In vitro production of rabbit macrophage tumor cytotoxin

MISSING IMAGE

Material Information

Title:
In vitro production of rabbit macrophage tumor cytotoxin its role in macrophage mediated tumor cell killing
Physical Description:
xi, 112 leaves : ill. ; 29 cm.
Language:
English
Creator:
Fisch, Harvey, 1946-
Publication Date:

Subjects

Subjects / Keywords:
Cytotoxins   ( mesh )
Macrophages   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida.
Bibliography:
Includes bibliographical references (leaves 103-111).
Statement of Responsibility:
by Harvey Fisch.
General Note:
Photocopy of typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 08039004
ocm08039004
System ID:
AA00011170:00001

Full Text














IN VITRO PRODUCTION OF RABBIT MACROPHAGE TUMOR CYTOTOXIN:
ITS ROLE IN MACROPHAGE MEDIATED TUMOR CELL KILLING













BY

HARVEY FISCH


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1981















To my parents, Irving and Martha Fisch, for their years of

love and concern and for setting the proper example.














ACKNOWLEDGEMENTS

First, I wish to thank my advisor, Dr. George E. Gifford, for

his guidance and patience throughout these studies, and for instilling

within me the elements of the scientific method. I also thank the

members of my committee for their helpful comments and suggestions.

The good nod to my fellow graduate students for their helpful

comments and encouragement as well. As a former colleague of mine once

said, "You guys are the only ones who know what it's really like."

Last, but not least, thanks to Mike Duke and Joe Brown for

technical assistance and light hearted humor.














TABLE OF CONTENTS

PAGE
ACKNOWLEDGEMENTS . . iii

LIST OF TABLES . .. ... ... vi

LIST OF FIGURES. .. .. ... . vii

ABBREVIATIONS USED . ....... ix

ABSTRACT . . . x

INTRODUCTION . . 1

Immune System Components: In Vitro Cytotoxicity for
Tumors. . . 1

The Mononuclear Phagocyte System: A Macrophage
Network . . 2

Significance of the Mononuclear Phagocyte System in
Tumor Regression. . . 5

The Process of Macrophage Activation. . 6

The Elicitation of Serum Tumor Necrotic and Cytotoxic
Activities. . . 9

Tumor Necrosis and Tumor Regression. . 13

Activated Macrophages Produce Soluble Cytotoxic
Activity in Vitro . ................. 15

Purpose of These Studies. .... . .. 15

MATERIALS AND METHODS. . ... 17

Animals . . 17

Cell Lines. . . 17

Culture Systems . ..... 17

Effector Cells (Macrophages). . .. 18

Production of Tumor Necrosis Serum. . 21

Production of Rabbit Macrophage Cytotoxins. 21

iv














Cytotoxicity Assays . . 22

Column Chromatography . . .. 28

Effect of Trypsin Inhibitors, Phenylmethyl Sulfonyl
Fluoride, Chloroquine and o-Phenanthrolene on
Rabbit Macrophage Cytotoxin Activities. ... 28

Effects of Prostaglandin E2 on Rabbit Macrophage
Cytotoxin Production. . . 29

Effects of Actinomycin D on Rabbit Pulmonary Lavage
Cell Cytotoxin Production . 29

Effect of Tunicamycin on Molecular Weight of Rabbit
Pulmonary Lavage Cell Cytotoxin . 29

Freeze-Thawing of Activated Rabbit Pulmonary Lavage
Cells . . 29

Susceptibility of Tumor Necrosis Serum Cytotoxin
Resistant L-929 Cells to Rabbit Macrophage
Cytotoxins and Pulmonary Lavage Cells . 30

Time Lapse Cinematography . . 30

RESULTS. . . . 32

Percent Macrophages in Rabbit Effector Cell Populations 32

Increasing the Macrophage Content of Bone Marrow
Cell Cultures . .. 32

In Vitro Cytotoxin Production by Rabbit Macrophages 35

Physicochemical and Biological Properties of Rabbit
Macrophage Cytotoxins and Tumor Necrosis Serum
Cytotoxin . . 63

Cocultivation of Mouse Peritoneal Exudate Cells or
Rabbit Pulmonary Lavage Cells with Actinomycin D
Pretreated L-929 Targets. . 75

DISCUSSION . . 93

BIBLIOGRAPHY . ... . 103

BIOGRAPHICAL SKETCH. . ... 112















LIST OF TABLES


TABLE PAGE

I. RABBIT CELL POPULATIONS UTILIZED AS MACROPHAGE
SOURCES . . 33

II. PERCENT RABBIT BONE MARROW CELLS WITH MACROPHAGE
PROPERTIES BEFORE AND AFTER TPA TREATMENT 34

III. ITEMS TESTED WITH LIMULUS AMEBOCYTE LYSATE REAGENT. 38

IV. RABBIT BONE MARROW CYTOTOXIN PRODUCTION: OPTIMUM
PLATING DENSITY . 43

V. EFFECT OF TPA ON TUMOR NECROSIS SERUM (TNS) TITER 45

VI. CYTOTOXIN PRODUCTION BY RABBIT MACROPHAGES FROM
BLOOD MONONUCLEAR, TPA-PRETREATED BONE
MARROW AND PULMONARY LAVAGE CELLS:
COMPARISON OF TITERS PRODUCED PER
1 x 10 CELLS . 46

VII. DATA FROM TABLE VI CONSIDERING PERCENT MACROPHAGE
CONTENT OF EACH CELL TYPE . 48

VIII. EFFECTS OF CHLOROQUINE AND o-PHENANTHROLENE ON
CYTOTOXIN ACTIVITY. . 72

IX. CELL SUSCEPTIBILITY PROFILES OF RABBIT CYTOTOXINS 74

X. SUMMARY: COMPARISON OF PHYSICOCHEMICAL AND
BIOLOGICAL PROPERTIES OF RABBIT TUMOR NECROSIS
SERUM CYTOTOXIN (TNSCT) WITH RABBIT
MACROPHAGE CYTOTOXINS . .. 76

XI. 51Cr RELEASE BY L-929 CELLS COCULTURED WITH
MOUSE PERITONEAL EXUDATE CELLS. .. 77

XII. EFFECTS OF TNSCT, PLCCT, BMCCT, AND PLC ON TNSCT-
RESISTANT AND CONVENTIONAL L-929 CELLS. 92













LIST OF FIGURES


FIGURE PAGE

1. Comparison of automated and manual photocell
in the photometric cytotoxicity assay. 25

2. Rabbit macrophage cytotoxin production with
and without added LPS. . 37

3. Rabbit macrophage cytotoxin production in the
absence and presence of varying levels of
exogenous LPS. . .. 41

4. Effects of bovine and fetal bovine serum on
pulmonary lavage cell cytotoxin
production. .. . 50

5. Comparison of effects of 1% and 10% fetal bovine
serum on cytotoxin production by rabbit
pulmonary lavage cell cultures containing
various levels of LPS .. .. 52

6. Time kinetics of rabbit pulmonary lavage cell
cytotoxin production. . 54

7. Time kinetics of cytotoxin production by
TPA-pretreated and washed rabbit bone
marrow cells. .. . 57

8. Effects of actinomycin D on cytotoxin
production by rabbit pulmonary lavage
cells . . 60

9. Effects of prostaglandin E2 and prostaglandin
antiserum on cytotoxin production by rabbit
pulmonary lavage cells . 62

10. Gel filtration of rabbit macrophage cytotoxins
on Sephacryl S-200 . 65

11. Ion exchange chromatography of rabbit macrophage
cytotoxins on DEAE-Sephadex. . 68

12. Thermal stability of rabbit TNS cytotoxin
and macrophage cytotoxins. . 71

13. C3H/HeN mouse peritoneal exudate cells on
actinomycin D pretreated L-929 cells:
ninety-six well tray . 79













14. Effector cell cytotoxicity for actinomycin D
pretreated L-929 targets: Determination
of S50 endpoint. . 82
15. C3H/HeN mouse peritoneal exudate cells on
actinomycin D pretreated L-929 cells:
Twenty-four well tray. . 84

16. Rabbit pulmonary lavage cells on actinomycin D
pretreated L-929 cells: Twenty-four
well tray. . 86

17. Linearity between number of plaques counted
vs. number of mouse peritoneal exudate
cells plated . . 88


viii












ABBREVIATIONS USED

BCG Mycobacterium bovis, strain BCG

BLM blood mononuclear cells

BLMCT blood monocyte cytotoxin

BMC bone marrow cells

BMCCT bone marrow cell cytotoxin

BPTI bovine pancreatic trypsin inhibitor

CT cytotoxin(s)

FMEM fortified Eagle's minimum essential medium

HEPES N-2-hydroxyethyl piperazine-N'-2-ethane
sulfonic acid

LPS lipopolysaccharide

MAF macrophage activating factor

MEF mouse embryo fibroblasts

MEM Eagle's minimum essential medium

MPS mononuclear phagocyte system

PEC peritoneal exudate cells

PGE prostaglandin E

PLC pulmonary lavage cells

PLCCT pulmonary lavage cell cytotoxin

PMSF phenylmethyl sulfonylfluoride

RES reticuloendothelial system

SBTI soybean trypsin inhibitor

TNF tumor necrosis factor

TNS tumor necrosis serum

TNSCT tumor necrosis serum cytotoxin

TPA 12-o-tetradecanoylphorbol 13-acetate















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



IN VITRO PRODUCTION OF RABBIT MACROPHAGE TUMOR CYTOTOXIN:
ITS ROLE IN MACROPHAGE MEDIATED TUMOR CELL KILLING

By

Harvey Fisch

December 1981

Chairman: George E. Gifford
Major Department: Immunology and Medical Microbiology

Rabbit macrophages derived from pulmonary lavage cells (PLC),

phorbol ester pretreated and washed bone arrow cells (BMC), or

blood mononuclear cells (BLM), cultured in the presence of LPS,

secreted tumor cell cytotoxins (CT) that were similar to each other

and tumor necrosis serum cytotoxin (TNSCT).

All CT's had molecular weights of approximately 48,000 D by

gel filtration and eluted from DEAE-Sephadex between 0.28 and 0.32 M

salt. All were stable to 560C for 60 minutes, but labile to 700C

for 20 minutes. Actinomycin D (AcD) enhanced sensitivity of L-929

cells to CT. B16C3 melanoma cells and mouse embryo fibroblasts

were resistant to CT.

By 3 hours in culture, all effector cells secreted detectable

levels of CT. Titers increased for PLC and BLM between 3 and 5

hours. After that time, BLM titers remained constant through 24

hours. PLC titers rose again from 18 to 30 hours. BMC cultures












demonstrated increasing titers from 3 through 14 hours. After that,

no increase in titer was demonstrated. No additional CT production

could be demonstrated after 30 hours in BMC or PLC cultures, despite

a change to fresh medium with LPS.

On an adjusted basis of 1 x 106 macrophages/ml, PLC produced

4 to 25 times more CT than BMC or BLM after 28 hours culture.

CT production by LPS-activated PLC could occur in serum-free

medium. Fetal bovine serum (10%) enhanced production up to 4-fold.

Bovine serum (>4%), or prostaglandin E2 (10-6M) inhibited CT

production. AcD (1 Ug/ml) added with LPS inhibited CT production

(>95%) by PLC. Delaying addition of AcD after LPS demonstrated

messenger RNA production for CT was completed by 2 to 6 hours.

AcD pretreatment of L-929 cells (2 ug/ml, 2 to 3 hours)

followed by washing enhanced their sensitivity to PLC killing. From

this, a plaque assay was generated demonstrating a linear

relationship between number of PLC plated and plaques formed.

Therefore, 1 PLC was capable of forming a plaque. Time lapse

cinematography demonstrated that contact between PLC and targets was

unnecessary for target cell killing.

B16C3 cells and L-929 cells selected for resistance to TNSCT

were resistant to PLC mediated cytotoxicity.

The conclusion drawn is that secreted CT may play a major role

in macrophage mediated tumor cytotoxicity.















INTRODUCTION

The problems encountered in biological disciplines are often

frustrating and at times seemingly insurmountable. This is the case in

the broad field of tumor immunology, an area filled with complexity and

apparent contradiction.

Immune System Components: In Vitro Cytotoxicity for Tumors

The interactions of the various humoral and cellular components of

the immune system with tumors as well as with each other in vivo and in

vitro have been well documented. This has been described as a "network"

(Jerne, 1974, 1977) as well as a surveillance system seeking out tumors

as they arise within the host (Thomas, 1959; Burnet, 1967, 1970). Some

investigators seriously question the validity of these sweeping concepts

in the face of spontaneously arising tumors (Hewitt et al., 1976).

