however, are apparently

absent from mouse peripheral blood (Lagunoff and Chi,

1980). The relationship of these cells to culture-derived

mast cells in several species, including the mouse, will be

further discussed in a later section of this review.


The study of mastocytomas has contributed to the

understanding of mast cell growth, differentiation, and

function, and has provided the bridge between complex in

vivo investigations and better-defined, clonal population

analyses of in vitro cultures. Spontaneous mastocytoma,

while common in such species as dogs (Cobb et al., 1975;

Yoffee et al., 1984, 1985), is a more infrequent condition

in other species, notably rats and mice (Lagunoff, 1985).

Efrati and colleagues (1957) described human mastocytoma

cells as large, lymphocyte-like elements similar to the

early mast cell progenitors described in lower mammals

(Maximow, 1906).

The link between in vivo and in vitro mast cell

studies was established in 1959 when Schindler and colleagues

(1959) reported the successful adaptation of the

methylcholanthrene-induced, murine mastocytoma P815 (Dunn

and Potter, 1957) to growth in culture. The latter

investigators had isolated the neoplasm from a disseminated

disease with foci in the spleen and subcutaneous tissue and

had subsequently adapted it to a highly transplantable

ascites form from which a number of observations were made.

Variations in cell morphology, granule size and density,

and nuclear morphology were noted in the descriptive study.

In vitro analysis (Schindler et al., 1959) demonstrated

that the cells synthesized histamine and serotonin.

Through extensive culture, some sublines of P815 lost both

granules and intracellular histamine content; in some

sublines, however, the condition was reversed by addition

of sodium butyrate to the culture medium (Mori et al.,


Focal mastocytomas have frequently been associated

with the presence of lymphocytes (Galli and Dvorak, 1979;

Askenase, 1980), supporting the once popular hypothesis

that the latter cells were precursors to the former

(Burnett, 1965, 1975, 1977). It was also noted that some

mastocytomas, like basophils and culture-derived mast

cells, synthesize chondroitin sulfate proteoglycan rather

than the heparin (Lewis et al., 1973) found in serosal mast


Some mastocytomas have been reported during the course

of experiments analyzing Abelson murine leukemia virus-

induced lymphomagenesis (Mendoza and Metzger, 1976; Risser

et al., 1978). The mastocytomas generally arose in

peritoneal oil granulomas evoked by tetramethylpentadecane

(Pristane) in mice which were inoculated with Abelson virus

thirty to forty days after administration of the oil

(Pierce et al., 1985). The tumors were generally

transplantable into syngeneic hosts and frequently could be

adapted to growth in vitro. In a yet unexplained, but

perhaps related note, Hasthorpe (1980) reported the

isolation and adaptation to in vitro growth of a factor-

dependent mast-like cell line from splenocytes of a DBA-2

mouse infected with Friend ervthroleukemia virus. Although

electron microscopy established the presence of budding C-

type virus particles on the highly granular cells,

Hasthorpe's FMP1.1 cell line was nontumorigenic upon

intraperitoneal or subcutaneous administration to syngeneic

hosts and could not proliferate in the absence of exogenous

growth factor in vitro.

Culture-Derived Mast Cells

The growing body of knowledge concerning mast cells is

due, in part, to the development of methodologies for the

selection, enrichment, and maintenance of culture-derived

mast cells. These cells possess many of the

characteristics of their in vivo correlates, the mucosal

mast cells, as demonstrated in the following text. A note

of caution, however, must be interjected into the seemingly

logical flow between naturally occurring in vivo mast cells

and culture-derived mast cells. Although similar by a

variety of criteria including morphology, biochemistry,

function, and growth factor dependence, culture-derived

mast cells are not "normal" in the sense that they have

been produced in anatomically foreign, although perhaps

physiologically sufficient, conditions. They may therefore

be considered models of their in vivo homologues until

definitive evidence permits us to conclude that the two

populations are completely identical. We will therefore

continue to use the term "culture-derived" mast cells, and

similarly distinctive terms, to maintain the tenor of this

caveat throughout the following discussion.

Adherent Feeder Layer Studies

The earliest reports of culture-derived mast cells by

Ginsburg (Ginsburg,1963; Ginsburg and Sachs, 1963) involved

a complex system of mouse thymocytes cultured on feeder

layers of mouse embryonic fibroblasts. It was apparent to

the authors that the monolayer was essential for

proliferation of mast cells. Culture of thymocytes in the

absence of the feeder layer failed to produce mast cells.

On the other hand, culture of embryonic skin fibroblasts

from eighteen day fetuses, without additional thymocytes,

infrequently resulted in mast cell outgrowth. The tissue

source of the mast cells, therefore, was disputable, and

quite possibly both thymocytes and embryonic feeder layers

contributed progenitors to the mast cell culture. The

issue was better defined several years later when the same

group reported that irradiated embryonic fibroblast

monolayers, which could no longer produce mast ceils when

cultured alone, could support the growth of culture-derived

mast cells from cocultured thymocytes (Ginsburg and

Lagunoff, 1967).

Thus, the culture-derived mast cell era was ushered in

with the observation that at least some mast cell

progenitors were present in lymphoid tissue and that

adherent cells were required for in vitro mast cell

differentiation and growth. In seminal attempts to clone

mast cell precursors in soft agar, Pluznik and Sachs (1965)

reported that embryonic feeder layers were again required

for outgrowth of mast cells from disaggregated splenocytes.

Ishizaka and colleagues (1977). investigating rat thymus-

derived mast cells in a system modeled after that of

Ginsburg, observed that clonal expansion was more prolific

in the presence of a feeder layer, although mast cells were,

indeed, isolated from cultures containing only thymocytes.

In the latter case, however, the mast cells were observed

to be associated with islands of fibroblast-like adherent

cells which arose in the thymocyte cultures.

The origin of mast cells in feeder layer cultures was

better resolved, almost twenty years after its initial

identification, by the definition of two morphologically

distinct populations of mast cells in mixed cultures of

adult lymphoid and embryonic feeder layer cells (Ginsburg

et al., 1982). Mast cells derived from the feeder layer

were morphologically similar to those found in connective

and serosal tissues, while those of lymphoid origin (lymph

node and thoracic duct) resembled mucosal mast cells, being

smaller in size with sparser, but larger granules than the

former cells. Pure mucosall mast cells" (in fact, culture-

derived mast cells of lymphoid origin) could be grown on

selected feeder layers which were free of serosal type

precursors. The mucosall mast cells" persisted, however,

only in the presence of T cell-derived factors. In

contrast, the mast cells derived from embryonic feeder

layers continued to persist, albeit without further

expansion, for six months or longer in the absence of

exogenous factors.

The same group (Davidson et al., 1983) later reported

that lymph node cells derived from unimmunized, horse

serum-immunized, and helminth-infected mice, grown in the

presence of conditioned media (from antigen-stimulated

mesenteric lymph node cells) but in the absence of

irradiated embryonic mouse fibroblast feeder layers,

proliferated (as large, vacuolated cells) but failed to

develop granules. When the undifferentiated, culture-

derived cells were transferred to fibroblast monolayers,

however, the cells developed metachromatic granules

containing histamine within seven days. Intimate contact

between the two populations of viable cells was apparently

essential to granule maturation, as neither fibroblast

conditioned media, fibroblast homogenates, glutaraldehyde-

fixed fibroblasts, nor separation of fibroblasts from

"large lymphocyte" mast ceil precursors by a membrane could

effect the change.

Despite the initial success in culturing mast cells

from lymphoid tissue cocultivated with adherent cell feeder

layers, reports in the literature of the technique's use

were limited to those of the previously cited groups. Two

factors probably contributed to the limited use of adherent

cell monolayers in mast cell culture. First, the system

was quite complex, requiring considerable time and

extensive subculture (or irradiation) of feeder layers to

eliminate the contribution of connective tissue type mast

cell precursors. Secondly, and perhaps more significantly,

the development of culture-derived mast cells in media

conditioned by activated lymphocytes by at least three

independent groups provided the opportunity to maintain

mast cells in the absence of a continuous monolayer of

feeder cells. As previously noted, however, even in the

absence of fibroblast feeder layers, islands of adherent

cells are observed in early cultures of lymphoid and

hematopoietic tissue-derived mast cells (confirmed in our

studies; see Chapters III and IV). Since the adherent

ceils are present in such cultures before the selection and

enrichment of mast cells, it is, at this juncture,

plausible to speculate that the adherent cells may assume a

transient, maturational role in mast cell differentiation.

Conditioned Media-Dependent Mast Cells

The development of techniques for culturing mast cells

from hematopoietic and lymphoid tissues has led to an

exponential increase in mast cell research and literature

citations. With the burst of scientific activity, however,

has come a concurrent increase in the number of terms used

to describe culture-derived mast cells, including P

(persisting) cells (Schrader and Nossal, 1980), histamine-

containing granular cells (Sredni et al., 1983), mucosal

mast cells, basophil/mast cells, and atypical mast cells.

Despite the discrepancy of terms, however, the long term

suspension cultures of mast cells appear to be strikingly

similar. Mast cells have thus been derived from mouse bone

marrow (Tertian et al., 1980, 1981; Nagao et al., 1981;

Razin et al., 1981a, 1982a,b,c; Schrader, 1981; Schrader et

al., 1981; Galli et al., 1982b; Crapper and Schrader, 1983;

Sredni et al., 1983; Wedling et al., 1983, 1985; Yung et

al., 1983; Suda et al., 1985), spleen (Hasthorpe, 1980;

Schrader and Nossal, 1980; Schrader et al., 1980, 1981;

Schrader, 1981; Tertian et al., 1981; Crapper and Schrader,

1983; Sredni et al., 1983; Pharr et al., 1984), fetal liver

(Nabel et al., 1981; Razin et al., 1984b), peripheral blood

(Crapper and Schrader, 1983; Suda et al., 1985), thymus

(Tertian et al., 1981; Schrader, 1981; Davidson et al.,

1983), lymph nodes (Ginsburg et al., 1978; Crapper and

Schrader, 1983), and intestinal mucosa (Schrader et al.,

1983b). Similar cells have been cultured from rat bone

marrow (Haig et al., 1982, 1983), peripheral blood (Zucker-

Franklin et al., 1981; Czarnetzki et al., 1983), and thymus

(Ishizaka et al., 1976, 1977) as well as human fetal liver

(Razin et al., 1981b), umbilical cord blood (Ogawa et al.,

1983), and adult peripheral blood (Denburg et al., 1983;

Czarnetzki et al., 1984).