Evidence also exists illustrating that some tumors grow as a result of an

immune response (Prehn, 1976).

Despite these difficulties, there has been much effort put forth to

investigate and understand just what the components of the immune system

can do when confronted with tumors both in the test tube and in the tumor

bearing host. More specifically, sensitized cytotoxic T lymphocytes

(Brunner and Cerottini, 1971), natural killer lymphocytes (Herbennan et

al., 1975), and armed and activated macrophages (Alexander, 1974)

represent cell types that can directly kill tumor cells subsequent to

contact with them. Some of the reported humoral activities cytotoxic for










tumor cells include specific antibody and complement (Hellstrom et al.,

1968), lymphotoxin, a product of lymphocytes (Granger and Kolb, 1968),

tumor necrosis factor, and cytotoxic factor (Carswell et al. 1975),

presumed macrophage products (Matthews, 1978; Mannel et al. 1979, 1980a,

1980b). Examples of humoral and cellular elements working together to

destroy tumor cells include specific antibody with nonimmune lymphocytes

(Pollack et al., 1972) and specific antibody with macrophages (Whitcomb,

1979).

The Mononuclear Phagocyte System: A Macrophage Network

The mononuclear phagocytes or macrophages represent a large,

morphologically heterogeneous group of cells found widely distributed

throughout the body (Vernon-Roberts, 1972). Metchnikoff (1905)

recognized the presence of mononuclear cells in the blood, lymph, as well

as throughout the various organs, that could engulf small particles such

as bacteria by projecting ameboid processes. He coined the term

phagocytosiss" (from the Greek word, "phagein," to eat) and designated

the large free and fixed phagocytic cells as "macrophages." He also

demonstrated the significance of these phagocytic cells concerning host

defense by reconfirming their presence in inflamed tissue as well as

noting their phagocytic capacity for bacteria.

Utilizing vital dyes which are preferentially taken up by

macrophages and to a lesser extent by other types of cells, Aschoff

(1924) demonstrated the distribution throughout the body of a functional

system of cells which included macrophages. He referred to this overall










network of cells as the "reticuloendothelial system (RES)." It was

believed that these cells were responsible for the formation of the

extrasinusoidal solid pulp of the lymph nodes and spleen reticulumm). In

addition, they were found lining the sinusoids of lymph nodes and other

organs (endothelium). Additional investigations extended the functional

RES to include the microglia of the central nervous system (del Rio

Hortega and de Asua, 1924) and blood monocytes (Marshall, 1956).

Later studies reviewed by van Furth et al. (1975) revealed that the

original concept of the RES as a unified system is inappropriate in that

it really consists of at least two distinct systems. One of these, the

"RES proper" is composed of cells that are nonphagocytic and make up the

framework of connective tissue. Distinct from these cells are the

macrophages, which often exist either fixed or free, in close proximity

to connective tissue. As a result, the term mononuclearr phagocyte

system" (MPS) has been proposed to denote a functional network made up

exclusively of macrophages (van Furth et al., 1975).

As previously mentioned, macrophages display considerable

morphological heterogeneity (Vernon-Roberts, 1972). This no doubt

created considerable confusion among early investigators regarding what

was and was not to be considered a macrophage. Perhaps the most

enlightening studies concerning this problem, as well as revealing the

source of macrophages, have dealt with their ontogeny. Macrophages

originate from a common source in the bone marrow, enter the circulation

and finally pass into the tissues, changing their morphological and many

of their functional characteristics as they proceed (van Furth et al.,









1975; van Oud Alblas and van Furth, 1979). Therefore, the varied

descendants of the bone marrow precursor merely represent different

stages or various end cells of a progressive differentiation process.

Aside from their mononuclear appearance, intense phagocytic

capability and common origin, macrophages share an array of other

morphological, chemical and functional characteristics. Included among

these properties are the ability to adhere tightly to glass, motility,

possession of abundant lysosomes containing various enzymes, production

of various immune modulators, complement components, superoxide free

radical and hydrogen peroxide and possession of Fc and complement

receptors (Eisen, 1980). In additional to morphological features,

heterogeneity among macrophages can be extended to these properties as

well since different macrophages vary qualitatively and quantitatively

with respect to these characteristics (van Oud Alblas and van Furth,

1979; Keller, 1980). This overall heterogeneity reflects the existence

of clearly separable subpopulations (Keller, 1980) with either distinct

genetic differences or changes that coae about with pluripotent

macrophage maturation or activation.

Many of these properties contribute to the ability of macrophages

to either act alone or in conjunction with other cells or products of the

"immune system" (T and B lymphocytes, antibody and complement) to help

maintain a state of homeostasis within the host. More specifically, this

involves elimination of exogenously introduced toxins, invading

microorganisms, or other foreign antigens, as well as the destruction and

eventual elimination of host cells that are effete or have undergone

certain alterations. This latter case includes the changes seen in cells

associated with neoplasia.










Significance of the Mononuclear Phagocyte
System in Tumor Regression

Nearly a century ago, Coley (1891) reported on the palliative (and

occasional curing) effects of using bacterial culture filtrates to treat

patients with neoplastic disease. The rationale for this treatment of

neoplasia was based on the observation (made long before that) that

cancer patients with concurrent infectious disease often displayed

regressing neoplasia or tumors that advanced less rapidly than those of

cancer patients without concurrent infection. More recent reviews

(Pearl, 1929; Jacobsen, 1934; Nauts et al., 1953) have discussed related

phenomena in detail.

Later investigations revealed that the MPS could expand and

increase its activity during the course of a neoplastic disease (Old et

al., 1960) or following the administration of such agents as

Mycobacterium bovis, strain BCG (Old et al., 1960; Leake and Myrvik,

1968), zymosan (Old et al., 1960), or Corynebacterium parvum (Halpern et

al., 1974). Similar studies also showed that such agents could protect

challenged animals against transplanted tumors or increase the survival

time of these animals (Old et al., 1959; 1960).

Further evidence that contributed credibility to the macrophage as

a critical component of the in vivo effector response to tumors was later

provided. Macrophages could be made nonspecifically cytotoxic to tumor

cells by culturing them in the presence of endotoxin (Alexander and

Evans, 1971), or supernatant fluids from antigen sensitized lymphocyte

cultures (Piessens et al., 1975; Leonard et al., 1978). Alternatively,

macrophages could be rendered specifically cytotoxic or "armed" to a










particular tumor by cocultivating them with spleen lymphocytes fran an

animal previously inoculated with the tumor (Alexander, 1974). These

macrophages could then be rendered nonspecifically cytotoxic (activated)

to different tumors by continued cocultivation with the original

immunizing tumors (Alexander, 1974).

More convincing evidence indicating that macrophages play a

significant role in the overall scheme of tumor defense mechanisms was

revealed by their actual presence within tumors (Evans, 1972; Alexander,

1974; Eccles and Alexander, 1974). Furthermore, tumor regression could

be correlated with macrophage infiltration (Russell et al., 1976).

The Process of Macrophage Activation

The term "activated" as it pertains to the macrophage was

introduced by MacKaness (1970) and it is a term which has received wide

usage by many investigators who study the macrophage. Depending upon

one's interests, the activated macrophage possesses many characteristics

which distinguish it from a macrophage which is not activated. Allison

and Davies (1975) have described biochemical and biological changes in

macrophages undergoing activation. These include increases in the

following: glucose oxidation, lysosomal enzyme levels, synthesis of

membrane components, pinocytosis, phagocytosis, adhesion, spreading and

motility. In addition, activated macrophages, as opposed to resting

macrophages, can inhibit the proliferation of and have demonstrated

cytotoxicity for tumor cells (Allison and Davies, 1975; Hibbs, 1976;

Keller, 1980).

The cytotoxicity that activated macrophages demonstrate for

transformed cells is most interesting. Aside from being nonspecific







7
as seen by the ability of activated macrophages to kill different types

of tumors from allogeneic, syngeneic or xenogeneic sources (Keller,

1980), most nontransformed cells are not killed by activated macrophages

(Hibbs, 1974a; Keller, 1980). Therefore, the activated macrophage

appears to be capable of distinguishing some common features) of the

tumor state from that of normal cells. Currently, the exact distinction

between tumor and nontransformed cells made by the activated macrophage

is unknown. However, some evidence suggests the difference lies within

the plasma membrane (Hibbs, 1974a).

The process of macrophage activation is not a simple, absolute,

one-step event, but involves a cascade of transient, intermediary

reactions (Hibbs et al., 1980; Meltzer et al., 1980) dependent upon the

presence of signals in the local environment of the macrophage (Hibbs et

al., 1977). Some of these signals are stimulatory while others are

inhibitory to macrophage activation (Hibbs et al., 1977; Chapman and

Hibbs, 1977). Ruco and Meltzer (1978) have extended this model by

describing it as a series of "priming" signals followed by activation

signalss.

As mentioned previously, some of the signals that can affect the

state of the macrophage and modulate activation have been identified.

They include endotoxin (Alexander and Evans, 1971), macrophage activating

factor, a T lymphocyte product (Leonard et al., 1978) and type I

interferon (Schultz et al., 1977; Jett et al., 1980). As work in this

field progresses, it is becoming more and more evident that many of these

activation signals are related to one another as part of the macrophage

activation cascade, particularly as priming or activation signals.










This is the case concerning macrophage activating factor (MAF) and the

active component of endotoxin, lipopolysaccharide (LPS). It has been

established using highly purified materials that MAF alone cannot

activate macrophages, but merely sensitizes or "primes" them to the

activating capabilities of LPS (Weinberg and Hibbs, 1979; Weinberg and

Hibbs, 1980; Hibbs et al., 1980; Weinberg, 1981; Pace and Russell, 1981).

Earlier studies concerning these two signals in this scheme were

complicated by the presence of small amounts of LPS contaminating the

lymphokine preparations as well as other materials (Hibbs et al., 1980).

Aside from clarifying the roles of MAF and LPS in the macrophage

activation scheme, many of these studies demonstrate the importance of

determining the presence of contaminating LPS when working with

macrophages (Levin et al., 1970; Fumarola, 1981).

Despite the apparent clarity of this simple model utilizing MAF (a

priming signal) and LPS (a triggering signal ), there are many points

regarding macrophage activation and its associated signals that are still

not clear. First, it is apparent that LPS is not the only triggering

signal. This has been demonstrated in several studies which either

exclude LPS from cultures or utilize macrophages derived from mice that

respond very poorly to LPS (Hibbs, 1976; Schultz and Chirigos, 1980;

Boraschi and Tagliabue, 1981). Secondly, type I interferon has been

described as a priming signal distinct from MAF (Boraschi and Tagliabue,

1981) as well as serving as the endogenous macrophage generated mediator

for LPS induced triggering of macrophage activation (Schultz and Chirigos,

1979). Further studies are needed to determine the identity of other










triggering signals as well as the precise role of interferon in the

scheme of macrophage activation.

The Elicitation of Serum Tumor Necrotic
and Cytotoxic Activities

Many of the foregoing mechanisms concerning macrophage activation

are now believed to represent accurate in vitro correlates of host

mediated responses to such agents as BCG, zymosan, Corynebacterium parvum

and LPS. Although the early investigators were not aware of the details

behind the host's response to these agents, it was quite apparent that

they had value in inducing favorable clinical responses in tumor bearing

animals. Ribi et al. (1975) gave tumor bearing guinea pigs both BCG and

LPS. The results of this double regimen were rather impressive. Test

animals that had received both BCG and LPS displayed a survival rate of

about 90% compared to a survival rate of about 66% for those receiving

BCG alone.

Carswell et al. (1975) found that they could cause a dramatic

necrosis of transplanted subcutaneous tumors in otherwise unmanipulated

mice that were injected with sera from other mice, rabbits, or rats that

had previously been inoculated intravenously with BCG, Corynebacterium

species or zymosan followed by an intravenous injection of LPS 2 weeks

later. They called the agents) in the sera responsible for this

phenomenon, tumor necrosis factor (TNF). It was also demonstrated by the

same group (Carswell et al., 1975; Helson et al., 1975; Green et al.,

1976) that crude tumor necrosis serum (TNS) or material that was

partially purified from TNS could kill many but not all types of tumor

cells in vivo and in vitro while being nontoxic for mouse embryo









fibroblasts in vitro. Others have reported similar findings for TNS

derived from the mouse (Mannel et al., 1980a; Mannel et al., 1980c) and

rabbit (Matthews and Watkins, 1978; Ostrove and Gifford, 1979). In

addition, some of these as well as other studies demonstrate that neither

tumor necrosis or in vitro cytotoxicity are due to LPS (Carswell et al.,

1975; Oettgen et al., 1980). LPS levels in the serum transferred from

donor animals are substantially below the levels required to cause tumor

necrosis. Furthermore, LPS is not toxic for tumor cells in vitro and has

a profile distinctly different from that of TNF regarding other

biological activities (Oettgen et al., 1980).

Biochemical and Physical Properties of Tumor
Necrosis Factor and Cytotoxins in Tumor
Necrosis Serum

Green et al. (1976) partially purified TNF from mouse TNS and

determined it to have a molecular weight of 125,000 to 150,000 D by gel

filtration. Further studies in the mouse (Mannel et al., 1980a; Mannel

et al., 1980c; Oettgen et al., 1980) demonstrated the molecular weight of

the cytotoxic activity (CT) and TNF respectively (by gel filtration) to

be about 60,000 D, after exposure to high salt concentrations (0.2 M or

greater). This may indicate ionic aggregation of polymeric subunits or

aggregation with some other molecular species found in the serum. Other

properties of mouse TNF and tumor necrosis serum cytotoxin (TNSCT)

include stability to 560 C exposure for 30 minutes, an isoelectric point

of pH 4.8, pronase sensitivity, and the presence of sialic acid and

galactosamine residues (Green et al., 1976; Mannel et al., 1980c; Oettgen

et al., 1980).







11
Studies in the rabbit employing gel filtration have determined the

molecular weight of TNSCT to be 55,000 D (Ruff and Gifford, 1980) and

39,000 D (Matthews et al., 1980). The two groups reported approximately

the same molecular weight (67,000-68,000 D) utilizing polyacrylamide gel

electrophoretic techniques. The consistently higher molecular weight as

determined by electrophoresis over that of gel filtration could possibly

be explained by alteration of configuration of the cytotoxic molecule by

denaturing conditions (Ruff and Gifford, 1980) or interaction of the

molecule with the electrophoretic matrix resulting in retardation of

migration (Matthews et al., 1980). The reasons for the difference between

the two groups concerning the molecular weight as determined by gel

filtration are unclear. Additional characteristics of rabbit TNSCT

include relative stability to 700C for 20 minutes (Matthews et al.,

1980), an isoelectric point of pH 5.1, and high sensitivity to pronase

and relative insensitivity to trypsin (Ruff and Gifford, 1980; Matthews

et al., 1980). These and other studies (Ruff and Gifford, 1981) suggest

that the rabbit CT molecule does not contain large amounts of exposed

carbohydrate contains little or no carbohydrate, as based upon inability

to bind cytotoxic activity to various plant lectins.

Are Tumor Necrosis Serum-Derived
Necrotic and Cytotoxic Activities
Mediated by the Same Effectors?

It is tempting to speculate from results of many of the

aforementioned studies that the TNS agents responsible for tumor necrosis

(TNF) and in vitro cytotoxicity are one and the same. However, the

complexity of TNS is underscored by the demonstration that a variety of

biological products such as interferon (Sauter and Gifford, 1966), B cell

differentiating factor (Hoffman et al., 1976), bone marrow colony











stimulating factor (Hoffman et al., 1977) and lysosomal enzymes (Sauter

and Gifford, 1966; Green et al., 1976; Old, 1976) are released into the

serum under conditions used for induction of tumor necrosis. It is

therefore possible that many mediators may contribute to tumor necrosis

and cytotoxicity and the mediator(s) responsible for one of these

phenomena may not be responsible for the other.

Initial studies employing crude TNS (Carswell et al., 1975) implied

that a single entity, TNF, could be responsible for both in vivo tumor

necrosis and in vitro cytotoxicity. As a result, many of the studies that

followed exclusively utilized either the in vivo assay for necrosis

Green et al., 1976) or an in vitro assay for cytotoxicity (Matthews et al.,

1980; Mannel et al., 1980d) to study purified effector material. It was

felt that either assay probably reflected the activity of the same

effector. Along these lines, Oettgen (1980) and Kull and Cuatrecasas

(1981), working in the mouse system, were able to partially copurify both

cytotoxic and necrotic activities from TNS. Additional studies

demonstrated that partially purified CT derived from rabbit TNS could

elicit tumor necrosis (Ruff and Gifford, 1980). The data of Kull and

Cuatrecasas (1981) are particularly interesting because they demonstrated

material with cytotoxic activities having molecular weights of 50,000,

160,000 and 225,000 D, but only the 160,000 D containing fractions also

displayed necrotic activity. In addition, they found that high molecular

weight species persisted after salt fractionation, and gel filtration of

the high molecular weight form in high salt led to loss of activity without

an increase in activity of low molecular weight species (50,000 D). This










is contrary to the hypothesis concerning high salt deaggregation of high

molecular weight forms (Mannel et al., 1980a; 1980c).

The results of the aforementioned studies do not definitively

demonstrate whether or not TNS necrotic and cytotoxic activities can be

attributed to a single effector molecule. Further investigation is

needed to clarify this point.

Tumor Necrosis and Tumor Regression

Early studies utilizing animals with transplanted subcutaneous

tumors demonstrated a "hemorrhagic necrosis" of their tumors after being

administered bacterial culture filtrates of such organisms as Escherichia

coli (Gratia and Linz, 1931). Later, Shear and Turner (1943) purified a

polysaccharide containing lipid material from Serratia marcescens (now

known to be LPS) which alone could cause a rapid and dramatic hemorrhagic

necrosis of tumors. Although the mechanism behind this phenomenon

remains unclear at this time, there is evidence suggesting an LPS induced

vasoconstriction and intravascular coagulation within the tumor (Algire

et al., 1952) as well as a need for lymphoreticular cells that are

sensitive to LPS (Mannel et al., 1979). As already mentioned, a similar

phenomenon can be seen in tumor bearing animals administered TNS

(Carswell et al.,1975) or partially purified TNF (Ruff and Gifford,

1980). It is not yet clear (but is circumstantially possible) that TNF

plays a role in the tumor necrosis induced by LPS alone. As rapid (often

less than a day) and dramatic as this phenomenon is, a ring of viable

tumor cells usually survives at the periphery of the neoplasm to grow and

kill the host (Shear, 1943). This demonstrates that tumor necrosis










in itself is usually insufficient to lead to total recovery fron

neoplastic disease.

Tumor regression, on the other hand, involves all those events

which contribute and ultimately lead to total recovery of the host from

naturally occurring or experimentally induced tumors. It is possible

that tumor necrosis may contribute to or allow subsequent tumor

regression to take place more readily by destroying those parts of the

tumor (the center) which effector mechanisms (such as activated

macrophages) may have difficulty reaching.

As previously discussed, much evidence has been provided concerning

the significance of the macrophage in overall tumor regression (Evans,

1972; Alexander, 1974; Eccles and Alexander, 1974; Russell et al., 1976).

It is possible that in many cases, the macrophage may be directly

involved in tumor regression that has been attributed to other types of

effector cells (natural killer cells, cytotoxic T lymphocytes). As an

example, studies by Berendt et al., (1978a; 1978b) have indicated that

LPS induced tumor regression requires concomitant T cell immunity and

that only those tumors that are immunogenic enough to evoke T cell

immunity will regress. At first inspection it is difficult to envision a

role for macrophages in such a scheme. However, the macrophage

activation scheme involves a series of caoplex events dependent upon

various signals such as MAF, a T lymphocyte product (Leonard et al.,

1978; Weinberg and Hibbs, 1979). It is quite possible that macrophages

may be "primed" by MAF produced by immunized T cells and subsequently

activated by the administered LPS, thereby allowing them to be

responsible at least in part for LPS induced tumor regression.










Activated Macrophages Produce Soluble Cytotoxic
Activity in Vitro

The involvement of the macrophage in the production of tumor

necrotic and cytotoxic serum activities is suggested by the coincident

hyperplasia of the MPS accompanying the administration of many TNS

eliciting agents (Carswell et al., 1975; Old, 1976). In addition,

evidence demonstrating that macrophages can kill tumor cells in vitro

(Alexander, 1974; Hibbs, 1976) prompted studies that sought to better

understand the relationship of TNF and TNSCT to activated macrophages.

Studies in the mouse have demonstrated that LPS sensitive cells are

needed to induce TNF in serum (Mannel et al., 1980d) and that biochemical

and physical properties of serum-derived TNSCT are similar to those of CT

derived from cultures of activated macrophages and macrophage cell lines

(Mannel et al., 1980a, 1980b, 1980d). In addition, these same studies

demonstrated that a rabbit antiserum directed against serum-derived TNSCT

inhibited all soluble cytotoxicity generated in the macrophage cultures.

Working in the rabbit, Matthews (1978) has demonstrated that blood

monocytes produce soluble tumor CT in vitro. This CT has striking

similarity to rabbit TNSCT in terms of selective toxicity for various

tumor cells and biochemical characteristics.

Purpose of These Studies

The aims of these studies are as follows:

1. To study the production of rabbit macrophage CT in vitro

utilizing macrophages from different sources (pulmonary alveolar, bone

marrow, and blood mononuclear cells).









16
2. To compare physicochemical and biological properties of these

CT with one another as well as with those of TNSCT.

3. To determine the significance these CT play in macrophage

mediated tumor cytotoxicity.














MATERIALS AND METHODS

Animals

New Zealand white female rabbits, ranging in age from 3 to 5 months

(mean weights of 3 to 5 kilograms) were employed. Mice (C3H/HeN and

C57B1/6, female) used ranged in age from 6 to 12 weeks. All animals
were obtained from commercial sources and housed and fed conventionally.

Cell Lines

L-929 cells (a C3H mouse cell line) were grown in Eagle's minimum

essential medium (MEM) supplemented with 10% bovine serumi and 250 units of

penicillin/ml and 125 pg of streptomycin/ml. B16C3 melanoma cells (a

C57B1/6 mouse cell line) were grown in fortified MEM supplemented with
10% fetal bovine serum. Fortified MEM contained 2x concentrations of

essential and nonessential amino acids and vitamins. Glutamine and sodium

pyruvate were also added. Primary mouse embryo fioroblasts (MEF) were

prepared by mincing near term embryos followed by trypsinization for 60

minutes (0.1% trypsin, 0.04% ethylene diamine tetraacetic acid), washing

in Gey's balanced salt solution and culturing in glass bottles containing

FMEM with 10% fetal bovine serum.

Culture Systems
Media

These have been described above (see Cell Lines). For use in

cultures where macrophages were employed, N-2-hydroxyethyl piperazine-N'-

2-ethane sulfonic acid (HEPES buffer, Calbiochem, LaJolla, CA) was added

at .03M.










Sera

Two lots of bovine sera were used: Gibco (Grand Island, NY), lot

number R290222 and Sterile Systems, Inc. (Logan, UT), lot number 400168.

The latter was certified to contain less than 1 ng lipopolysaccharide/ml .

Unless otherwise stated, this was the lot used. One lot of fetal bovine

serum was used for all experiments in this study (Kansas City Biological,

Lenexa. KS).

Lipopolysaccharide-Free Culture System

Using a Limulus amebocyte lysate reagent kit (Pyrogent, Mallinckrodt

Inc., St. Louis, MO) to test various media components for contaminating

lipopolysaccharide (LPS), a culture medium was established which contained

less than 0.025 ng LPS/ml (see Table III for the various materials

screened). This "LPS-free" culture system was used for most of the

studies to be discussed.

Effector Cells (Macrophages)

Rabbit Macrophage Sources

Rabbit macrophages were derived from three sources, pulmonary lavage

cells (PLC), bone marrow cells (BMC), and blood mononuclear cells (BLM).

For harvesting PLC and BMC, rabbits were anesthetized with a combination

general anesthetic consisting of ketamine hydrochloride (Bristol

Laboratories, Syracuse, NY) and xylazine hydrochloride (Cutter

Laboratories, Shawnee, KS) followed by intravenous air embolism sacrifice.

After hair removal and cleansing of the skin with 70% ethanol, the trachea

and/or femurs were exposed. For PLC harvest, the trachea was transected

and the lungs flushed out with six aliquots of 50 ml cold sterile pyrogen-

free saline using a sterile disposable syringe. Cells were immediately

removed from the cold saline by centrifugation and resuspended









in cold MEM supplemented with 10% fetal bovine serum and .03M HEPES

buffer. Cells were generally washed twice by centrifugation in medium.

About 20 to 40 x 106 PLC were obtained per rabbit. For BMC, femurs were

removed and trimmed free of as much soft tissue as possible. After

the bones had been rinsed well with sterile, pyrogen-free saline, the ends

were removed with sterilized bone cutters and the contents of the marrow

cavity flushed out with FMEM supplemented with 10% fetal bovine serum and

.03M HEPES buffer, using a 16 gauge needle attached to a 20 ml sterile

disposable syringe. Up to 2 x 109 bone marrow cells were obtained per

rabbit.