The development of in vitro methods for mast cell

culture provided additional means of detecting embryonic

mast cell precursors. As previously noted, Ginsburg (1963)

observed occasional mast cell outgrowth in cultures of day

eighteen embryonic mouse skin. Similar experiments with

the rat model demonstrated that embryonic rat thymus,

isolated between eighteen and twenty days post coitum and

cocultured with adult rat thymocvtes or thymocyte

conditioned media, contained cells capable of

differentiating into mast cells (Ishizaka et al., 1976). A

third group (Nabel et al., 1981; Galli et al., 1982a) was

able to culture murine mast cells derived from day thirteen

fetal liver suspensions cultured in lymphocyte conditioned

media. Using the adherent cell system, Ginsburg and

colleagues (1982) were also able to demonstrate that mast

cells could be derived in culture from disaggregated mouse

embryos between ten and thirteen days of gestation. It was

thus apparent that the precursors of culture-derived mast

cells were present in the mouse embryo at ten days post

coitem, several days before mast ceils per se were

observable in the embryo (Kitamura et al., 1979c).

Interleukin 3 From Lymphocyte-Conditioned Media And Other


The analysis of culture-derived mast cells and their

precursors evolved from studies of hematopoietic cell

growth factors produced by lymphocytes. Reports of mast

cell-supporting factors (which, for the sake of convention,

we will commonly call interleukin 3) in the supernatants of

mitogen-stimulated splenocytes began to surface at the

beginning of the present decade (Burgess et al., 1980;

Hasthorpe, 1980). Since that time, a number of other

investigators have utilized media conditioned by a variety

of means to support the differentiation and growth of

culture-derived mast cells. Such media have thus been

derived from splenocytes activated by concanavalin A

(Clark-Lewis and Schrader, 1981; Tertian et al., 1981; Yung

et al., 1981; Schrader et al., 1981; Nakahata et al.,

1982b; Yung and Moore, 1982; Sredni et al., 1983), by

pokeweed mitogen (Hasthorpe, 1980; Nakahata et al., 1982b,

Wedling et al., 1983, 1984; Pharr et al., 1984), by

phytohemagglutinin A (Ogawa et al., 1983), by bacterial

lipopolysaccharide (Nakahata et al., 1982b), and by mixed

lymphocyte reactions augmented by lectin (Razin et al.,

1981a, 1982a, 1982c). Conditioned media with analogous

activity have been elicited from concanavalin A-stimulated

mesenteric lymph node cells of parasitized animals

(McMenamin et al., 1985). Phytohemagglutinin A- or

concanavalin A-stimulated human blood lymphocytes also

produced activities which supported the growth of human

basophilic cells and a growth factor-dependent mouse cell

line (Tadokoro et al., 1983; Stadler et al., 1985).

The phenotype of the murine lymphoid cell which

produces interleukin 3 was deduced from the activities of

conditioned media of a number of related T cell clones. In

contrast to Lyt 1+2+, Lyt 1-2+, and Lyt 1-2- cells, which

did not support culture-derived mast cell growth, the

supernatants of Lyt 1+2- T cell clones, corresponding to

the inducer T lymphocyte subset, supported the

proliferation of such cells (Nabel et al., 1981). The

observations of Nabel and colleagues were subsequently

confirmed by other investigators (Fung et al., 1984; Yokota

et al., 1984), and by the observation that rat mesenteric

lymph node cells expressing differentiation markers of

helper T lymphocytes (OX19+, W3/25+, 0X8-) were responsible

for the production of a factor with analogous activity to

mouse interleukin 3 (McMenamin et al., 1985).

A number of permanent cell lines also produce factors

which support the growth of culture-derived mast cells.

The best known of these cell lines is the myelomonocytic

WEHI-3 line, which constitutively produces high levels of

interleukin 3 (Nagao et al., 1981; Schrader et al., 1981;

Yung et al., 1981; Yung and Moore, 1982). Mast cell-

promoting activities have also been demonstrated in

conditioned media from concanavalin A-stimulated T cell

hybridoma cells (Clark-Lewis and Schrader, 1981), cloned T

cell lines (Nabel et al., 1981), iectin-stimulated T

leukemias (Yung et al., 1981; Yung and Moore, 1982; Metcalf

and Kelso, 1985), and B lymphoma cells (Clark-Lewis et al.,


The biologically active factor in lymphocyte and WEHI-

3 conditioned media has been given a variety of names. It

was first termed "multi-CSF" by Burgess and colleagues

(1980), due to its ability to support the differentiation

of multiple hematopoietic lineages in vitro. Schrader and

colleagues (Schrader and Nossal, 1980; Schrader, 1981;

Schrader et al., 1981; Clark-Lewis and Schrader, 1981)

reported that "P cell stimulating factor" (PSF), which

supported the growth of persisting, mast-like cells in

vitro, was present in concanavalin A spleen conditioned

media. Ihle and colleagues (1981, 1982) proposed the name

"interleukin 3", or EL 3, for the factor which induced the

enzyme 20-alpha-hydroxysteroid dehydrogenase in nude mouse

spleen cells as well as effecting a number of

differentiative and supportive activities in multiple cell

lineages. Although Ihle's terminology has achieved the

greatest usage in recent literature, alternative

nomenclature for the same activity has been proposed by

Bazill and colleagues ("multi-hematopoietic cell growth

factor", MCGF; Bazill et al., 1983) and Iscove

("multilineage hematopoietic growth factor", multi-HGF;

Iscove, 1985).

interleukin 3, derived from a variety of sources, has

been purified to homogeneity (Yung and Moore, 1982; Ihle et

al., 1982b; Bazill et al., 1983; Clark-Lewis et al., 1984)

and, more recently, the genes for IL 3 have been molecularly

cloned and expressed (Fung et al., 1984; Yokota et al.,

1984; Rennick et al., 1985). Extensive reviews of the

literature describing interleukin 3 (per se, and the

related activities called by various other names) have been

published recently and should be consulted for additional

information (Clark-Lewis et al., 1985; Ihle, 1985; Iscove,

1985; Schrader et al., 1985; Whetton et al., 1985; Yung and

Moore, 1985).

Interleukin 3-Independent Mast Cells

Several reports of interleukin 3-independent mast cells

have appeared in the literature in recent years.

Schrader's group (Schrader and Crapper, 1983; Schrader et

al., 1983a) observed the emergence of factor-independent

variants from factor-dependent cells. In one experiment,

factor-dependent ceils were plated in agar in the absence

of exogenous interleukin 3. From these cultures, several

colonies of autonomous culture-derived mast cells (P cells)

arose which, after several weeks, were subsequently adapted

to growth in interleukin 3-free liquid media. The

autonomous colonies secreted interleukin 3 into the culture

media, but also retained their receptors for the factor.

Furthermore, the autonomous cells generated more colonies

when plated at low density in the presence of exogenous

interleukin 3 in agar than in the absence of the growth

factor. It seems likely that the autogenouss" cells were

not truly factor-independent, but rather were variants

which were able to proliferate, albeit at lower efficiency,

at the low levels of interleukin 3 provided by an autocrine


Similar factor-independent culture-derived mast cells

have been reported by a second group (Ball et al., 1983;

Conscience and Fisher, 1985). Several long-term bone

marrow-derived cultured mast cell lines were found, after

eleven months in culture, to contain variants which

proliferated at higher rates than similar cultures in the

presence of interleukin 3. The more proliferative cells

were able to continue cell growth in the absence of

exogenous interleukin 3, but the doubling time was

increased 160 percent when compared to cells of the same

line maintained in conditioned media. In contrast to the

autogenous cells of Schrader, the media conditioned by the

factor-independent mast cells described in the more recent

studies failed to support the growth of other factor-

dependent culture-derived mast cells. It is possible,

however, that the cells were able to maintain their growth

in the presence of interleukin 3 at levels below those

detected in the assay. Interestingly, both sets of factor-

independent mast cells were tumorigenic in syngeneic mice.

Although no retroviral particles were observed in one of

these cell lines (Ball et al., 1983), the observations are,

by the criteria of tumorigenicity and factor independence,

similar to those of a recent report of Abelson murine

leukemia virus-transformation of culture-derived mast cells

(Pierce et al., 1985). Two possible mechanisms could

reconcile the yet unexplained results. First, the

activation of a latent replication-defective viral genome

in the long-term culture-derived mast cells of Ball and

colleagues could be the missing link. Under such

circumstances, no viral particles would be detected, but

the cells could become both factor independent and

tumorigenic by virtue of the viral transforming gene

product. Similarly, the activation of a cellular homologue

of a viral transforming gene, like c-abl (the normal

function of which is unknown), could be invoked to activate

the mechanisms necessary to generate the phenotype of the

factor-independent, tumorigenic mast cells.