For harvesting BLM, rabbits were anesthetized and bled from the

marginal ear vein into sterile tubes containing pyrogen-free heparin.

After collection, the blood was underlayered with half its volume of

Histopaque-1077 (Sigma Chemical Co., St. Louis, MO) and centrifuged at 400

g for 40 minutes at room temperature. The leukocytes at the interface

were collected, washed 3 times with cold pyrogen-free saline and suspended

in MEM, supplemented as for PLC. Generally, 50 to 100 x 106 BLM were

obtained from one rabbit. Where necessary to minimize effector cell

clumping and to insure a majority of single effector cells (see Plaque

Assay), PLC were pushed through a 19, 23 and finally a 25 gauge hypodermic

needle, using a syringe.

Leake and Myrvik (1968) have described a technique to significantly

increase the yield of PLC from rabbit lungs. When this technique was

modified by intravenous injection of 0.25 ml Freund's complete adjuvant

(Gibco, Grand Island, NY) instead of Mycobacterium bovis, strain BCG

(Leake and Myrvik, 1968), PLC yields were significantly increased up to

about 200 x 106 per rabbit, if cells were harvested 2 to 3 weeks after

the injection. These rabbits will be referred to as "primed" throughout










these studies. Those rabbits which were previously unmanipulated prior to

sacrifice will be referred to as "normal."

Mouse Macrophage Source

Mouse macrophages were derived froa peritoneal exudate cells (PEC).

C3H/HeN and C57B1/6 mice were injected intraperitoneally with 2.5 to

3.0 ml of fluid thioglycollate medium (Bacto, Difco, Detroit, MI). Three

days later, the mice were sacrificed by cervical dislocation and peri-

toneal exudate cells (PEC) harvested by washing out the peritoneal

cavities with 8 to 10 ml of cold medium. The cells were washed twice by

centrifugation and resuspended in fresh medium. As with rabbit PLC, where

necessary to minimize cell clumping and insure a majority of single cells,

PEC were pushed through 19, 23 and finally 25 gauge hypodermic needles,

using a syringe.

Activation of Macrophages in Culture

LPS (Escherichia coli 055:B5, Westphal, Difco, Detroit, MI) was used

to induce macrophage activation in all cultures. Unless otherwise stated,

culture conditions were as follows: for mouse PEC or rabbit PLC and BLM,

LPS was added at 10 ug/ml in MEM, supplemented with 10% fetal bovine serum

plus .03M HEPES buffer; for rabbit BMC, cells were cultured in FMEM

supplemented with 10% fetal bovine serum plus .03M HEPES buffer. Further-

more, concerning BMC, the phorbol ester, 12-o-tetradecanoylphorbol

13-acetate (TPA) was added at 0.1 or 1 pg/ml and the cells generally

cultured overnight, after which time they were washed free of TPA (4 to 6

times) and refed with fresh medium containing LPS (see Increasing the

Macrophage Content of Bone Marrow Cultures, under Results). Unless other-

wise stated, all effector cells were cultured at 1 x 106 cells/ml.








21
Viability of all effector cell populations was determined by trypan

blue dye exclusion. All effector cells used had 95% viability or greater

by this criterion and were always cultured at a density based on the

percentage of viable cells.

Identification of Macrophages in Effector
Cell Populations

Identification of macrophages was based on morphological criteria

(Giemsa's blood stain), phagocytic capability of latex beads (Dow Chemical

Co., Indianapolis, IN) and peroxidase content (Leukocyte Peroxidase,

Histozyme Kit No. 390A, Sigma Chemical Co., St. Louis, MO).

Production of Tumor Necrosis Serum

The basic technique has been described by Carswell et al. (1975).

Rabbits were injected intravenously with viable BCG (Tice strain, 3 x

108 organisms). Two to three weeks later, LPS was injected

intravenously and the rabbits bled 2 to 3 hours later. The blood was

allowed to clot and the serum was withdrawn, further clarified by

centrifugation, and aliquoted. Aliquots were frozen at -200C until used.

Once thawed, tumor necrosis serum (TNS) could maintain cytotoxicity titer

(see assays) when kept at 4C for several months.

Production of Rabbit Macrophage Cytotoxins

Rabbit PLC, BLM or TPA-pretreated and washed BMC were cultured in

appropriate medium containing LPS. (Occasionally BMC were cultured in 175

cm2 plastic tissue culture flasks at 2 to 5 x 106 cells/mi to increase

titer yields. Otherwise, PLC and BLM were generally cultured at 1 x

106 cells/ml in 25 cm2 plastic tissue culture flasks or 24 well

plastic tissue culture plates.) After an appropriate time at 370C,

supernatants were harvested and cells removed by centrifugation.








Supernatants were titered and used within 48 hours if stored at 4C (no

appreciable loss of titer occurred if stored at this temperature for this

time) or frozen at -800C. Supernatants frozen at -800C maintained titers

for several months. Cytotoxins (CT) generated by rabbit macrophages are

designated as follows: pulmonary lavage cells (PLCCT), TPA-pretreated and

washed bone marrow cells (BMCCT), and blood mononuclear cells (BLMCT).

Cytotoxicity Assays
Photometric Assay

A basic photometric technique has been previously described (Ruff

and Gifford, 1980). L-929 cells were plated in 96-well tissue culture

trays at 50,000 cells/0.1 ml to establish a dense monolayer. Dilutions of

CT in 0.1 ml were added along with actinomycin D (Calbiochem, LaJolla, CA)

to give a final concentration of 1 ug/ml and the trays incubated at 370C

for 18 hours in a humidified 5% CO2 atmosphere. After the culture

period, the plates were stained for 10 minutes with crystal violet (0.2%

in 2% ethanol), washed with water and allowed to dry.

Destruction of monolayers was evaluated by placing trays on an X-ray

viewing screen and measuring light transmission through monolayers with a

photocell. The reciprocal of the dilution of CT required to kill 50% of

the targets was taken as the 50% endpoint (S50). Two kinds of

photocells were used in these studies. A manually operated photocell unit

was constructed using a photoconductive cell (cadmium sulfide, No. 05HL,

Clairex Electronics, Mt. Vernon, NY). The cell was cemented to a small

piece of copper tubing that had an outside diameter approximating the

inside diameter of culture wells. The photocell was attached to a

multipurpose meter and read as a variable resistance at a fixed voltage as

a function of light transmitted through the monolayers. For convenience









an automated photocell (Titertek Multiskan, Flow Laboratories, Inc.,

Inglewood, CA) was also used for some of the experiments. Figure 1

compares the two photocells reading the same sample wells. They are

fairly similar, although it appears that the automated photocell is

slightly more sensitive and can detect very small numbers of target cells

retained at lower CT dilutions that the manual photocell cannot.

Nevertheless, since an entire experiment was always read with the same

photocell, and the enhanced sensitivity deiiionstrated by the automated

photocell was not much greater than the manual one (about 15% greater, see

Figure 1), this difference in photocell sensitivity is insignificant.

This photometric technique was slightly modified to measure effector

cell killing of targets. The only difference from the aforementioned

procedure used to titer CT involved adding actinomycin D at 2 ug/ml for 2

to 3 hours, followed by removal of this drug from the medium by decanting,

washing the monolayer with medium and replacing with fresh medium

containing effector cells. It was previously determined that 2 washings

were sufficient to remove virtually all extracellular actinomycin D

(washings at this point could not inhibit H3-uridine incorporation into

L-929 cells).

Plaque Assay

L-929 cells were grown to monolayers in plastic 24 well flat bottom

tissue culture trays, plating 250,000 cells in 1 ml medium. After the

monolayers were pretreated with actinomycin D (2 ug/ml) for 2 to 3 hours,

they were washed twice with medium. Small numbers of well mixed effector

cells (mouse PEC or rabbit PLC) in 2-fold increments were added carefully

to washed and drained monolayers of targets in a final volume of 0.2 ml

medium. Control target monolayers received only 0.2 ml medium. No agar





























Figure 1. Comparison of automated and manual photocell in
the photometric cytotoxicity assay.

Each panel (upper and lower) represents titration of a
single sample by each of the photocells. Each point represents
percent target cell survival of a single culture well.

S = number of cells surviving in cultures, So = number
of cells plated.









* Automated Photocell


0 Monua I


Photocell


S 1 I I 1 1 I I 1


160 640 2560 10,240

[CYTOTOXIN DILUTION] -I


100-



80-



60-



40-


_J


rr


_J
_J
w
0



II
0
0


o( I)
OI


20-







100-



80-


60-



40-


20-









or methyl cellulose overlays were used. LPS was added at a final

concentration of 10 pg/ml to PEC and PLC cultures. Plates were cultured

undisturbed at 370C for 18 to 20 hours. Supernatant contents were then

decanted and the plates were stained for 10 minutes with crystal violet,

washed with water and allowed to dry. In all cases, at least 95% of the

effector cells plated were single cells and the number plated was based on

number found to be viable (always >95%).

Cr51 Release Assay

L-929 cells were suspended into two 1 ml aliquots of 2 x 106

cells. One aliquot was treated with actinomycin D for 1 hour. At that

time, 100 vCi sodium51 chromate (51Cr; ICN Chemical, Waltham, MA)

was added to each aliquot and the cells gently mixed by shaking. After 1

hour at 37C the target cells were washed 5 times by repeated centrifuga-

tion and resuspension in fresh medium (MEM with 10% fetal bovine serum,

gentamicin and HEPES buffer) to remove unincorporated 51Cr and from

one of the aliquots, actinomycin D. The concentration of the suspension

was adjusted with medium to 3 x 104 cells/0.1 ml. B16C3 cells were

labelled in a similar way only in the absence of actinomycin D.

These labelled targets were used to determine cytotoxicity of mouse

PEC (L-929) or CT generated by rabbit macrophage cultures (L-929,

B16C3).

Mouse PEC were seeded into 96 well flat bottom plastic tissue

culture plates in MEM with 15% fetal bovine serum, gentamicin, and HEPES

buffer at 3 x 105 cells in 0.1 ml volumes. Plates were centrifuged at

200 g for 30 seconds to sediment the cells to well bottoms. After 2 hours

at 370C in a humidified air atmosphere containing 5% CO2, the nonad-

herent cells were removed by decanting the supernatants and washing











the adherent PEC monolayers twice gently with warm medium. The great

majority of these adherent PEC, greater than 90%, had macrophage

morphology as revealed by staining with Giemsa's blood stain and could

phagocytize latex particles. After the second wash, 0.1 ml MEM with 10%

fetal bovine serum, HEPES buffer and gentamicin was added to each well.

In an additional 0.1 ml of similar medium, 51Cr-labelled targets were

added. LPS was then added. Plates were centrifuged at 200 g for 30

seconds to sediment target cells upon effector monolayers. Plates were

incubated at 370C, in a humidified air atmosphere containing 5% CO2. At

appropriate times for harvesting the supernatants, 0.05 ml of medium was

added to each well to insure an even distribution of released 51Cr

throughout the cultures. Cr51 labelled targets (L-929 or B16C3

cells) were added to microtiter wells along with TNSCT, PLCCT, or BMCCT.

Again, at harvest times, 0.05 ml of medium was added to wells to evenly

distribute released Cr51.

For harvest, aliquots of 0.1 ml were removed and counted in a gamma

counter (Gamma 300 Radiation Counter, Beckman Instruments, Inc.,

Fullerton, CA). For each time period, the percent specific 51Cr

released was determined by use of the following formula: percent release

= [(E-S 4 (T-S)] x 100, where E = counts released front labelled targets in

test cultures containing effectors and targets, S = spontaneous counts

released from labelled targets cultured alone and T = total releasable

counts from labelled targets, as determined by maximum lysis of targets

with 1% dodecyl sodium sulfate. Appropriate volume correction factors

were employed to calculate counts released in each case.









Column Chromatography
Gel Filtration

All gel filtration chromatography was performed using a single,

calibrated column of Sephacryl, S-200 (Pharmacia Fine Chemicals,

Piscataway, NJ) of gel bed dimensions 88 x 2.5 cm. Flow rate was 21 ml/hr

using phosphate buffered saline. Samples were concentrated by pressure

dialysis in an Amicon Diaflo ultrafiltration unit (Amicon Corp.,

Lexington, MA) prior to loading 1 ml samples onto the column. Fractions

(2 ml) were collected and assayed for cytotoxicity on L-929 cells.

Ion Exchange Chromatography

Fractions with cytotoxicity for L-929 cells from the Sephacryl S-200

column were pooled and concentrated by pressure dialysis and loaded onto a

column of DEAE-Sephadex (Pharmacia Fine Chemicals, Piscataway, NJ)

previously equilibrated with phosphate buffered saline (pH 7.2, 0.15 M).

The column was washed with 2 column volumes of starting buffer to remove

unbound proteins. A linear salt gradient (0.15 to 0.60 M) was applied to

the column and fractions (4 ml) collected and assayed for cytotoxicity.

Flow rate was 20 ml/hr.

All chromatography was performed at 4C.

Effect of Trypsin Inhibitors, Phenylmethyl Sulfonyl-
fluoride, Chloroquine and o-Phenanthrolene on
Rabbit Macrophage Cytotoxin Activities

Various drugs were tested in the actinomycin D cytotoxicity assay to

determine their effects on CT activity. The following were tested:

soybean trypsin inhibitor (SBTI, Worthington Biochemical Corp., Freehold,

NJ), bovine pancreatic trypsin inhibitor (BPTI, Sigma Chemical Co., St.

Louis, MO), at 100 pg/ml each; phenylmethyl sulfonylfluoride (PMSF) at

10-4M; chloroquine at 10-4M, and o-phenanthrolene at 10-5M

(last 3 drugs from Sigma Chemical Co., St. Louis, MO).










Effects of Prostaglandin E2 on Rabbit Macrophage
Cytotoxin Production

Prostaglandin E2 (PGE2, Sigma Chemical Co., St. Louis, MO) was

dissolved in ethanol and diluted in serum-free medium immediately before

addition to duplicate PLC cultures in serum-free medium. PGE2 was added

at 10-6M, 6 hours after initiation of the cultures with LPS (10

pg/ml). Controls received no PGE2. PGE2 antiserum (0.1 ml of lyoph-

ilized stock reconstituted with 2.5 ml buffer, Sigma Chemical Co., St.

Louis, MO) was added to another set of cultures at the same time as LPS.

Controls for this received no PGE2 antiserum. These latter experiments

were performed in the conventional MEM containing fetal bovine serum.

Effects of Actinomycin D on Rabbit Pulmonary
Lavage Cell Cytotoxin Production

Actinomycin D was added at 1 ug/ml at the same time and at various

times after addition of LPS to PLC cultures. Cultures were incubated for

30 hours, supernatants harvested and assayed.

Effect of Tunicamycin on Molecular Weight of Rabbit
Pulmonary Lavage Cell Cytotoxin

Tunicamycin (Sigma Chemical Co., St. Louis, MO) was added to test

PLC cultures at 0.5 and 5 pg/ml both 2 hours before and at the same time

as LPS addition. Cultures were incubated for 30 hours at 370C at which

time, supernatants were harvested and titrated for cytotoxic activities.

Cytotoxin was subjected to gel filtration chromatography on Sephacryl

S-200 (previously described).

Freeze-Thawing of Activated Rabbit Pulmonary Lavage Cells

Pulmonary lavage cells were cultured in the presence of LPS (10

vg/ml) and at various times, replicate cultures were terminated. From

some, supernatants were harvested and titered as previously described.









Other cultures were subjected to five freeze-thaw cycles in a dry

ice-ethanol bath. After freeze-thawing, the culture contents were

centrifuged and both supernatants and pellets titrated for cytotoxicity.

Susceptibility of Tumor Necrosis Serum Cytotoxin
Resistant L-929 Cells to Rabbit Macrophage
Cytotoxins and Pulmonary Lavage Cells

Matthews (1978) has demonstrated that it is possible to select for

TNSCT-resistant L-929 cells by growing L-929 cells in the presence of

small amounts of TNSCT. Using a lot of TNS which had an S50 titer

(See Cytotoxicity Assays, Photometric Assay) in the actinomycin D

cytotoxicity photometric assay of approximately 8000 units, L-929

resistant cells were selected for by growing cells in the presence of a

1:500 dilution of this CT. After about 2 weeks, during which time dead

cells were decanted and viable cells were refed twice a week with fresh

medium containing TNS, resistant cells were trypsinized, washed and plated

into microtiter wells (10,000/well). Conventional L-929 cells were plated

into another set of wells. After allowing cells to adhere overnight in

conventional medium, TNSCT, PLCCT, BMCCT, or PLC (100 or 1000 cells) were

added to test wells (control wells received only medium). In those wells

containing PLC, LPS (10 pg/ml) was added. After 48 hours, plates were

stained and read as previously described.

Time Lapse Cinematography

L-929 cells were plated into 25 cm2 tissue culture flasks and

allowed to adhere and form a monolayer overnight. The cells were

treated with actinomycin D (2 yg/ml) for 2 to 3 hours, then washed

extensively. Rabbit PLC were added in small numbers (10 x 103 to 50 x

103) to the washed monolayers in 3 ml medium at which time, LPS was

added (10 ug/ml). Plates were allowed to equilibrate for an hour in a

lucite incubator chamber mounted on an inverted phase contrast microscope










connected to a time lapse motion picture camera. A temperature of 370 to

38C was maintained within the lucite chamber by a heated air curtain

with a thermostat sensing device. An appropriate field was located

(containing as few macrophages as possible). The field of view was

photographed through a green filter. Photography was generally commenced

4 to 6 hours after addition of effector cells and LPS. Exposure rate for

the first 2 to 3 hours of filming was 1 frame/15 seconds. Generally after

that time, exposure rate was increased to 1 frame/8 seconds. Several

films were made at magnifications of 150 and 200 X. Films were viewed at

24 frames/second.














RESULTS

Percent Macrophages in Rabbit Effector Cell Populations

Table I illustrates the percentage of cells that are macrophages

from three sources, in several rabbits, as judged by morphological and

phagocytic criteria. Pulmonary lavage cells represented a population

made up almost exclusively of macrophages (90-95%). Blood mononuclear

cells were 12-20% macrophages while bone marrow cells contained no more

than 0.1-0.5% macrophages.

Increasing the Macrophage Content of Bone
Marrow Cell Cultures

It is obvious from Table I that bone marrow cells provided only

negligible amounts of macrophages. Lotem and Sachs (1979) have demon-

strated that the phorbol ester, TPA, can cause both mouse and human

myeloid leukemic cells to differentiate into cells with varying degrees

of several macrophage characteristics (adherence to surfaces, Fc and C3

receptors, phagocytic capability, lysozyme activity, and macrophage

morphology).

Treatment of rabbit bone marrow cell cultures with TPA (1 ug/ml)

resulted in adherence of many cells to the culture dish by 30 minutes.

By 5 hours, over 90% of the cells were adherent (Table II). In addition,

about one-third of the adherent cells had macrophage morphology and

phagocytized latex beads. Twenty percent stained positive for

peroxidase. (Later studies indicated that 0.1 ug/ml TPA was adequate to

cause similar differentiation changes in rabbit bone marrow cells.)












TABLE I

RABBIT CELL POPULATIONS UTILIZED AS MACROPHAGE SOURCES


Percent
Source Macrophagesa


Pulmonary Lavage Cells 90 95

Blood Mononuclear Cells 12 20

Bone Marrow Cells <1

aMacrophage morphology (Giemsa's blood stain) and phagocytosis of latex
beads.











TABLE II

PERCENT RABBIT BONE MARROW CELLS WITH MACROPHAGE PROPERTIES BEFORE AND
AFTER TPA TREATMENT



Percent of Cells
Macrophage Property Before After TPAa


Adherence <1 92b

Morphology <1 32C

Phagocytosis (latex beads) NTd 35c

Peroxidase positive NT 20C


a5-28 hours after adding TPA at 1 yg/ml.

percent of total bone marrow cells added to culture.

CPercent of total adherent bone marrow cells.

dNot tested.








Considering the variable degrees of each property present, this

process can be envisioned as a spectrum of variable differentiation in

the direction of the mature macrophage state.

In Vitro Cytotoxin Production by Rabbit Macrophages

Requirement for Activating Agent

Matthews (1978) has reported on the apparently consistent, spontane-

ous production of CT by rabbit BLM cultures in the absence of known added

activating agent (e.g., LPS). These studies did not determine the possible

presence of contaminating LPS in the culture medium. This is often a prob-

lem in studying macrophages and can lead to serious misinterpretation of

experimental results (Fumarola, 1981). Figure 2 demonstrates the results

from experiments employing rabbit PLC and BLM cultures, where no attention

was given to the possibility of contaminating LPS. Depending upon the time

cell supernatants were harvested, the cytotoxin titers in the cultures to

which LPS was added (1 ug/ml) were consistently greater than titers gener-

ated in cultures to which no LPS was added (3 to 20 times greater for BLM,

10 to 50 times greater for PLC, replicate cultures varied from mean values

by 50%).

Using the Limulus amebocyte lysate test (Levin et al., 1970), it was

determined that the culture system used for the experiments demonstrated in

Figure 2 contained LPS (>0.025 ng/ml). Table III illustrates the various

materials tested. The recycled glassware used to store prepared media

tested positive for LPS. This was probably contaminated with the LPS-

positive bovine serum lot (Gibco) also shown in Table III. Other than the

glassware and Gibco bovine serum, only the trypsin stock concentrate

(Microbiological Associates) and thioglycollate broth base powder (Difco)

tested positive in the Limulus assay.

If no LPS was added to culture systems testing negative for LPS

(<0.025 ng/ml) by the Limulus assay, rabbit PLC generally produced no CT






























Figure 2. Rabbit macrophage cytotoxin production with and
without added LPS.

Cytotoxin production by rabbit macrophage cultures in the
presence (1 ~g/ml) or absence of added LPS. Culture
supernatants were harvested for assay at times indicated. Blood
monocytes were cultured at 4 x 106/ml, pulmonary lavage cells
at 1 x 106/rl.







37


PULMONARY LAVAGE CELLS
-0- -0--
LPS ADDED NO LPS ADDED

4 BLOOD MONOCYTES

LPS ADDED NO LPS ADDED







3







- /
I /
55

I0 /
2- / / -











/
0/
I-I



5 0











5 10 15 20 25


HOURS











TABLE III

ITEMS TESTED WITH LIMULUS AMEBOCYTE LYSATE REAGENT



Test
Item Resultsa


Bovine calf serum
Gibco; lot No. R290222 +
Sterile Systems, Inc., lot No. 400168 0

Fetal bovine serum
K.C. Biol., lot No. 309447 0

Trypsin
Stock concentrate, Microbiol. Assoc. +
Crystalline, Sigma; bovine, porcine pancreas 0

Eagle's minimum essential medium, base concentrate 0

Glutamine, stock concentrate 0

Gentamicin, stock concentrate 0

HEPES buffer, stock powder 0

Deionized water from ion exchange tap 0

Thioglycollate broth, base powder, Difco +

Reusable autoclaved laboratory glasswareb +

aLimulus lysate test results interpreted as follows: (+) = contains at
least 0.025 ng/ml LPS; (0) = contains less than 0.025 ng/ml LPS.
bSterilized glassware rinsed with pyrogen free water; washings used for
Limulus assay.







(Figure 3). However, one out of the ten rabbits sacrificed for these

studies (only four rabbits are shown in Figure 3) had PLC that could

produce detectable levels of CT in LPS-free cultures. Pasturella species

was cultured from lung washings from this rabbit as well as from its PLC

cultures. The presence of a Gram negative infection in the lungs probably

resulted in in vivo activation of PLC. Despite this problem, the great

majority (90%) of the rabbits utilized provided PLC that would not produce

CT unless LPS was added to the culture (Figure 3). As little as 0.1 to 1

ng/ml could induce CT production. As a precaution, PLC removed from all

rabbits to be utilized in other experiments (other than to merely generate

CT for physicochemical studies) were tested in LPS-free culture systems for

their ability to produce CT in the absence of added LPS. This insured that

CT was not produced unless LPS was added to the culture system, allowing

control over PLC activation.

As previously mentioned, many agents such as BCG (Old et al., 1960),

Corynebacterium parvum (Halpern et al., 1974), etc., can cause a general-

ized macrophage hyperplasia in vivo. Leake and Myrvik (1968) have demon-

strated this locally in the lungs of rabbits that were injected intraven-

ously with BCG in mineral oil. They reported up to a 10-fold increase in

the number of PLC obtained from rabbits injected intravenously over 1 week

before sacrifice. They also reported qualitative changes in the PLC (an

increase in cell lysozyme content as well as ultrastructural changes).