Mast Cell Precursors

The development of in vitro techniques for the

differentiation and maturation of mast cells from

phenotypically immature progenitors permitted the analysis

of mast cells based upon a number of criteria. First, the

number of culture-derived mast cell precursors in

particular tissues was determined quantitatively. Prior to

the development of mast cell culture techniques which

utilized sources of interleukin 3 for mast cell

proliferation, Pluznik and Sachs (1965) enumerated mast

cell clones in soft agar with feeder layers. The authors

reported approximately thirty mast cell colony forming

units per million spleen cells seeded, while the frequency

of culture-derived mast ceil precursors in embryos, thymus,

and lymph nodes were five per million, three per hundred

million, and less than one per fifty million cells,

respectively. Schrader and colleagues (1981) cloned mast

cell precursors from mouse bone marrow in soft agar with

WEHI-3 conditioned media and found thirty to two-hundred

progenitors per million Thy 1-negative cells. These

results were contradicted by a more recent report (Sredni

et al., 1983) which found 600 to 700 bone marrow

precursors, 400 to 500 spleen precursors, 25 to 30 thymus

precursors, and 100 to 200 lymph node precursors per

million seeded cells. The latter results, however, were

generated in a system using concanavalin A-activated

splenocyte conditioned media as a source of interleukin 3

and different strains of mice, thereby making comparison

difficult. Interestingly, the latter authors also observed

that athymic nude mice had as many mast cell precursors as

syngeneic wild type controls. Since athymic mice do not

exhibit the profound mucosal mastocytosis found in

appropriate controls (Olson and Levy, 1976), the last

observations could be interpreted to indicate that athymic

mice lack the inductive conditions required for the

proliferation of seemingly normal numbers of mast cell


Mast cell precursors were enumerated by Nakahata and

colleagues (1982), using a modification of the semisolid

methylcellulose media culture system previously used to

identify erythroid and myeloid precursors. In the latter

and subsequent report using this system (Pharr et al.,

1984), which used pokeweed mitogen activated spleen

conditioned media for a source of interleukin 3, the

investigators found between twenty and 140 mast cell

precursors per million BDF1 mouse spleen cells and 200 mast

cell precursors per million bone marrow cells. Suda and

colleagues (1985), using the Nakahata culture system,

demonstrated that W/Wv mice, which are severely deficient in

mast cells of both serosal and mucosal subsets, had the

same number of peripheral blood mast cell precursors as

wild type mice (approximately thirty per million nucleated

cells), thus indicating that the mast cell defect was in a

homing or developmental step, rather than at the stem cell

or migratory level.

A third method for the enumeration of culture-derived

mast cell precursors from various tissues was reported by

Crapper and Schrader (1983). Using limiting dilution

analysis of cells in liquid culture containing WEHI-3

conditioned media, The authors were able to enumerate mast

cell precursors in bone marrow, spleen, mononuclear

peripheral blood cells, and lymph nodes. All of the data

recorded in the latter experiments concurred with the

previously cited results of Schrader and colleagues (1981)

as well as those reported by the Nakahata group (Nakahata

et al, 1982b; Pharr et al., 1984). Furthermore, Crapper

and Schrader were able to substantiate the findings of Suda

and colleagues that mast ceil deficient mice had similar

numbers of culture-derived mast cell precursors (in bone

marrow and spleen) when compared to appropriate wild-type

controls, although the former authors used wf/Wf and the

latter authors used W/Wv.

The development of techniques for the propagation of

culture-derived mast cells has also permitted the

characterization of such cells at various stages of

differentiation. Thus, Ginsburg and colleagues (Ginsburg,

1963; Ginsburg and Sachs, 1963; Ginsburg and Lagunoff,

1967; Ginsburg et al., 1982; Davidson et al., 1983)

reported a progression of characteristics of cultured mast

cells, starting with large, mononuclear "stem" cells.

After six to ten days in culture, large, lymphocyte-like
"mastoblasts" with round and bilobed nuclei and a narrow

rim of metachromatic cytoplasm, similar to those described

by Maximow (1906) in stained tissue sections, arose and

became dominant in cell culture. The cytoplasm of the

monoblasts continued to increase in size as the nucleus

became more distinctly indented, with a chromophobic region

in the concave aspect of the nucleus. At twelve to

thirteen days in culture, a foamy region in the cytoplasm

was seen to spread, sometimes encompassing the entire

cytoplasm. Metachromatic material first appeared in the

foamy region as faint, amorphous substance in vacuoles,

later increasing in size and staining intensity to well-

defined granules. These latter cells were described as

"young mast cells", possessing granular, metachromatic

cytoplasm, with actively mitotic, kidney-shaped, round or

oval nuclei; such ceils dominated the cultures between days

twelve and twenty-two. After three weeks in culture, the

majority of the cells were mature, round mast cells with

round to oval, eccentric or centered amitotic nuclei and

abundant, metachromatic cytoplasmic granules, similar in

morphology and histochemistry to mucosal mast cells.

Similar staging of rat culture-derived mast cell precursors

and intermediates has been reported (Ishizaka et al., 1976;

Zucker-Franklin et al., 1981; Sterry and Czarnetzki, 1982;

Czarnetzki et al, 1983), corroborating in situ observations

(Combs et al.. 1965).

Characteristics of Culture-Derived Mast Cells

The development of refined analytical methods, first

applied to other hematopoietic cells, aided in the

characterization of culture-derived mast cells. Yung and

colleagues (1983) analyzed bone marrow cells by

centrifugation techniques and reported that interleukin 3-

responsive cells could be isolated by nature of their

median buoyant density (1.033) from interleukin 2-

responsive cells (1.075). The authors also noted that the

buoyant densities of long-term bone marrow-derived cultured

mast ceils (1.062 to 1.095 g/ml) were similar to those

determined by Pretlow and Cassidy (1970), who analyzed

heterogeneous populations of freshly isolated peritoneal

mast cells and reported that immature mast cells have a

median buoyant density of 1.087. Interleukin 3-

responsive cells were separated by sedimentation velocity

analysis from the in vitro precursors of macrophages (CFU-

M) and the pluripotent colony-forming cell of the spleen

(CFU-S), but not the bipotent precursor of granulocytes and

macrophages (CFU-GM).

Culture-derived mast cells and their precursors have

been characterized for the expression of a broad variety of

hematopoietic differentiation markers in attempts to assign

them to a particular lineage. Bone marrow and intestinal

precursors to culture-derived mast cells were observed to

lack the T lineage antigens Thy 1, Lyt 1 and Lyt 2

(Schrader et al., 1983b; Guy-Grand et al., 1984). Like

their precursors, more differentiated culture-derived mast

ceils are deficient in Thy 1, Lyt 1, and Lyt 2 ( Ginsburg

et al., 1981; Nabel et al., 1981; Schrader, 1981; Schrader

et al., 1981; Tertian et al., 1981; Davidson et al., 1983;

Sredni et al., 1983; Ghiara et al., 1985), although

Schrader and colleagues (1982) reported that Thy 1 may be

transiently expressed on these cells. Similarly, culture-

derived mast cells lack surface immunoglobulin, a B lineage

marker (Ginsburg et al., 1981; Schrader et al., 1981;

Tertian et al., 1981; Sredni et al., 1983), NK-1, a marker

of natural killer cells (Nabel et al., 1981), complement

receptors and MAC-1, a differentiation antigen of

mononuclear phagocytes which is also expressed by some

natural killer and T lymphoma cells (Tertian et al., 1981).

Like their in vivo correlates, murine culture-derived

mast cells express surface receptors for IgE (Ginsburg et

al., 1978: Nagao et al., 1981; Schrader, 1981; Schrader et

al., 1981; Tertian et al., 1981; Ginsburg et al., 1982;

Nakahata et al., 1982b; Sredni et al., 1983; Wedling et

al., 1983, 1985), which induce the anaphylactic release of

histamine when cross-linked by IgE and homologous antigen

(Ginsburg et al., 1978; Sredni et al., 1983) or anti-IgE

(Ginsburg et al., 1982). The number of IgE receptors per

cell has been estimated to be 2 to 3 x 105, similar to the

number on serosal and mucosal mast cells (Razin et ai.,

1981a; Ginsburg et al., 1982). Some investigators have

also noted the presence of receptors for IgG (Schrader,

1981; Tertian et al., 1981); the cells, however, did not

phagocytose opsonized or unopsonized targets (Schrader,

1981; Sredni et al., 1983). Liquid cultured mast cells in

conconavalin A-stimulated spleen conditioned media also

express the lymphocyte marker Ly 5 (Nabel et al., 1981;

Tertian et al., 1981). Contradictory observations of

histocompatability Class II (Ia) antigens, or the lack

thereof, on culture-derived mast cells were resolved by

Wong and colleagues (1982), who showed that such cells,

grown in the presence of immune interferon (interferon

gammma, found in the supernatants of concanavalin A-

stimulated splenocytes) expressed the marker while cells

grown in the absence of interferon (as in WEHI-3

conditioned media) were devoid of Ia. Culture-derived mast

cells were also shown to express Class I histocompatability

antigens and receptors for peanut agglutinin (Schrader,

1981; Schrader et al., 1981; Tertian et al., 1981).

Russell and colleagues developed a panel of rat

monoclonal antibodies against murine mononuclear phagocytes

which could discriminate between culture-derived mast ceils

and connective tissue mast cells (Leblanc et al., 1982).

The same group observed that culture-derived mast cells

expressed the phenotype Bl.1-/B23.1+/ B54.2+, while in

contrast peritoneal mast cells were B1.1+/ B23.1-/B54.2+

(Katz et al., 1983). Although the Forsmann glycolipid

recognized by monoclonal antibody B1.1 is undetectable on

culture-derived mast cells, the latter cells do express the

antigen precursor, globotetrasylceramide (Katz et al.,

1985b), and may therefore be deficient or defective in the

glycosyltransferase required for the synthesis of the

mature antigen.

Culture-derived mast cells have been extensively

characterized biochemically as well. The histamine content

of cultured mast cells, like that of mucosal mast cells,

has been estimated between 450 and 500 nanograms per

million cells, at least ten-fold less than the histamine

content of comparable numbers of peritoneal mast cells

(Nabel et al., 1981; Nagao et al., 1981; Razin et al.,

1981a; Galli et al., 1982b; Sredni et al., 1983; Wedling et

al., 1985). Mouse culture-derived mast cell (and mucosal

mast cell) granules stain blue when treated with alcian

blue and safranin (Ginsburg and Lagunoff, 1967), indicating

the presence of weakly sulfated mucopolysaccharides,

whereas serosal mast cell granules, containing strongly

sulfated heparin proteoglycan, stain red. Razin and

colleagues (1982c) analyzed the proteoglycan of murine

culture-derived mast cells and found they contained

glucuronic acid-N-acetylgalactosamine-4,6-disulfate, or

chondroitin sulfate proteoglycan E, a unique

glycosaminoglycan which could not be detected in basophilic

leukemia cells, peritoneal mast cells, or chondrocytes.