Figure 3 shows that PLC from rabbits injected intravenously with complete

Freund's adjuvant 2 to 3 weeks previously, produced no CT in the absence of

added LPS and variable amounts of CT per 1 x 106 cells cultured in the

presence of different levels of LPS. In addition, the normal rabbit shown

(not administered Freund's adjuvant) produced significantly greater levels

of CT at lower doses of LPS. Other experiments employing single dose

levels of LPS in cultures of PLC demonstrated variable and overlapping






























Figure 3. Rabbit macrophage cytotoxin production in the
absence and presence of varying levels of exogenous LPS.

Pulmonary lavage cells from normal rabbits, or rabbits
primed with Freund's complete adjuvant 2 to 3 weeks previously,
were cultured in the absence or presence of varying levels of
LPS. Supernatants were harvested after 24 hours. Culture
systems contained <0.025 ng/ml LPS prior to LPS addition.
















O Normal Rabbit
* Primed Rabbit a
A Primed Rabbit b
* Primed Rabbit c


(\J
'0
x


IC
t(5






LU

w
x
o
I-

0
I-

1-,


0 0.1 I 10 100


1,000 10,000


LPS CONC., ng/ml







results for CT production by LPS-activated PLC from normal and primed

rabbits (data not shown). From this, it appears that both primed and

normal rabbit PLC required the presence of LPS to produce CT in vitro. In

addition, CT production by each cell type occurred to a similarly variable

extent. Primed rabbits yielded 4 to 10 times more PLC (up to 200 x 106

cells) than normal rabbits, 80 to 90% of which had macrophage morphology

(data not shown). In addition, LPS added to PLC cultures from primed or

normal rabbits in amounts exceeding 10 ug/ml produced no more CT than at 10

ug/ml LPS (data not shown).

As with PLC, TPA-pretreated and washed BMC would not produce CT in

LPS-free cultures unless LPS was added. In addition, no CT was produced

from bone marrow cells that were not pretreated with TPA although exposed

to LPS (bone marrow contains <1% macrophages, see Table I). Optimum CT

production resulted when BMC were pretreated with TPA (0.1 and 1 Ug/ml

yielded similar results) for a minimum of 7 hours, then washed extensively

(4 to 6 times) prior to activation with LPS (data not shown). This demon-

strates the prerequisites for production of significant levels of CT by

BMC. These are differentiation of sufficient bone marrow precursor cells

to macrophages and subsequent activation of these macrophages by LPS.

Taking advantage of the strong and efficient adherence of BMC

induced by TPA, BMC were plated at varying densities to determine the most

efficient plating density for CT production (Table IV). Cultures producing

the most CT had cells plated at approximately 0.5 x 106 and 1.5 x

106/cm2. This range of plating densities yielded consistently high CT

titers for PLC as well (data not shown).

It was important that the TPA-pretreated BMC were washed extensively

prior to adding LPS. Initial experiments with BMC were performed by adding

LPS to TPA treated BMC without washing out the TPA. Virtually no CT was









TABLE IV

RABBIT BONE MARROW CYTOTOXIN PRODUCTION: OPTIMUM PLATING DENSITYa


Number Cells S50Titerb
x 106/cm2 S50Titerb per 10 x 106 cells


8.80 2290 104

5.92 2500 169

2.96 2500 338

1.48 2089 564

0.54 933 691

aCells were cultured in MEM with 10% fetal bovine serum and 1 pg/ml TPA
for 12 hours. TPA was removed by washing cultures extensively and LPS
(1 ug/ml) was added.
b30 hour titers.







produced by these cultures. In order to determine if TPA could have an

inhibitory effect on the CT, TPA was added to a TNSCT assay (Table V). A

linear relationship was found between the logarithm of the amount of TPA

added to the assay and the decrease in the titer of TNSCT. Pretreating the

target cells (L-929) used in the assay with 100 ng TPA for 18 hours and

then washing out the TPA prior to adding TNSCT to the cultures resulted in

less inhibition of TNSCT titer (21% instead of 36% where only 2.4 ng TPA

was left in the assay with TNSCT). These results demonstrate that TPA

exerts an inhibitory effect on CT activity in general.

Comparison of Cytotoxin Production Capabilities
by Rabbit Macrophages Derived from the
Different Sources

Rabbit PLC, BMC and BLM were cultured under conditions that optim-

ized CT production for each of the cell types. BMC cultured in a fortified

medium (see Materials and Methods) generated about 4 times more CT than if

cultured in the standard medium used for PLC and BLM. This was done to

determine which of the cell types had the greatest potential for generating

highest CT titers. Based upon 1 x 106 cells cultured per ml of medium,

PLC were most efficient under the tested conditions (Table VI, 20 to 140

times more CT than BMC or BLM produced). Also as previously seen regarding

sensitivity to LPS, PLC demonstrated much variability (about 7-fold) when

comparing different rabbits with each other. BMC demonstrated much less

variability (about 2-fold).

The overall cell populations used varied greatly in their macrophage

content (Tables I and II). In addition, the BMC demonstrated a broad

range of cells with macrophage characteristics depending upon which prop-

erty was being utilized to call a cell a macrophage. It was therefore

necessary to normalize the different cell populations for their macro-

phage content to compare the ability of macrophages from these











TABLE V

EFFECT OF TPA ON TUMOR NECROSIS SERUM (TNS) TITER


Percent Inhibition
Experiment TPA(ng) TNSCT Activity


TNS titrated in presence of TPAa 2.4 36

9.75 48

78 70

TNS titrated on TpA-pretreated and 100 21
washed targets


aTPA was added to the standard actinomycin
assay (see Materials and Methods).


D serum cytotoxin photometric


bL-929 cells were preincubated with TPA for 18 hours, then washed. Serum
cytotoxin photometric assay was then performed.










TABLE VI


CYTOTOXIN PRODUCTION BY RABBIT MACROPHAGES FROM BLOOD
PRETREATED BONE MARROW AND PULMONARY LAVAGE CELLS:
TITERS PRODUCED PER 1 x 106 CELLS


MONONUCLEAR, TPA-
COMPARISON OF


S50Titer
Cell Typea Rabbit per 1 x 106 Cells

Pulmonary lavageb 1 14,000
2 4,562 526
3 2,200 100

TPA-pretreated bone marrow 4 70 4d
5 77 6d
6 119 Od

Blood monocytesb 1 79


aAll cells derived fran normal rabbits and cultured at 1 x 106/ml in
medium containing 10% fetal bovine serum, LPS at 10 ug/ml and 3 x 10-2M
HEPES buffer. Plating densities, 0.5-1.5 x 10 /cmL.

bCultured in Eagle's minimum essential medium.

cCultured in Eagle's minimum essential medium fortified with 2 X vita-
mins and essential amino acids, nonessential amino acids and sodium
pyruvate.
dCorrected for loss of bone marrow cells that do not adhere following
TPA-pretreatment (<10%).








47

different sources to generate CT. Considering the potential ranges of

percent macrophages in each population as well as the extremes of CT

titer produced by each of the cell types, the PLC were still most

efficient at generating CT (see Table VII) on a basis of 1 x 106

macrophages/ml culture medium (4 to 250 times greater than BMC and 4 to

40 times greater than BLM).

Effects of Bovine and Fetal Bovine Serum on
Rabbit Macrophage Cytotoxin Production

The cytotoxic capabilities of macrophages can be enhanced or

inhibited by various serum components (Hibbs et al., 1977; Chapman and

Hibbs, 1977). Mechanisms which could help explain this phenomenon may

involve production or function of macrophage CT. PLC were cultured in

the absence or presence of varying levels of bovine or fetal bovine serum

(Figure 4). Under the conditions tested, it can be seen that PLC could

generate CT in serum-free medium (PLC from two different rabbits generated

330 to 400 S50 CT units under the culture conditions described for

Figure 4, panel a). Bovine serum was inhibitory to CT production at levels

generally greater than 4%. Fetal bovine serum enhanced CT production by

PLC up to 3-fold (at 10%). Figure 5 further illustrates that 10% fetal

bovine serum was more effective than lower levels (1%) at potentiating

LPS-induced PLCCT production. This was the case at 3 levels of LPS

concentration. It can also be concluded that to the extent to which they

were tested, LPS at 10 ug/ml with 10% fetal bovine serum were optimal

together for potentiating the activation of PLC by LPS.

Time Kinetics of Macrophage
Cytotoxin Production
Figure 6 demonstrates in vitro production of CT by LPS-activated

PLC with respect to time. Generally, significant levels of CT appeared











TABLE VII

DATA FROM TABLE VI CONSIDERING PERCENT MACROPHAGE
CONTENT OF EACH CELL TYPE


Percent Theoretical Range of
Macrophages S50Titer per
Cell Type Possible Range 1 x 106 Macrophages


Pulmonary lavage 90 95 14,737 15,556
2,316 2,444

TPA-pretreated bone marrow 20 100a 70 350
119 595

Blood monocytes 12 20 395 658

aConsidering all macrophage properties of TPA-pretreated and washed
adherent bone marrow cells (e.g. peroxidase staining positive cells
represent 20% of total adherent cell population. See Table I).





























Figure 4. Effects of bovine and fetal bovine serum on
pulmonary lavage cell cytotoxin production.

Rabbit pulmonary lavage cells were cultured for 24 hours
in the absence or presence of various levels of bovine or fetal
bovine serum. Experiments employing cells from 2 rabbits are
shown: a. LPS present at 1 Ug/ml; b. LPS present at 10 ug/ml.
Control cultures contained no serum.











W No Serum

y Bovine Serum

Fetol Bovine Serum


200
U,
w
to




x ,

0
0 2 5
o % SERUM

b
-J
0 100-
Z
z
0
80-

0t
60-



40



20



0 4 8 16 32

% SERUM


10






























Figure 5. Comparison of effects of 1% and 10% fetal
bovine serum on cytotoxin production by rabbit pulmonary lavage
cell cultures containing various levels of LPS.

Supernatants were harvested from triplicate cultures that
were incubated for 30 hours.











Fetal Bovine Serum:

EI % D 10o%


5-

0
x
4-
w


0
Ln 3
1-


z

0 2
0









3.3 10


p.g/ml


LPS,






























Figure 6. Time kinetics of rabbit pulmonary lavage cell
cytotoxin production.

Pulmonary lavage cells from a nonmal rabbit or rabbits
primed with Freund's complete adjuvant 2 to 3 weeks previously
were cultured in the presence of LPS. Supernatants were
harvested at times indicated.










o PRIMED
o PRIMED
A NORMAL


RABBIT I
RABBIT #2
RABBIT


HOURS


3


Q:
w
I-
I-
0
u)

Z2
x
0
I-
0
0-
C-)
0o


O


5 10 15 20 25 30









by 3 hours and titers increased significantly between 3 and 5 hours.

Levels remained fairly constant at times tested between 5 to 7 and 18

hours and then increased significantly again after about 18 hours in

culture. Variation of replicate cultures was mean 20% for the Freund's

adjuvant primed rabbits and mean 50% for the normal rabbit. Figure 6

also demonstrates the variation among rabbits (normal and primed) in the

CT titers generated by their PLC at any single given time. Other experi-

ments involving culturing PLC from several normal and primed rabbits under

identical conditions for the same time resulted in overlapping and simi-

larly variable CT titers for both groups (data not shown).

Time kinetics of CT production by TPA-pretreated and washed, LPS-

activated BMC are given in Figure 7. Detectable levels of CT (20 to 26

S50 units) appeared by 3 hours and increased through 14 hours. Levels

measured at 14 and 30 hours were similar.

Time kinetics of CT production by LPS-activated BLM are included in

Figure 2. Detectable levels of CT were present by 3 hours, increased

between 3 and 7 hours and were similar at 7 and 24 hours. These observa-

tions are consistent with the data of Matthews (1978). From these observa-

tions, it is apparent that time kinetics of CT production by PLC, BMC and

BLM are somewhat different after an initial production time during the

first 3 hours. After 30 hours culture, essentially no CT was produced by

PLC and BMC that were refed with fresh LPS-containing medium (data not

shown).

Effects of Actinomycin D on Cytotoxin
Production by Rabbit Pulmonary Lavage Cells

Cameron (1980) has described an actinomycin D-induced inhibition of

cytotoxicity of human macrophages for tumor cells. The data presented





























Figure 7. Time kinetics of cytotoxin production by
TPA-pretreated and washed rabbit bone marrow cells.

Rabbit bone marrow cells were plated into 24-well tissue
culture dishes and treated with TPA (0.1 ug/ml) overnight.
After removing TPA from cultures (washed 5 times), fresh medium
containing LPS (10 /g/ml) was added. Supernatants were
harvested at various times.















Rabbit A
ARobbit B
120




100-



cr
w
n-
S80





x 60-





40-
I.--








20-






5 10 15 20 25 30

HOURS









in Figure 8 show that actinomycin D (1 jg/iml) consistently inhibited

PLCCT production if added to cultures at the same time as LPS.

(Inhibition of CT titers of test cultures was greater than 95% below

those of control cultures to which no actinomycin D was added.) Delaying

addition of actinomycin D resulted in a decrease of inhibition of CT

titers. This varied considerably among rabbits, especially when compar-

ing primed to normal rabbits. From the data of Figure 8, it is apparent

that in no case was there active CT messenger RNA or active CT in PLC

prior to the addition of LPS to the cultures. In addition, active CT

messenger RNA was fully synthesized by 2 to 6 hours after LPS activation

of primed PLC and about 3 hours after LPS-activation of nonnal PLC.

Effects of Prostaglandin E (PGE) and Prostaglandin
Antiserum on Cytotoxin Production by Rabbit
Pulmonary Lavage Cells

It has been demonstrated that PGE can be produced by macrophages in

culture and can shut off macrophage cytotoxicity for tumor cells

apparently after a period of macrophage activation (Taffet and Russell,

1981). Figure 9 demonstrates that PGE (10-6 M) added to PLC cultures

6 hours after LPS (lOug/ml) resulted in inhibition of CT production down

to about 30% of control culture levels. In addition, adding PGE anti-

serum to cultures at the same time as LPS increased CT production by 50%

over the control levels. These data are consistent with the conclusions

of Taffet and Russell (1981) concerning regulation of macrophage cytotox-

icity by endogenously produced PGE.

Disruption of Rabbit Pulmonary Lavage Cells at Various
Times after LPS Activation: Failure to Demonstrate
Significant Active Intracellular Cytotoxin Pool

Unanue and Kiely (1977) have demonstrated that a macrophage

protein, mitogenic for thy.mocytes, accumulates within the macrophage in






























Figure 8. Effects of actinomycin D on cytotoxin
production by rabbit pulmonary lavage cells.

Actinomycin D (l g/ml) was added to cultures of rabbit
pulmonary lavage cells at the same time and various times after
addition of LPS. Culture supernatants were harvested for assay
after 30 hours. Panel a represents 2 primed rabbits, panel b, 2
normal rabbits.








O Primed Robbit I
APrimed Rabbit 2


* Rabbit I
A Rabbit 2


I 2 3
HOURS ACTINOMYCIN D


4
ADDED


5 6
AFTER LPS


200


150


100


w

I-


0
O





0
I-
0
0


50







150


100



50





























Figure 9. Effects of prostaglandin E2 and prostaglandin
antiserum on cytotoxin production by rabbit pulmonary lavage
cells.

Rabbit pulmonary lavage cells were cultured in the
presence of LPS at lOug/ml. To 1 set of cultures in serum-free
FMEM, prostaglandin E2 was added at 10-6 M. To another
set of cultures in standard MEM supplemented with 10% fetal
bovine serum, prostaglandin antiserum was added (see Materials
and Methods). Cultures were incubated for 28 hours and
supernatants harvested.














Control
Cultures


Test
Cultures


PGE
CONTROLS


PGE


CONTROLS


ANTI-PGE









the form of an intracellular pool. Only a small portion of this pool may

be secreted by the macrophage depending upon its state of activation. By

repeated freeze-thawing of thioglycollate-stimulated mouse PEC, they

could demonstrate higher levels of mitogenic protein within the cell than

were secreted under certain conditions.

Rabbit PLC were activated by LPS and at various times after LPS

addition (30 minutes, 1, 3, 7 and 24 hours), culture supernatants and

freeze-thawed lysates were analyzed for CT. As previously, supernatant

CT became measurable at 3 hours but at no time, except one (1 hour after

addition of LPS) did freeze-thawed lysate contain measurable CT (data not

shown). At 1 hour after LPS addition to cultures, only about 2 S50

units of CT appeared in freeze-thawed lysates while none appeared in the

culture supernatant. Appropriate controls indicated total stability of

PLCCT to repeated freeze-thawing in this culture system. In addition,

assaying the freeze-thawed PLC (none viable by inability to exclude

trypan blue dye) on L-929 cells in the presence of actinomycin 0 revealed

no cytotoxicity. These studies indicate that no significant levels of

active freeze-thaw-releasable CT were retained by the activated

macrophage.

Physicochemical and Biological Properties of Rabbit
Macrophage Cytotoxins and Tumor Necrosis
Serum Cytotoxin
Gel Filtration Chromatography

All 4 of the cytotoxins were eluted through the same Sephacryl
S-200 column. All eluted as single peaks. The elution volumes for all

corresponded to a molecular weight of about 48,000 D (Figure 10).






























Figure 10. Gel filtration of rabbit macrophage cytotoxins
on Sephacryl S-200.

Macrophage culture supernatants were concentrated by
pressure dialysis prior to gel filtration. The column was
calibrated with standards of known molecular weight,
ribonuclease (not shown), chymotrypsin (C), ovalbumin (OV),
bovine serum albumin (BSA).








65





*TNS
OPLC Cytotoxic
ABMC Activities
ABLM

BSA OV C
50-

40

30

20-
10-


E 24
(r
W 18

f 12

6-
U)

z 320-
x
o 240- 0.6
o
S160- 0.4

80- -0.2


150 125 200 225 250 275 300


ELUTION VOLUME (ml)









Ion Exchange Chromatography

The 4 cytotoxins were eluted through freshly poured DEAE-Sephadex

columns with gel beds of approximately similar size. All eluted with the

same linear salt gradient between 0.28 and 0.32 M NaCl (Figure 11).

Effect of Tunicamycin on the Molecular Weight
of Cytotoxin Generated by Rabbit Pulmonary
Lavage Cells

Previous studies failed to demonstrate significant levels of

exposed carbohydrate on rabbit TNSCT (Matthews et al., 1980; Ruff and

Gifford, 1981). These studies drew their conclusions fran the inability

to bind TNSCT to a variety of sugars and plant lectins.

Tunicamycin has been shown to inhibit N-glycosylation of

glycoproteins by blocking formation of N-acetylglucosamine-lipid

intermediates (Hickman et al., 1977). This has been demonstrated for

several types of glycoprotein including interferon (Fujisawa et al.,

1978) and immunoglobulin (Hickman et al., 1977). Rabbit PLC were

cultured with LPS both in the presence and absence of tunicamycin.

Tunicamycin at 1 ug/ml had no effect on production of CT, but at 5 ug/ml,

CT titers were decreased by 30% when compared to control cultures which

contained no tunicamycin. As high as 5 ug/ml tunicamycin had no effect

on the molecular weight of CT produced as determined by gel filtration

chromatography on Sephacryl S-200 (data not shown). This demonstrates

that CT has no covalently linked carbohydrate or has a covalently linked

carbohydrate of very small molecular weight or could have a carbohydrate

portion of substantial size that is covalently linked to the rest of the

molecule by a tunicamycin resistant mechanism.






























Figure 11. Ion exchange chromatography of rabbit
macrophage cytotoxins on DEAE-Sephadex.

Rabbit TNS or pooled cytotoxin containing fractions from
the Sephacryl S-200 column were concentrated by pressure
dialysis and loaded in starting buffer. Unbound proteins were
washed off the column, followed by elution with a linear salt
gradient from 150 to 600 mM phosphate buffered saline.







68



TNS
o PLC CYTOTOXIC
ABMC ACTIVITIES
ABLM


40 600

30 450

20 300

10 150


4 20 40 60 80
r u600 !
S3 450
F- 2
-_ 300

Lo 6 450
O 4 .-------- 50
S20 40 60 80 600 Z
6 0
0 6- 450
F-A
O 4- --- 300-
H --2-- 0 O
>- 2 x.-x-- 150 o
0 z

4 20 40 0 o600

3 -450
2~ -x
2 300

I X-------- 150

20 40 680
FRACTION NUMBER









Temperature Stability of Rabbit
Macrophage Cytotoxins

Each of the cytotoxins were subjected to 56% and 70C in separate

experiments (Figure 12). All four of the cytotoxins were stable to 560C

for 60 minutes. In addition, the titers of all were greater than the

controls, which were maintained at 40C during the course of the

experiment. This may suggest a heat labile inhibitor of CT present. At

70C, all cytotoxins demonstrated decay kinetics that were fairly similar

and by 60 minutes, all lost about 80% of activity compared to controls.

Effect of Serine Protease Inhibitors and
o-Phenanthrolene on Cytotoxin Activities

There is evidence in the murine system that serine proteases may

play a role in macrophage mediated tumor cytolysis (Adams, 1980; Adams et

al., 1980). Rabbit PLCCT was not inactivated by phenylmethyl sulfonyl

fluoride (10-4M), an inhibitor of serine proteases (data not shown).

In addition, neither bovine pancreatic trypsin inhibitor nor soybean

trypsin inhibitor inactivated PLCCT or BMCCT activity (data now shown).

None of these enzyme inhibitors inactivated TNSCT either, consistent with

the findings previously reported (Ruff and Gifford, 1981). On the other

hand as this latter group has reported for TNSCT, o-phenanthrolene, a

metal chelator had significant inhibitory effects on the macrophage CT

(Table VIII).

Effect of Chloroquine on Rabbit
Macrophage Cytotoxin Activities

Chloroquine is a drug which can apparently stabilize cell lysozomes

(Lie and Schofield, 1973). Previous work with TNSCT (Ruff and Gifford,

1981) has demonstrated an inhibitory effect of this drug on cytotoxin

activity in the actinomycin D assay. All 4 cytotoxins were tested in the

standard actinomycin D assay with chloroquine (10-4M) and all were

inhibited (Table VIII).







































Q4-






0
E






C)








4-)



4-)
c-








(0


4-)
0
o-)
(0


I-
SI
0



4-J
r-













C-)
4-
0















E-.







3 tn

o
+>
0
co
j


0)
L*
3 .4) *
4> CL



.C E.
r- 0 -
r4-




o o

- 0 4-


C/ N,
-o Lo



S *-
OC






0
4O S-
to tc
C 0- (0
4-0 >-







C C


(0 S- to

= 4- s
-3 3




On 0 L
-4--) :-
4-


00

Excc
4-)







cU 0 /)





*r 4.)
-T C
U0
E c
























o 2
2-J
m m






z) U
zJ
2 CL


"1OdlNO3 %


I II I I a 1 I 1
o 0 0 0 0 0 0 0
0CI O O (D cj


I


t


IL^^^$^


63111 OSs









TABLE VIII

EFFECTS OF CHLOROQUINE AND o-PHENANTHROLENE ON CYTOTOXIN ACTIVITY


Percent of Control S50Titer
Drug PLCa BMCb BLMC TNSd


Chloroquine, 10-4M 55 38 37 31

o-Phenanthrolene, 10-5M 46 48 35 50


Cytotoxins from:
aPLC, pulmonary lavage cells

bBMC, TPA-pretreated bone marrow cells

CBLM, blood monocytes
dTNS, tumor necrosis serum









Effects of High Arginine Levels on Rabbit
Macrophage Cytotoxin Activities

Many reports indicate that arginase plays some role in macrophage

mediated tumor cytotoxicity (Currie and Basham, 1975, 1978; Currie, 1978;

Chen and Broome, 1980). These same reports indicate inhibition of

macrophage merdiated tumor cytotoxicity by adding excess arginine to the

medium. Arginine added to our cytotoxin assay at 2 x 10-2M (5 to 10

times the level employed by these latter investigators) did not inhibit

cytotoxicity of any of the CT (data not shown). From this, we conclude

that none of our cytotoxins is arginase.

Relative Titers of Rabbit Macrophage Cytotoxins
in the Presence and Absence of Actinomycin D
in the Cytotoxicity Assay

Ostrove and Gifford (1979) have reported an enhanced sensitivity of

L-929 cells to TNSCT in the presence of actinomycin D. Similar results

were demonstrated with PLCCT and BMCCT. Target cells demonstrated up to

250 times more sensitivity to CT in the presence of actinomycin D than if

actinomycin D was not employed (data not shown).

Target Cell Susceptibility Profiles of Rabbit
Macrophage Cytotoxins and Tumor Necrosis
Serum Cytotoxin

Table IX compares PLCCT and BMCCT with TNSCT target cell

susceptibility. Both macrophage derived cytotoxins are not toxic for

mouse embryo fibroblasts or B16C3 melanoma, two types of TNSCT-

resistant cell, while L-929 cells, a TNSCT-sensitive cell, is also

sensitive to macrophage CT. Chromium51 release studies demonstrated

L-929 sensitivity to PLCCT and BMCCT, while B16C3 cells were resistant.