Chondroitin sulfate proteoglycan E was shown in this study

to be chemically distinct from heparin by a number of

criteria including sensitivity to enzymatic degradation and

molecular weight (chondroitin sulfate proteoglycan E has an

estimated molecular weight of 200 kilodaltons, in contrast

to heparin, which has a molecular weight of 750

kilodaltons). These results have been confirmed in the

literature (Razin et al., 1983; Sredni et al., 1983).

Mouse culture-derived mast cell granules also contain a

number of other in vivo mast cell-associated biological

mediators, including serotonin (5-hydroxytryptamine) and

dopamine (Tertian et al., 1981).

Arachidonic acid metabolites, the prostaglandins and

leukotrienes, are important biological mediators

associated with metachromatic cells. IgE-dependent

activation of mouse culture-derived mast cells results in

the synthesis and release of leukotriene C4 (Razin et al.,

1982b, 1983), a component of the slow releasing substance

of anaphylaxis (Austin, 1984). These studies also showed

that bone marrow-derived cultured mast cells generated

approximately twenty-five times more leukotriene C4 than

prostaglandin D2 upon activation by calcium ionophore

A23187 or IgE receptor-mediated pathways. In contrast, rat

peritoneal mast cells preferentially synthesized and

released prostaglandin D2 in forty-fold excess over

leukotriene C4.

Culture-derived mast cells are ultrastructurally

distinct from serosal mast cells. Cells of the former

category thus possess granules which are more heterogeneous

in size and electron density than the latter (Ginsburg and

Luganoff, 1967; Ginsburg et al., 1978; Nabel et al., 1981;

Galli et al., 1982b; Wedling et al., 1985). Granules in

culture-derived mast cells are ovoid and are frequently

associated with small vessicles (Sredni et al., 1983) and a

well-developed Golgi apparatus (Ginsburg and Lagunoff,

1967). The substance of mouse culture-derived mast cell

granules is crystalline (Razin et al., 1982a), while that

of human origin (resembling basophils, rather than mast

cells) is more particulate (Razin et al., 1981b). Mouse

bone marrow-derived mast cells form membrane channels after

activation with IgE and anti-IgE through which granules may

reach the cell surface (Razin et al., 1982a), similar to

human lung mast cells (Caulfield et al., 1980). The

cytoplasmic membrane of culture-derived mast cells is

characterized by numerous, fine protrusions (Galli et al.,

1982b; Wedling et al, 1985) which are absent from

peritoneal mast cell membranes.

The response of culture-derived mast cells to

secretagogues is similar to that of mucosal mast cells, but

not serosal mast cells. As previously discussed,

leukotriene C4 is synthesized and released as a result of

activation of the IgE receptor-mediated pathway. In

addition, histamine, chondroitin sulfate proteoglycan E,

and beta-hexosaminadase are released by immune activation

(Razin et al., 1983). Like both serosal and mucosal mast

cells, culture-derived mast cells are induced to

degranulate by the calcium ionophore A23187 (Razin et al.,

1982a, 1982b; Sredni et al., 1983; Robin et al., 1985).

Culture derived mast cells, however, mimic mucosal mast

cells in their lack of response to compound 48/80 (Sredni

et al, 1983). In contrast, mouse peritoneal mast cells are

degranulated by compound 48/80.

Many of the characteristics previously described for

mouse culture-derived mast cells have been reported in

analogous rat and human systems. Ishizaka and colleagues

(1976) cultured rat thymocytes in the presence of rat

embryonic fibroblast monolayers and observed the outgrowth

of cells with receptors for IgE and metachromatic granules.

Haig and colleagues (1982, 1983) grew rat bone marrow-

derived cultured mast cells in the presence of media

conditioned by mesenteric lymph node cells. The

investigators observed that, similar to culture-derived

murine mast cells, rat culture-derived mast cells were

smaller than peritoneal mast cells, possessed sparse

granules of heterogeneous size, and expressed surface

receptors for IgE. Rat culture-derived mast cell granules

stained blue by the alcian blue-safranin technique and were

metachromatic when stained with toluidine blue. The

granule proteoglycan was identified as non-heparin,

although no report of its precise chemical composition has

been published to date. Such cells also contained

immunochemically detectable levels of rat mast ceil

protease II, which was previously described as a marker of

mucosal, but not serosai mast cells. Rat mesenteric lymph

nodes (Denburg et al., 1980) and peripheral blood (Zucker-

Franklin et al., 1981) have also been shown to contain

precursors of culture-derived mast cells.

Reports of mast cells derived from human tissues are

clouded by difficulties in distinguishing between the

various types of basophilic cells when compared to in vivo

correlates, namely mucosal mast ceils, connective tissue

mast cells, and basophils. Granulated cells with receptors

for IgE and low levels of histamine (50 to 450 nanograms

per million cells), thus resembling mouse culture-derived

mast cells, have been observed in cultures of human fetal

liver grown in unconditioned media (Razin et al., 1981b).

Adherent human blood mononuclear ceils and pleural exudate

cells, which were propagated with L-ceil conditioned media,

exhibited similar characteristics (Czarnetzki et al., 1983,

1984; Kruger et al., 1983). Ceils with high affinity

receptors for [gE and slightly higher levels of histamine

(480 to 1600 nanograms per million cells) were cultured

from human umbilical cord blood grown in phytohemagglutinin

A-stimulated human T cell conditioned media (Ogawa et al.,

1983). Horton and O'Brien (1983) reported the culture of

granulated cells with centrally placed, round or indented

nuclei from human bone marrow harvested from a patient with

systemic mastocytosis. The culture-derived mast cells

showed no growth advantage with a number of conditioned

media, but required an adherent, bone marrow-derived feeder

layer to persist.

Other In Vitro-Derived Metachromatic Cells

As noted previously, some of the confusion in the

terminology applied to culture-derived mast ceils has been

generated by the appearance of characteristics of one or

more of the in vivo basophilic cell correlates. Thus,

although basophils in mice are lacking (Lagunoff and Chi,

1980) or extremely rare (Dvorak et al., 1982), a report of

cloned, basophil-like cell lines with IgE receptors has

appeared in the literature (Galli et al., 1982a). The cell

line in question, derived from mouse splenocytes cultured

in concanavalin A-stimulated splenocyte conditioned media,

lacked histamine and had both natural killer cell

differentiation markers and natural killer activity.

In a series of studies on the rat, Czarnetzki and

colleagues described the growth of connective tissue-like

mast cells in vitro. In the earliest study (Czarnetzki et

al., 1979), mast cell-free peritoneal exudate cells were

harvested from rats which were previously injected with

sterile water (intraperitoneally). The peritoneal cells

were cultured in L-cell conditioned media with sodium

butyrate. The mast cells which grew out of this

population, although initially possessing blue-staining

granules by the alcian biue-safranin technique, later had

red staining granules and released histamine in response to

compound 48/80. Despite their serosal mast cell

characteristics, these culture-derived mast ceils were

similar to the classical description of mucosal mast cells

in that they contained low levels of histamine (500

nanograms per million cells) and survived for only short

periods. Similarly described cells were subsequently

isolated by the same group from rat peritoneal cells

(Czarnetzki and Behrendt, 1981), and rat mononuclear

phagocytes (Czarnetzki et al., 1981, 1982; Sterry and

Czarnetzki, 1982).

In related studies, in vitro cultivated human

peripheral blood mononuclear cells (Denburg et al., 1983)

and guinea pig bone marrow ceils (Denburg et al., 1980)

developed into metachromatic cells with segmented nuclei

which were more characteristic of basophils than mast

cells. The human cells, in particular, possessed the

polymorphonuclear structure with mature chromatin, Golgi

and microtubules which are more characteristic of basophils

than mast cells. This conclusion, however, directly

disagreed with Zucker-Franklin (1980), who contended that

human mast cells and basophils share common ultrastrucural


Tadokoro and colleagues (1983) cultured cells with

metachromatic granules and lobulated nuclei from normal

human bone marrow in conditioned media from lectin-

stimulated blood lymphocytes. The culture-derived cells

contained 500 to 2000 nanograms of histamine per million

cells and were responsive to IgE-anti-IgE- and calcium

ionophore-mediated histamine release but were refractory to

the effects of compound 48/80. The authors concluded that

their conditioned media contained a basophil-promoting

activity which furthermore had a molecular weight of 25 to

40 kilodaltons and was distinct from interleukin 2. The

same group recently reported that media conditioned by

phytohemagglutinin A- and concanavalin A-stimulated human

blood lymphocytes could support the interleukin 3-dependent

mouse cell line 32Dci as well as promote the growth of

human culture-derived basophils (Stadler et al., 1985).

Furthermore, the interleukin 3 and basophil-promoting

activities, which were also found in media conditioned by

the growth of E-rosetting T lymphocytes and the MoT cell

line, were biochemically distinct by at least five

different criteria. The isolated human interleukin 3 was

shown to promote the growth of mast cells which were unable

to proliferate in the presence of mouse interleukin 3 (in

WEHI-3 conditioned media). Thus, in the human system, the

culture-derived mast cell-basophil dilemma is no longer

solely a matter of terminology and mistaken identity, but,

in fact, appears to involve multiple growth promoters and,

most likely, multiple progenitor cells.

Relationship of In Vivo- and In Vitro-Derived Mast Cells

A small body of evidence supports the theory that

culture-derived mast cells are more than circumstantially

related to mucosal, and perhaps serosal, mast cells. A

number of characteristics, including morphology,

histochemical fixation and staining, dependence of

proliferation on T cell-derived factors, biogenic amine

content, protease content, presence of receptors for IgE,

and sensitivity to secretogogues, have been noted in this

review and cited by many of the authors as proof of the

relationship between culture-derived mast cells and the

mast cells of the mucosal surfaces. There is preliminary

evidence that culture-derived mast cells have natural

cytotoxic activity against tumors such as WEHI-3 and Meth A

which is enhanced by interleukin 3 (Ghiara et al., 1985).