L929 cells released up to 18% of their label by 8 hours (taking into

consideration spontaneous release of label by target cells) while B16C3

cells released none of their label by 21 hours (data not shown).










TABLE IX

CELL SUSCEPTIBILITY PROFILES OF RABBIT CYTOTOXINSa


Cytotoxin
Target Cell Source S50Titerc


L-929 TNSb >128
PLCb >128
BMCb >128

Mouse embryo fibroblasts TNS <2
PLC <2
BMC <2

B16C3 TNS <2
PLC <2
BMC <2

aphotometric assay without added actinomycin D was used to titer the
same aliquot of each cytotoxin on the cells indicated.
bTNS, tumor necrosis serum; PLC, pulmonary lavage cells; BMC, bone
marrow cells (TPA pretreated).


cForty-eight hour titers.







75
Table X summarizes biological and physicochemical characteristics

of rabbit macrophage cytotoxins and TNSCT.

Cocultivation of Mouse Peritoneal Exudate Cells or
Rabbit Pulmonary Lavage Cells with Actinomycin D
Pretreated L-929 Targets
Target cells can be rendered increasingly sensitive to TNSCT (a

putative macrophage product) by actinanycin D (Ostrove and Gifford, 1979;

Ruff and Gifford, 1981). In addition, Kunkel and Welsh (1981) have

reported that pretreatment of L-929 cells with actinomycin D followed by

washing rendered them increasingly sensitive to lysis by natural killer

cells. Considering these observations, it was of interest to determine

what effect actinomycin D pretreatment of L-929 cells followed by washing

would have on macrophage mediated killing of L-929 targets. Table XI

illustrates increased sensitivity of L-929 cells to mouse PEC cytotox-

icity using a 51Cr release assay. By 6 hours, nearly 40% of the

51Cr was released by actinomycin D pretreated targets cocultured with

LPS activated PEC, while by 12 hours, those targets not pretreated with

actinomycin D released less than 10% of their label in the presence of

effector cells. In this experiment, the effector:target cell ratio was

between 1:1 and 10:1 (see Materials and Methods).

Demonstration of Cytotoxicity of Actinomycin D
Pretreated L-929 Cells with Effector:Target Ratios <1

Actinomycin D pretreatment and washing of monolayers was followed

by addition of small numbers of mouse PEC or rabbit PLC activated by the

addition of LPS. Significant levels of killing could be detected by 18

hours using the photometric assay (see Figure 13). If actinomycin D was

omitted, no killing was detected at 18 hours. It was possible to

determine the number of effector cells required to kill 50% of the






















-1 CO Q 0 Q 0

u/ 0 ho




0Q 0 U-)
I 4- -
Z I- CO o'D E
CO 0 0
-- I 0 .
I *.e .. .-


I- +3) m O ,- u .
0 U)C CO -0
o0 0 "> f- -*- 0 "0
:C 0) CO LUC-- 0 u
0- X L.. ciJ 4- --- C 3
SCO 0 O "-- -O
U-y- O C Z Z000 0 1- 0 0

) a) U o 0 4- L 0
-- 4- 4 .> a) 4 C
LLJ0 r- r- *- *- C *r- +
F1 -- ao 4- > 4. 4 *(D U *r-
C .-- O r a+4> o I a e )
S< 4- | t) 43 *r- *r- 0ca 4- 4-t C 0
OL '0 U) r







-4'



< F-r
F- 0 -- nU -.
c.- i O O 3 U O) CO Cn 4-3 4-
S:. C -c 2- < -< Z C/) 3 r 0



C *-4 -





.C co,
X -Ij C












0- E <4-
C
C 3: *< >






0 4-
C: L O-
4- C 0.1u 0





:) c (a 4- 1-J 0
O)LJ 0 0 C .u
o .> 0 ( 0











w Lo r- a) 0 S-
< 0 Z L S.-

04- 4-). E C)
1-fX C C
CD~ O


co *r- a >: *i
-- Io > 0
C-- 4- 4 4 I -0















>a-.0 4
0 O0 > (0 U 4. -
o o t -~- "o o an i.-

S- *- C C- a 01 ) 0
Z "3 +> f0 a *r- I O U
oL ( 3 -?o -o- -M C > -I & a















(a 0 4-) W a )U ea S)E3
+C J X (00 O <0 C S- 0 )
0 L S.- -- 0 o 0 -
.< *- -0 .0 > C > 0
WC 4- Cr M e- 4.-0 -) < 0
2 >C E 0 ) 0- -+ 4r- I- E -
0 r-- QS J> t I L- *. m C- CO



X+: < (D F- I- 0 C u -- m .0 u "a
O- 01 0 CC *) 0 0
>> UJO *i-- S <- l. CN 4- > <) OQ G

L. L '0 r-> +3 +- l < 0 +


S0 CL 15t *n CO -- X U c. 4 .
0 0.> < CL C. *l l- 0 +









TABLE XI

51Cr RELEASE BY L-929 CELLS COCULTURED WITH MOUSE PERITONEAL
EXUDATE CELLS


Percent 51Cr Releaseda,b
Actinomycin D No Actinomycin D
Time (hours) Pretreatment Pretreatment


3 1.3 0 0

6 38.3 + 1.8 1.9 0

12c 46.8 + 1.3 9.4 0

percent specific 51Cr released into supernatants = E S
T -S (100)
where E = experimental release, S = spontaneous release and T = total
releasable counts.

S = amount of 51Cr released by target cells cultured alone.

T = total releasable counts by lysis of 3 x 104 51Cr labelled
L-929 cells with 1% dodecyl sodium sulfate.
bSamples run in triplicate for determination of experimental and
spontaneous counts released. Samples run in quadruplicate for total
releasable counts.

cMaximum spontaneous counts released = 22% of total releasable counts
by 12 hours.










































t
C


0







C
*-

C,







=
0

4e




0C


U-
cu
0r-.
OC




I











S-
30
oC
E -'
S)


-r
r-

I








L.


L. 0
Q -
C,D0
+


U-1


S. -- -
= -U -0
03 4- *
c- 0 r- X
U) EU
C -. 1-.0


+- 0 C 0,
4- 3 ,0 C0
CJ C C -I 0
L QO C
*r- +-






a m 0

c C *0- o
r- Q00*
C .. C -0


rr-







4- ) ( U)
, S- g) 3 -
0c 0
C 3
*r- I 0 C o


u re- O J
S.e r- *






4Jr- 3
3( EC(
3 -CflUNr'-








On (n .- 4-4 C






a) 0 1-S.-
CO OS











r0 CU).4-)
r L 4-)M C
0 E (0a





E 0) X S,- -
S3 U) *U-













) C4- 0 (D
) I 4- 4- 3 I'
5 3 0C -






S+J = -0
SU *r- O

( c: o c 0

.C O0
N *r- .C r-

e 01 *r- )
1o *r- QL-UJ (-








- ~ S r


D N /,



0 H
" 0:: .'
* : *^ ---;r.- < -f~ ..^ v. u g





eeN

,'0,10


ii-I H
o


0r
Ct~ c43 ~ oz


i ,





ii [

[i I,






L't









targets. Figure 14 demonstrates this point. Approximately 175 rabbit

PLC and 300 mouse PEC could kill 50% of the actinomycin D pretreated

monolayers. If we take into account the fact that 95% of the rabbit PLC

and 70% of the mouse (C3H/HeN) PEC were macrophages, we can then

determine the average number of targets killed by a single macrophage.

Knowing that approximately 50,000 L-929 cells composed a monolayer, it is

determined that a single rabbit macrophage killed an average of 25,000r

(175x0.95) or 150 targets and a single mouse macrophage killed an average

of 25,000+(300x0.70) or 119 targets. These are only average determin-

ations, since, as will be shown, not every macrophage necessarily killed

targets, nor were the number of targets killed by a single macrophage the

same.

Controls for these experiments previously determined that the

manual photocell employed for these studies could at no time detect the

small number of effector cells plated. In addition, removal of adherent

cell populations from mouse PEC virtually eliminated their ability to

generate killing. We therefore concluded that the macrophage was the

effector cell in the PEC population.

Demonstration of Single Effector Cell
Cytotoxicity: Plaque Assay

Repeating the same experiment on a larger monolayer surface (24

well plates) resulted in a greater dispersal of the effector cells.

Here, killing was detectable by the formation of plaques (see Figures 15

and 16). If the number of effector cells plated was compared with the

number of plaques that resulted, a linear relationship between the two

was obtained (Figure 17). This dictates that a single effector cell was

capable of forming a single plaque, though not every effector cell






























Figure 14. Effector cell cytotoxicity for actinamycin D
pretreated L-929 targets: Determination of S50 endpoint.

L-929 monolayers were pretreated with actinomycin D
(2pg/ml) for 2 to 3 hours. After washing monolayers, effector
cells (a, mouse peritoneal exudate cells; b, rabbit pulmonary
lavage cells) were added in 2-fold increments. Control target
monolayers received no effector cells. LPS was added at 10
pg/ml. Target monolayers were stained after 18 hours.














0
O.
7e


7,
S
7-.


*/


10 20 40 80 160 320 640 1280 2560
NO. MOUSE PERITONEAL EXUDATE CELLS PLATED


LJ
o 100
1-

ag 80-



60-



40



20


10 20 40


80 160 320 640 1280 2580


NO. RABBIT PULMONARY LAVAGE CELLS PLATED








































u
>1

e:
*r-










0






x

4-'




_4-
4-)
0
3 >














U-)
*A 4-






U r-
4->









o- 4

CL
U) )
p0
O y












4-'
SCi










(U -

0 o
a




L.
n.


0
c-
Q-
01 <


I t)
....C L



CU)

0 r- .
U) -aU

0 4-


tj C





ra 0-
I- -



S- 4-U) CL
(0
0 U0

(ALr-C


cu2 W





=- 4-0 >
O *. I
-Q 0-


n U




Cn r- c


L0 0 U-














4-> S 0
S(U




DEL-







0 0C Q.



CD.Q*r
i- ^ S-
O"3 0
L1> Q





CN
CN

04



CC





0
00












CC
ro w



H
0E






















cI
U







0

Z
0
CO





u- <







































a,









c-,



C
4-3
I0






o
S-




ra










,--
r>


C
E r-
+-
0




















c







CL
0













I










*r- 1

LL.)


CLJ
JO
1..
>L


















(U



















x N
.. .


-;1
H





ofo
ncn



I ,I, ,.= N .
-..m-.






i ...






IF
~0


---~- --






























Figure 17. Linearity between number of plaques counted
vs. number of mouse peritoneal exudate cells plated.

Each point represents the mean of 4 replicates 1
standard deviation. Panel a, represents peritoneal exudate
cells from C57B1/6 mouse; panel b, represents peritoneal exudate
cells from C3H/HeN mouse. Botn lines represent the best fit
by linear regression and nave correlation coefficients of
>0.990.
















100



80



60



40




O
0, 20


Q-
u.
0
r,
w 100


2z
80



60



40



20


20 40 60 80 100 120 140

NUMBER OF PERITONEAL EXUDATE CELLS PLATED









necessarily formed a plaque. Therefore, one plaque represented the

killing events associated with a single macrophage.

For convenience, this assay was developed in mice and then utilized

in the rabbit system. One of the most noticeable observations about the

rabbit system concerns its variability (sensitivity of PLC to LPS when

comparing different rabbits; see Figures 3 and 6). This plaque assay was

performed using PLC from several rabbits. In only about 20% of the

rabbits used did the number of plaques generated conform closely with the

number of PLC plated. This probably demonstrated variation among rabbits

concerning the number of macrophages in the overall PLC population that

could be sufficiently activated to make a plaque. In addition, as in the

mouse system, plaques varied in size, demonstrating variation in the

ability of each macrophage to kill targets.

Other observations made in the rabbit plaque system included the

inability of 2 x 10-2M arginine in the medium to inhibit PLC plaque

formation and the requirement for the addition of LPS to get plaques.

L-929 cells for these latter studies were maintained in a culture system

which tested negative (<0.025 ng/ml) for LPS (see Table III). Finally,

if actinomycin D was added at the same time as PLC and LPS to actinomycin

D pretreated and washed L-929 monolayers, no plaques were produced.

However, if actinomycin D was added to the system 2 hours or more after

the PLC and LPS were added to actinomycin D pretreated and washed

monolayers, the same number of plaques were formed when compared to

controls to which no actinomycin D was added either with or following PLC

(data not shown).

Time Lapse Cinematography of Small Numbers of
Rabbit PLC on Actinomycin D Pretreated L-929 Cells

Using the conditions described in Materials and Methods for

cocultivation of effector cells on actinomycin D pretreated targets, time