Investigators have noted that in vivo-derived mast cells

exhibited similar tumoricidal activity (Farram and Nelson,

1980) and that in vivo-derived cells with the

Thy l-/Lyt 1-/ Lyt 2- phenotype, which are demonstrably

cytotoxic, are sensitive to the proliferative activities of

interleukin 3 (Djeu et al., 1983; Lattime et al., 1983).

Evidence of more direct relationships between in

vitro-derived mast cells and their in vivo correlates has

been elusive. Several investigators have associated the

high incidence of culture-derived mast cell precursors and

mucosal mast cells in the intestine of normal mice (Crapper

and Schrader, 1983; Guy-Grand et al., 1984). The role of

antigenic stimulation and T cell function in the

proliferation of mucosal mast cells per se has been

thoroughly described in the literature (for reviews see

Jarrett and Haig, 1984; Shanahan et al., 1984; Bienenstock

et al., 1983). Guy-Grand and colleagues (1984) also showed

that the number of mast cells which could be cultured from

intestinal mucosa increased with antigenic stimulation and

WEHI-3 tumor burden, implicating the role of interleukin 3

in the in vivo proliferation of mast cell precursors. The

results, however, associated the in vivo and in vitro mast

cell precursors by existence in the same tissue, and did

not directly show that the populations involved were


The most suggestive evidence to date of the

relationship between in vivo- and in vitro-derived mast

cells involves the demonstration that culture-derived mast

cells, when injected into mast cell-deficient mice,

populated both mucosal and connective tissue-serosal

compartments. Nakano and colleagues (1985) injected

culture-derived mast cells and partially purified (30 to 40

percent) peritoneal mast cells into W/Wv mice by various

routes. At high levels of inoculum (105 to 106 cells),

intravenously- or intraperitoneally-injected cultured mast

cells populated the spleen and stomach mucosaa and muscle),

while at lower levels of inoculum (102 to 104cells), only

intraperitoneallv injected cells were able to populate the

same anatomical sites. Both cultured and peritoneal mast

cells were relatively inefficient at populating the skin,

however, possibly due to the presence of mast cells, or

their precursors, which were already present in the skin

(Kitamura et al., 1977). An interesting note to the Nakano

studies was the evolution of mucosal or serosal mast cell

characteristics from injected cells (regardless of origin)

depending on the anatomical site of subsequent lodging.

Thus, the granules of mast cells isolated from the

peritoneal cavity, spleen, skin, and gastric muscularis

propria of reconstituted animals stained preferentially

with safranin (which has an affinity for highly sulfated

mucopolysaccharides like heparin) and the fluorescent dye

berberine sulfate (which also binds to heparin), and were

ultrastructurally homogeneous in size and electron density,

while mast ceils identified in the glandular stomach mucosa

stained preferentially with alcian blue and were unstained

by berberine sulfate. Mast cell phenotype, therefore, may

be functionally regulated at the level of the tissue

microenvironment in which a multipotent mast cell precursor

or intermediate develops.


Despite an apparent wealth of literature available on

the subject, the potential still exists for scholarly.

significant contributions to the body of knowledge which

describes mast cells. The lineage relationship between the

mast cells found in vivo mucosall and serosal) is still

poorly defined, ond only recently have preliminary studies

approached the relationship between the aforementioned

cells and their putative correlate, the culture-derived

mast cell. Little is known of the phenotype of the cells

which give rise to mast cells in culture, and the ontogeny

of the mast cell in early embryonic tissues is documented

in scant and unsystematic reports.

In the course of the remaining chapters of this

dissertation, we describe our recent contributions to the

study of the mast cell. Beginning with the observation of

cell lines with basophilic granules, we have characterized

Abelson murine leukemia virus-transformed mast cell-like

lines of midgestational, embryonic origin using panels of

monoclonal antibodies as well as biochemical and molecular

biological techniques (Chapter II). Although mast cells

were not detected in homologous, uninfected tissues,

culture-derived mast cells could be propagated from

embryonic sources in the presence of exogenously supplied

interleukin 3. These studies (and parallel experiments on

adult bone marrow-derived mast ceils) also provide the

first detailed analysis of hematopoietic marker expression

of cultures progressing from heterogeneous to homogeneous

populations of mast cells (Chapter III). We have

subsequently analyzed the frequency of mast cell precursors

in embryonic placenta, nonplacental embryonic tissues, and

adult tissues, demonstrating the earliest reported mast

cell precursors as well as a heretofore unreported rich

source of such cells, the placenta (Chapter IV). In the

same chapter, we have characterized the cell surface of

mast cells grown in semisolid agar media and have presented

encouraging preliminary results of experiments designed to

sort mast cell precursors on the basis of differentiation

antigen expression.



Abelson murine leukemia virus (A-MuLV) is a

replication-defective transforming retrovirus which was

isolated from a tumor in a steroid-treated BALB/c mouse

inoculated with the Moloney murine leukemia virus (Abelson

and Rabstein, 1970). Molecular analysis of the A-MuLV

genome has revealed that the virus arose by recombination

between the thymotropic Moloney virus genome and a cellular

gene termed c-abl (Goff et al., 1980; Shields et al.,

1979). The recombinant virus can infect and immortalize

hematopoietic cells in vivo and in vitro, and can transform

certain fibroblast cell lines in vitro (Scher and Siegler,

1975). The virus has demonstrated the ability to transform

in vivo mature cells of the B lineage (Potter et al., 1973;

Premkumar et al., 1975) as well as those of the T (Cook,

1982), myelomonocytic (Raschke et al., 1978; Ralph et al.,

1976), and mast cell (Risser et al., 1978; Mendoza and

Metzger, 1976) lineages. In contrast, in vitro infection

of adult or embryonic primary hematopoietic tissues

followed by clonal selection in semisolid media results in

the immortalization of cells exhibiting characteristics of

the earliest stages of B lymphoid differentiation

(Rosenberg et al., 1975; Rosenberg and Baltimore, 1976b;

Boss et al., 1979; Siden et al., 1979; Alt et al., 1981).

Abelson virus can also induce agar colony-forming ceils

which express erythroid characteristics (Waneck and

Rosenberg, 1981); the latter cells, however, fail to

proliferate as permanent cell lines in liquid culture.

Based upon Abelson virus's propensity to immortalize B

lineage precursors in vitro, experiments were designed to

study early embryonic lymphoid precursors. The

midgestation embryonic placenta has been reported to be the

earliest source of B cell precursors (Melchers and

Abramczuk, 1980; Melchers, 1979) in the mouse. The

successful development of permanent cell lines from

midgestation embryonic placenta transformed in vitro by A-

MuLV has recently been reported (Siegel et al., 1985).

Primary agar colony counts indicated that the frequency of

A-MuLV targets is highest at ten days of gestation. Unlike

previously reported A-MuLV-transformed embryonic cell

lines, the genomes of the placental cells contain a

germline immunoglobulin heavy chain locus characteristic of

nonlymphoid cells and perhaps very immature lymphoid cell


We proceeded to analyze this novel group of A-MuLV

embryonic transformants to better ascribe them to cells of

a particular lineage. This chapter summarizes our efforts

to characterize the A-MuLV transformants derived from

midgestation embryonic tissues and presents several

significant observations. First, unlike previously

reported Abelson virus-transformed embryonic cells, the

placental cell lines isolated in our laboratory exhibit

many characteristics of culture-derived and mucosal mast

cells including differentiation antigens, high affinity

receptors for igE, and metachromatic granules containing

histamine and sulfated proteoglycans. Second, we describe

the development of a simple, sensitive, nonisotopic,

nonfluorometric method to detect membrane receptors for

immunoglobulins. Third, all of the cell lines analyzed had

at least one, and sometimes more than one integrated A-MuLV

provirus. Finally, the cell lines proliferated independent

of exogenous mast cell growth factor, and no growth factor

could be detected in either cell lysates or the

supernatants of exponentially growing cultures maintained

at high cell density.

Materials and Methods

Cell Lines

Cloned A-MuLV-transformed embryonic cell lines were

established in our laboratory as previously reported

(Siegel et al., 1985). All had been adapted to growth in

liquid culture and had been successfully maintained in 10P

(RPMI 1640 (GIBCO, Grand Island, NY) with 5x10-5 M 2-

mercaptoethanol (Sigma Chemical Co., St. Louis, MO)) and 10

percent heat-inactivated fetal bovine serum (Sterile

Systems, Logan, UT)) for at least two years with twice

weekly passages prior to analysis. The nomenclature used

to designate each cell line included the number of days of

gestation (detection of vaginal plug on day 0) followed by

a suffix ("P" for placenta of (BALB/cAN x B10.BR/SgSn)Fl

origin, "PC" for placenta of (BALB/cAN x CBA/J) origin, and

a clone number. Cell line 10P12, for example, was the

twelfth clone isolated from 10 day placental cells derived

from matings of BALB/cAN females and B10.BR males.

Mouse tumor cell lines used as experimental controls in

the studies were chosen to represent the major

hematopoietic lineages:

1. B lymphoid

a. FLEI-4, an pre-B cell line derived by E.J.

Siden by infection of day 15 BALB/c fetal

liver with A-MuLV strain P120;

b. 18-81 (Siden et al., 1979), a pre-B cell line

induced by infection of bone marrow ceils with

A-MuLV strain P120;

2. T lymphoid

a. RLd 11, a radiation-induced leukemia

obtained from N. Rosenberg;

b. B2-4-4, a Moloney virus-induced leukemia

obtained from N. Rosenberg;

c. YAC-1, a Moloney leukemia virus-induced

lymphoma obtained from R. Weiner (University

of Florida);

3. Monocyte-macrophage

a. WEHI-3, a myelomonocytic leukemia obtained

from M. Norcross (University of Florida);

b. P388D1, a methylcholanthrene-induced monocytic

tumor obtained from American Type Culture

Collection (Rockville, MD);

4. Basophil-mast cell

a. P815, a methylcholanthrene-induced mastocytoma

obtained from S. Noga (University of Florida);

b. CB6ABMC4, an A-MuLV-induced mastocytoma

obtained from M. Potter (National Cancer

Institute, Bethesda, MD);

c. BALABMC20, an A-MuLV-induced mastocytoma

obtained from M. Potter.

All of the above cell lines were maintained in 10P with

twice-weekly passage.

Cell line DA-1, which was used in the analysis of

interleukin 3 production, was generously supplied by J.N.

Ihle (National Cancer Institute, Frederick, MD) and was

maintained in a one-to-one mixture of WEHI-3 conditioned

media and enriched media (EM; Razin et al., 1984a).

Analysis of Histamine Content

Histamine content was determined by a modification of

the enzymatic-isotopic microassay of Taylor et al. (1980).

Histamine methyltransferase (HMT) reagent was prepared from

BALB/cAN mouse brain which was homogenized in iced 5mM

sodium phosphate buffer, pH7.9 (10 ml per gram of brain)

with an iced, sintered glass homogenizer. The crude

homogenate was transferred to an Oak Ridge type tube and

cleared centrifugally at 50,000xg for ten minutes at 4C.

Histamine standard solutions and tumor cell lysates were

prepared in PIPES-BSA which contained 25mM PIPES, pH 7.4

(Sigma Chemical Co., St. Louis, MO), 0.4 mM magnesium

chloride, 5 mM potassium chloride, 120 mM sodium chloride,

1 mM calcium chloride, 5.6 nm glucose (Fisher Scientific

Co., Fairlawn, NJ), 0.1 percent BSA (Sigma); standard

solutions and cells were boiled for ten minutes and cleared

for five minutes in a microfuge (Brinkmann Instruments,

Westbury, NY). Immediately before starting the reaction,

HMT-SAMe was formulated: 500 microliters of HMT reagent

was mixed with 10 microliters (5 microcuries) S-[methyl-

3H]-adenosylmethionine (76.1 Ci per mMol; New England

Nuclear, Boston, MA) and 5 microliters unlabeled S-

adenosylmethionine (SAMe, Sigma Chemical Co.; 16 micrograms

per ml). Each reaction mixture consisted of 10 microliters

of HMT-SAMe and 20 microliters of histamine standard

solution or tumor cell lysate. Following incubation for

two hours at 37C, each reaction was stopped with 0.1 ml of

1 M sodium hydroxide and saturated with solid sodium

chloride. The contents of each tube were extracted with

0.25 ml chloroform in the presence of 10 microliters of

unlabeled SAMe (16 micrograms per ml)and the phases were

separated by brief centrifugation. The aqueous phase was

carefully removed with a pasteur pipet and the organic

phase was re-extracted with an equal volume of 1 M sodium

hydroxide and unlabeled SAMe. Aliquots (150 microliters)

of the organic phase were added to toluene-based

scintillation fluid (10 ml) and counted on a Beckman liquid

scintillation counter (Beckman Instruments, Palo Alto, CA).

All samples and standards were prepared in duplicate or

triplicate. Histamine content was determined by comparison

of tritiated counts in dilutions of tumor lysates to those

in histamine standard solutions and were expressed as

nanograms (ng) per 104 cells.

Cytological Staining

Staining procedures were performed on air dried

cytocentrifuge (Shandon Scientific Company, Ltd., London,

U.K.) preparations on 25x75 mm precleaned glass slides.

Slides were fixed in methanol for one to two minutes prior

to staining with either Wright's Giemsa (Schalm et al.,

1975) or May-Gruenwald Giemsa (Thompson and Hunt, 1966).

Metachromatic cell granules were identified by staining for

five minutes with 0.1% w/v toluidine blue in 30% v/v

ethanol, pH 0.5 after fixation for two minutes in Mota's

fixative (lead subacetate in acidic ethanol)(Yam et al.,

1971). All stained slides were coverslipped with Permount

(Fisher Scientific Co.) prior to observation by light



Lineage-specific determinants on the surfaces of

placental and control cell lines were probed with a panel

of monoclonal (mAb) and polyvalent antibodies whose

significant features are summarized in Table II-l. All

monoclonal antibody preparations, unless otherwise noted,

were used as filtered hybridoma supernatants from

stationary phase cultures grown (in our facility) in

Dulbecco's Modified Eagle's Minimal Essential Media (GIBCO)

with 10 percent heat-inactivated fetal bovine serum and

5x10-5 M 2-mercaptoethanol. Preliminary work with mAb

B23.1 was performed with partially purified antibodies

generously supplied by P. LeBlanc (University of Florida);

later studies were performed with filtered hybridoma

supernatants which had been grown in RPMI 1640 with 10

percent fetal bovine serum, generously supplied by G. Place

in the laboratory of S. Russell (University of Florida).

Monoclonal antibody FT-1 was generously supplied by M.

Kasai (National Institute of Health, Tokyo, Japan); the

nature of its production and processing have been

previously reported (Kasai et al., 1983).

Table II-1.


Lineage-Specific Antibodies Used in Surface
Marker Analysis

Specificity (Literature Citation)





anti-Asialo GM1




200 Kd Ly5 antigen on lymphoid cells from spleen,
lymph node, and bone marrow (Kincade et al., 1981b).

Ly5 antigen on lymphoid cells from spleen, bone
marrow, lymph node, and plasma cells but not on
thymocytes or CFU-S (Coffman and Weissman, 1981a).

220 Kd Ly5 variant (B220) on B lineage cells on
spleen, lymph node, and bone marrow, but not on
thymocytes (Coffman and Weissman, 1981b).

Probably recognizes Dolichus biflorus agglutinin
receptor on early fetal thymocytes and on some
thymic leukemia cells (Kasai et al., 1983).

Neutral glycolipid asialo GM1 on early fetal
thymocytes, some fetal liver cells, and few adult
bone marrow, spleen, and lymph node cells as well
as some thymic leukemia cells and natural killer
cells. Not on adult and embryonic, Thy-i positive
thymocytes (Habu et al., 1980; Kasai et al., 1980;
Young et al., 1980).

Thy 1 glycoprotein on thymocytes and T cells from
spleen but not on bone marrow prothymocytes
(Dennert et al., 1980).

Abelson transforming antigen on bone marrow targets
of A-MuLV, on most thymocytes, some bone marrow and
spleen cells, fetal liver erythrocytes and pre-B
cells, and bone marrow pre-B cells, but not on lymph
node cells, CFU-S or stem cells committed to myeloid
lineages (Shinefeld et al., 1980).

Antigen on resident and elicited macrophages, adherent
cultured bone marrow cells and macrophage-like cell
lines as well as on culture-derived mast cells but not
on resident peritoneal mast cells (Katz et al., 1983;
Leblanc et al., 1982).

Rabbit anti-asialo GM1 was received from W.W. Young

(University of Virginia Medical Center) as a delipidated

serum. Rabbit anti-rat immunoglobulins was purchased as a

lyophilized powder (IgG fraction of rabbit anti-rat IgG

(heavy and light chains). Miles Laboratories, Elkhardt, IN)

and reconstituted per manufacturer's specifications.

Rabbit anti-mouse immunoglobulins was purchased from

Gateway Biologicals (St. Louis, MO).

Normal rat serum was prepared from cardiac blood of an

unimmunized animal. Normal mouse serum was prepared from

pooled specimens of multiple unimmunized BALB/cAN mice.

Cell Surface Markers

Cell surface differentiation antigens were detected by

a modification of the method of Uchanska-Ziegler and

colleagues (1982). Formalin-treated, heat killed

Staphylococcus aureus (S.aureus) Cowan I (The Enzyme

Center, Inc.,Boston, MA, and later a generous gift of Dr.

Michael D.P. Boyle) were first coated with rabbit anti-rat

IgG (RAMIgG, Miles Laboratories, Elkhart, IN) and then with

rat monoclonal antibodies specific for mouse

differentiation markers. Either hybridoma cell culture

media or partially purified immunoglobulins were used as a

source of the latter antibodies.

Mouse cells (1x105 or fewer) selected for analysis

were pelleted in round-bottom, PVC microtiter wells

(Dynatech, Alexandria, VA) and resuspended in a small

volume (5 microliters) of S.aureus-RAMIgG-monoclonal

antibody sandwiches. Following a thirty-minute incubation

on ice, the contents of each well were washed six times and

a sample was prepared as a cytocentrifuge mount. Slides

were stained with May-Gruenwald Giemsa or Wright's Giemsa.

Fc Receptor Assays

Receptors for the Fc domain of mouse IgE and IgG were

detected by a novel modification of the S. aureus method

used to detect other surface antigens (Siden and Siegel,

1986). Indicator bacteria were prepared by incubating 25

microliters of packed S. aureus or Escherichia coli (E.

coli) with 0.175 ml of 2,4,6-trinitrobenzene sulfonic acid

(3.7 mg per ml in 0.28M cacodylate buffer, pH6.9) for 10

minutes at room temperature. The 1.5 ml microfuge tube

containing the reactants was wrapped in foil to retard

photodecomposition and taped to a rotator during

incubation. After four washes with 0.01 M phosphate-

buffered 0.15 M sodium chloride, pH 7.3 (PBS; Mishell and

Shiigi, 1980), the TNP-S. aureus or TNP-E. coli were

resuspended to their original volume in balanced salt

solution (BSS; Mishell and Shiigi, 1980) with 10 mM HEPES

(pH 7.35), 0.1% w/v sodium azide, 5x10-5 M 2-

mercaptoethanol (Sigma Chemical Co.) and 1% v/v fetal

bovine serum (H1OBNFl).

Hybridoma supernatants containing mouse IgE anti-

(DNP)2 (IgELa2, American Type Culture Collection) and

partially purified mouse IgG anti-TNP (generous gift of M.

Rittenberg, Oregon Health Center University) were cleared

by centrifugation at 12,000xg for 15 minutes at 4C. The

latter was diluted in Dulbecco's modified Eagle's minimal

essential medium (GIBCO) with 10% v/v heat-inactivated

fetal bovine serum, and both contained 0.1% w/v sodium

azide. Cells to be analyzed were suspended in Hl0BNF1 at

ixl06 per ml, aliquoted at 0.1 ml per well of a 96-well PVC

cluster (Dynatech, Inc., Alexandria, VA) and pelleted by

centrifugation (2 minutes, 10Oxg, 4C). The cluster was

"flicked" and briefly vortexed (five to ten momentary

touches) and the cells were resuspended in 0.1 ml of

cleared IgE or IgG anti-TNP. Clusters were covered with

plastic wrap and placed in a 37C, 5% carbon dioxide

incubator for one hour. Following incubation, the treated

cells were washed twice with Hl0BNF1 by centrifugation.

The cell pellets were dispersed by vortexing and 5

microliters of 10% w/v TNP-S. aureus or TNP-E. coli were

added to each well. The covered microtiter clusters were

placed on ice for thirty minutes and the contents of the

wells were washed six times as previously described. The

pellets were resuspended in 0.1 ml of HO10BNFl. Ten to

thirty microliter samples of each well were

cytocentrifuged, stained with May-Gruenwald Giemsa and

observed by bright field microscopy.


BALB/cAn mice used in these studies were bred in our

colony. BlO.BR/SgSn and CBA/J mice were purchased from The

Jackson Laboratory (Bar Harbor, ME). For studies involving

timed pregnancies, females were placed in the cages housing

one or two males at ratios of one to three females per

male. Females were observed for vaginal plugs following

overnight cohabitation and the date of detection was noted

as day zero of gestation.

Analysis of A-MuLv Proviral Integration

High molecular weight DNA was extracted from tumor

cells by the method of Steffan et al. (1979). Tumor cells

were harvested from liquid culture by centrifugation and

washed in cold PBS. The cells were resuspended at ten

million per milliliter in TES (10 mM Tris, 5 mM EDTA, 100 mM

sodium chloride, pH 7.5) and added dropwise to an equal

volume of lysis buffer consisting of TES with one percent

w/v SDS (Sigma Chemical Co.) and 0.04 percent Proteinase K

(Fisher Scientific Co.). The lysates were digested

overnight at 37C with gentle mixing. DNA was extracted

from the lysate. twice for thirty minutes with an equal

volume of glass-distilled phenol and twice again with

chloroform containing 4 percent isoamyl alcohol (Fisher

Scientific Co.). The extracted DNA was precipitated in 2.5

volumes of 95 percent ethanol or isopropanol, dissolved in

TIOE1 buffer (10 mM Tris, 1 mM EDTA, pH 7.5), and dialyzed

extensively against the same buffer. Normal embryonic

tissue, prepared in an identical manner, provided control

DNA. All DNA was quantitated spectrophotometrically by the

OD260/OD280 method (Maniatis et al., 1982).

Ten micrograms of DNA were incubated with 10 units of

restriction endonuclease BAM HI (New England Biolabs,

Beverly, MA) in TA buffer (O'Farrell et al., 1980);

completeness of digestion was monitored by the addition of

one microgram of bacteriophage lambda DNA to a duplicate

sample. The digested DNA was mixed with sample buffer (50 mM

Tris, pH 7.5, 5 mM EDTA, 25% w/v Ficoll, 0.05% w/v

bromophenol blue, at 5X concentration) and electrophoresed

for eighteen hours at 40 volts D.C. on an 0.8% agarose gel

in TEA buffer (40 mM Tris-acetate, 2 mM EDTA, pH 7.8).

Both the gel and the electrophoresis buffer contained 0.5

micrograms of ethidium bromide per milliliter. Samples

were organized such that a set of digested cellular DNAs

were pipetted into wells on one side of the gel and

duplicates containing bacteriophage lambda DNA were

pipetted into wells on the other side of the same gel.

Following electrophoresis, the gel was photographed

under ultraviolet light to visualize the restricted DNA and

verify that all of the lambda DNA had been digested to

completion. A ruler placed alongside the gel was

photographed at the same time to provide a scale of DNA

fragment sizes for later reference. The lanes containing

lambda DNA were generally cut away and discarded. The gel

was then exposed to shortwave ultraviolet light for an

additional ten minutes to break the DNA and thereby

facilitate transfer. DNA in the gel was denatured with 0.5

M sodium hydroxide, 0.6 M sodium chloride for one hour,

neutralized in 1 M Tris, pH 7.4, 1.5 M sodium chloride (two

changes of 150 to 200 milliliters for 30 to 45 minutes

each), and transferred to nitrocellulose (Schleicher and

Scheull, Keene, NH) by the method of Southern (1975).

Probes for A-MuLV-related sequences were prepared from

the virus-specific recombinant plasmid pAB3Sub3 (Goff et

al., 1980) by nick translation (Rigby et al., 1977) to a

specific activity of 108 dpm per microgram. The 32p_

labeled sequences were hybridized to the nitrocellulose-

immobilized DNA for twenty hours at 68C by the method of

Wahl (1979). The nitrocellulose blot was then washed

extensively under stringent conditions (0.015 M sodium

chloride, 0.0015 M sodium citrate at 68C). The probed

blots were autoradiographed on XAR-5 film (Eastman-Kodak,

Rochester, NY) with two calcium-tungstate-phosphor

intensifying screens (Cronex Lightning Plus, E.I.DuPont de

Nemours and Co., Wilmington, DE) for two to five days.

Conditioned Media

Embryonic tumor cells were cultured under conditions

similar to those used to generate WEHI-3 conditioned media

(W3CM) (Razin et al., 1984a). Cells from log phase

cultures were seeded at Ix106 per ml in o10P in bacterial-

grade petri dishes (Fisher Scientific Co.). Following four

days (92 to 100 hours) of incubation in humidified five

percent carbon dioxide, the conditioned media were

harvested by centrifugation and filtered through tissue

culture grade 0.2 micron nitrocellulose filters.

Conditioned media were concentrated by stirred-cell

ultrafiltration (Amicon, Danvers, MA), by dialysis against

polyethylene glycol 20,000 (Fisher Scientific), or by

ammonium sulfate (Fisher Scientific Co.) precipitation when

such procedures were desired.

Cell lysates were prepared from ceils grown to 7 to

9x105 per ml as follows: Cell cultures were centrifuged

(ten minutes at 200xg, 40) and the cells were washed twice

in PBS. The cells were resuspended in one milliliter 10P

and subjected to three rounds of alternate freezing (in dry

ice-ethanol) and thawing (at 37C). The lysates, which

contained no viable cells upon microscopic examination,

were cleared centrifugally at 2000xg for fifteen minutes

(4C) and at 12,000xg for fifteen minutes (4C), and were

filtered through 0.45 micron sterile, disposable filters

(Gelman Sciences, Ann Arbor, MI). Lysates were stored at

-20C prior to use.

Assay for Interleukin 3

Interleukin 3-like activity was analyzed by a

modification of the method of Razin and colleagues (1984a).

The proliferation of cell line DA-1 (generously provided by

J.N. [hle), which requires interleukin 3 for growth, served

as an assay for interleukin 3. DA-1 cells were

centrifuged, washed twice in PBS, and resuspended at 5x105

per ml in appropriate condititioned media or cell lysates.

One-tenth milliliter aliquots were pipeted in triplicate

into separate compartments of a 96-well microtiter cluster

(Linbro, Flow Laboratories, McLean, VA). The cells were

incubated (37C, five percent carbon dioxide) for sixteen

hours, at which time each well was pulsed with one

microcurie of 3H-TdR (5 Ci per mMol, Amersham Corporation,

Arlington Heights, IL) in ten microliters of EM. Following

six hours additional incubation, the ceils were collected

on Whatman glass microfiber filter strips (Whatman Paper

Ltd., Maidstone, U.K.) in water by multiple automated

sample harvester (MASH, Otto Hiller Company, Madison, WI).

The filter strips were air dried. 3H-TdR which was

incorporated into filter-immobilized DNA was counted in

toluene-based scintillation fluid in a liquid scintillation



Histamine Content of Transformed Placental Cells

Initial light microscopic examination of some of the

transformed cell lines revealed large mononuclear cells

with basophilic granules which stained metachromatically

with acidic toluidine blue (H.R. Katz, personal

communication). Since the cells fit the working definition

of mast cells by these criteria, we began our

characterizations by assaying for intracellular histamine.

All but two of the placental cell lines analyzed contained

histamine as detected by a sensitive, isotopic-enzymatic

microassay (Table 11-2). The quantity of histamine detected

in the cells (5 to greater than 500 nanograms per million

cells) was similar to that found in cultured mast cells

derived from mouse spleen (450 to 500 nanograms per million

cells; Razin et al., 1981a), bone marrow (80 to 150

nanograms per million cells; Razin et al., 1982a), and

fetal liver (200 to 1400 nanograms per million cells; Nabel

et al., 1981), and mucosal mast cells (160 to 2000

nanograms per million cells; Befus et al., 1982b;

Bienenstock et al., 1982), but one order of magnitude less

than that found in serosal mast cells (15 micrograms per

million cells; Bienenstock et al., 1982). Histamine was

not detected in lymphoid or myelomonocytic cell controls,

nor was it detected in the mastocytoma P815, which

reportedly has variants which are devoid of mast cell

granules (Mori et al., 1979). Histamine biosynthesis was

confirmed by chromatographic identification (Galli et al.,

1976) of [3H]-histamine in extracts of [3H]-histidine-

labeled cells (data not shown).

Table 11-2. Histamine Content of Embryonic Tumor Cell Lines
and Control Tumor Cell Lines

(ng/104 cells)a



























a: Histamine content was determined as
Materials and Methods.

described in

Analysis of Hematopoietic Lineage Markers

In view of the morphological and biochemical

similarity of the placental cell lines to mucosal and

culture-derived mast cells, we sought to further confirm

the relationship by analysis of the cell surface

differentiation antigens with an antibody which

discriminated between serosal and culture-derived mast

cells (Katz et al., 1983). This and other cell surface

antigens were analyzed by a simple and sensitive S. aureus-

antibody sandwich method (Uchanska-Ziegler et al., 1982).

Figure II-1A shows the binding reaction of the cell line

10P12 with bacteria coated with anti-cultured mast cell

antibody B23.1, while Figure II-IB shows that the same cell

line, incubated with bacteria coated with unimmmunized rat

immunoglobulin, resulted in no bacteria bound to the mouse

cell surfaces.

Cell surfaces of the remaining A-MuLV placental

transformants and of a number of control tumor cell lines

were probed with monoclonal antibody B23.1 as well as a

panel of other monoclonal and conventional antibodies

reported to be specific for hematopoietic differentiation

markers (Table II-l). Table IT-3 summarizes the results of

the cell surface analyses. Every virus-transformed

embryonic cell line expressed the B23.1 epitope, which is



Figure II-i.

Detection of Cell Surface Determinants on Abelson
Murine Leukemia Virus-Transformed Embryonic Cells.

Killed Staphylococcus aureus bacteria coated with anti-mast
cell/monocyte B23.1 antibodies (A) and normal rat serum
antibodies (B) were reacted with cells from line 10P12 as
indicated in Materials and Methods.



+ + + +

o co L

* 0 o -
41J PL W




i -1


z- <

CI% A-


o ,-4
4-1 U)

*H '-

4-J ,-4
w. 0

u 0

U 0



0 C
ca-I 0
o 6

>-4 e
0 E
z 4
01 ;
o r..

+ +

> --<

< H



Q Q<
o > o


o- 0 ,--1
o '.-IO
C) --

+ +

.-I o)

XI z

0L -Cr
'-4 P.

0 u



-cc E-i
'0 u


+ 4: 4:

t: t: : gz

Q 4-

z 4- + +

z + z

+ z

+ + 4

+ : 4:

rx I:L





+ u



+ + a)



E- F-4


l- l >-'
U ~ Q -4
~ ~,0 -

00 I4
--I ai

+ +




*H a)

e Q)

V.) 0u

(n 3
u u

c xi



1-1 -i


also found on cells of the monocyte lineage (Leblanc et

al.. 1982) as well as on culture-derived mast ceils. This

marker was also detected on control mastocytoma (P815) and

myelomonocytic leukemia cells (WEHI-3).

Two of the cell lines expressed three related

antigenic determinants of the 200 to 220 kilodalton surface

glycoprotein family which is found on mouse lymphoid cells.

The cell lines 11P0-1 and 10P12, as well as all of the

control lymphoid and adult-derived mastocytomas which we

examined, expressed the antigens defined by the monoclonal

antibodies 14.8 and RA3-2C2. These epitopes are expressed

on B cells and their surface immunoglobulin-negative

precursors. The progenitor cells react with the monoclonal

antibody RA3-3A1 as well. The latter antibody, however,

did not recognize a 220 kilodalton glycoprotein on tumor

cells of the T lineage which was detected by 14.8 and RA3-

2C2. The selective reactivity of RA3-3A1 has been

confirmed previously (Coffman and Weissman, 1981b). It is

interesting to note, however, the novel expression of this

previously characterized B cell differentiation antigen on

mastocytoma P815 as well as our embryo-derived mast cell


Placental and control tumor cell lines were also

analyzed for the expression of other primitive lymphoid

markers. The neutral glycolipid asialo-GMl, which is

expressed on the early embryonic, Thy-l-negative thymocytes

(Habu et al., 1980) and for the expression of receptors for

the Dolichos biflorus agglutinin, which is expressed on

early fetal thymocytes and on some thymic leukemia cells

(Muramatsu et al., 1980). The monoclonal antibody M6

(Kasai et al., 1983) reportedly recognizes cells bearing

receptors for Dolichos biflorus agglutinin and probably

binds to the receptor itself. Table 11-3 shows that only

one of the embryonic cell lines, 11P0-1, expressed the

early thymocyte differentiation marker recognized by anti-

FT-1. The same ceil line also bears surface asialo-GMl, as

do two other embryonic cell lines, 9P1 and 10P6. It is

also interesting to note that two A-MuLV-induced adult

mouse mastocytomas, CB6ABMC4 and BALABMC20 were both

positive for FT-1, and that the latter cell line also

expressed the neutral glycolipid asialo-GMl. These

observations will be discussed later in this chapter.

Despite the presence of markers specific for early T

lineage cells on several embryonic cell lines, mature T

lineage markers were only detected on three control cell

lines. Lack of reactivity with monoclonal antibody

T24/31.7 (Dennert et al., 1980) indicated that none of the

cell lines expressed the Thy-l differentiation marker.

Also conspicuously absent from A-MuLV-transformed embryonic

cell lines was the antigen recognized by the monoclonal

antibody 5H1, which is expressed on Abelson murine leukemia

virus transformation-sensitive targets from mouse bone

marrow as well as on thymocytes, fetal liver red blood

cells, and pre-B cells (Shinefeld et al., 1980; E. Siden

and L. Shinefeld, unpublished observations).

Analysis of Receptors for IgE and IgG

High affinity membrane receptors for the Fc regions of

immunoglobulins were detected with a sensitive, isotype-

specific assay developed in this laboratory (Siden and

Siegel, 1986). Coated bacteria were prepared by

derivatizing the bacterial surfaces with the hapten

trinitrophenol (TNP) and then reacting the modified

bacteria with mouse monoclonal anti-TNP antibodies of the

IgE or IgG classes. Experimental and control cell lines

were incubated with the immunoglobulin-coated bacteria at

37C for one hour and processed for cytocentrifugation and

staining. Typical positive and negative reactions are

illustrated in Figure 11-2. Observations are summarized in

Table 11-4. IgE receptors were detected on all but two of

the embryonic cell lines analyzed, while none of the cell

lines expressed high affinity receptors for IgG. Receptors

for IgG were detected, however, on the myelomonocytic

leukemia cell line WEHI-3 (Table 11-4) and on the

macrophage-like tumor cell line P388D1 (data not shown).

The results of the bacterial assay for IgG receptors were

duplicated by the method of Schrader (1981), which used

xenogeneic (rabbit) 7S antibodies immobilized on sheep red



Figure 11-2.

Detection of Surface Receptors for IgE on A
MuLV-Transformed Embryonic Cells.

Killed Staphylococcus aureus bacteria haptenated with
trinitrophenol (TNP) were incubated with cells from line
10P8 which had previously been incubated with monoclonal
mouse IgE (A) and IgG (B) as described in Materials and Methods.

0 0 -i

Table 11-4.


Analysis of Surface Membrane Receptors for IgE
and IgG on A-MuLV-Transformed Embryonic
Cell Lines and on Control Tumor Cell Lines


Mouse IgEa

Mouse IgGa

Rabbit IgGb







RL 11







a: Membrane receptors for allogeneic (mouse) IgE and IgG were
detected by the TNP-bacteria/anti-TNP method (see Materials
and Methods).
b: Membrane receptors for xenogeneic (rabbit) IgG were detected
by the rabbit anti-sheep RBC/SRBC method of Schrader (1981).
c: ND indicates analysis not performed.





blood cells. Rosettes of sheep red blood cells on IgG Fc

receptor-positive cells were detected by phase microscopy

after staining with crystal violet.

Metachromatic Granules in Transformed Cell Lines

Initial examination of several of the embryonic cell

lines, which provided the first evidence for their

relationship to mast cells, was followed by a more detailed

study of the remaining cells. Cells were fixed in Mota's

lead subacetate and stained in acidic toluidine blue to

detect granules rich in basophilic glycosaminoglycan. As

seen in Table 11-5, the majority of the original (BALB/c x

BI0.BR)F1 embryonic cell lines did not possess

metachromatic granules, although two cell lines with high

histamine content (9P1 and 11P62) did express the

characteristic. Additionally, four of six embryonic cell

lines derived from a different paternal background, and all

long-term culture-derived mast cells (see Chapter III),

stained metachromatically with toluidine blue. Many of the

A-MuLV-transformed embryonic cell lines synthesize and

secrete chondroitin-4,6-disulfate proteoglycan (D. Levitt,

R. Porter, and E. Siden, manuscript in preparation), in

support of our observations.

Analysis of A-MuLV Provirus Integration

Although the embryonic tumor cell lines were derived

from cells infected with Abelson murine Leukemia virus, we

sought to confirm the presence of A-MuLV-specific sequences

Table 11-5. Metachromatic Granules in A-MuLV-Transformed
Embryonic Cell Lines and Control Tumor Cell Lines.

Cell Line






a. Cytocentrifuged smears of cells were stained with
toluidine blue as indicated in Materials and Methods.
Cells with metachromatic granules (+) and without
metachromatic granules (-) were scored.

in cellular DNA by identifying the proviral genome. DNA

isolated from four of the embryonic cell lines was probed

with the Abelson virus-specific recombinant plasmid

pAB3Sub3. The cellular DNA was first digested with

endonuclease BamHI, which recognizes no restriction sites

within the virus-specific sequences of the plasmid. Thus,

each integrated A-MuLV provirus detected was represented by

a single band on the autoradiographed blot. All but one of

the cell lines so analyzed showed two copies of the A-MuLV

provirus integrated into high molecular weight DNA (Figure

11-3); cell line 11P62 showed only one copy. In addition to

integrated viral sequences, the probe also detected two

germ line fragments of the endogenous c-abl gene, which

contains cellular sequences and BamHI sites not present in

the plasmid. The v-abl and c-abl patterns of the control

cell line 160N54, from which the infecting virus was

isolated, are also shown.

Two of the embryonic cell lines, 10P12 and 11PO-1,

were injected into mice to determine whether the cells were

tumorigenic. Both cell lines caused tumors in syngeneic

mice inoculated within two weeks of birth. DNA was

prepared from tumors isolated from the mice (output DNA).

Restriction analysis of the input (original cell line) and

output DNA was performed to determine whether the lesions

were due to cell line proliferation or to subsequent

infection of host cells by shed virus. Tumor cells derived