r. 242:: '; _"“-

A ,Ih— g
. ‘ -42.; <‘

~ -I..::‘

_ 1.;~:~. urn»;
73E“ 33:.5‘3522’r-‘i' ~

:3:
32%;.

To.

;. ‘e
21:397-

1&5

m

. L __‘.- .0“.

l
V II
‘33“ "MR3, \ ..

{wk}: HI

“W M_ - 3

, It? :1‘ ‘1“:33‘3‘0

' q€I3I‘;I ‘1 ”1333" 33-1334}?

'43 ‘ ‘( 3431335!) '1 ILLL'" ""3; 1351 g
. ‘1'?“ 33313339 L»

{$33 ‘GIII’I 11-31533?"
.- ['lw'i'hxd iqin my 'Iwfl'fik
3‘3“ 123% I33“ ‘1'} 26‘3‘31
uklili'dfl‘w .- III? 3'"
3'“ 133“" It?
A l I 0137 1. '
IHH3J1H3‘3‘ [H l’l 3’31.-
WWWHI
133‘ f133‘3311ig333 "3 '3 3‘31:

“1:.-. >. h:

....
3.1.1:" .

9/59“}.

an?

‘;.i.

4.2:
'5“.

3n

4‘1“ 1: 333333; I

3133:“ 3f '
33144 '25! I

‘72::

9-1:-

“3;;
‘l

#2::

aka

7%.—

.n~

”211‘ ‘ .-.
‘gga' a » 3 — W

. ft?“ . :4}:
2?

PM

A?
”M...—

“Eta
‘A .m- .—
5: I W

-—::""‘;-: z

w: 45g:_
40-.“

3&3;

L - '
.-.. . p
__— 3-13" . 4%»;
.9... I", ‘L ' . . .
r' ‘ v ' K
mm‘fi‘ A -‘
. “-3.:- < 'qgf‘shd w .
43E? .

- E. “2:25:
”rm-E... , :
é§%§§§

“m 3 4 ' H
.39"; 1; ,,:

,4;
a
“It...
“" _ L

‘1- A».

W

mw' '. n
—:v-. . a! A

r:

=53.

~‘~‘_.sri:s..

«fab—“35.6
""""‘.T“&#f:‘¢
#21:.

v“ _,
“i. ’34 n.
A." *—
rm:

w 'v.._‘

' :x-v: i=7“ -
WWW
-; Lie.

w-..

“w

.._-_,‘... -
..._L
":2

m

,M"

—‘ 55': les;
”L31”;
”1: 7% w 1%- 1- 3

Fr
Mew -
2g
4;.

.7

, ~£“
L'

..r.;_~:¢.. A.’ “-

5-..“, . .
. :4“: ’.« L
....jn‘:‘

1‘2:
:..- ;~"‘:. - . ‘

- r, r:

H"...—

42:73"
1 .zyf-W‘L‘M ‘ u‘,

.14.:

\_ _x*,3:r "

--~: ‘2‘ -

M .. - -
«L4,:-
arr $3341:
3,9...
.14

1 .JJ.
‘ if?!
- a.“

H331
9‘333‘ 3"‘3‘:

'5”?!
g
3;";

112a
Ifi‘

I‘M”

2'2"? ’ - ...5.:,—:
_~:;?f='~ffi’§é.:§
“=2 . .x‘
9

.~.;

in

II‘
"3333", 313333
I HI.
{3313“
”1"12'"):

......, .:_|,: _' x

:3 ' - . >4 . _‘ ...
‘ 55 .

1?:— “9:22;; $14.5?“

.. “mfi1””*f‘:w-

': :-::£;E:f;-g._ “9’2." :1": '1. ‘ r'
* “m 2‘ 377%: ’5 I}: {
-—- 1‘“ .r. “.7 ‘F' ~ ~—
,3: M _.(I“" 1.: 4: A » «5":1 ‘- .err
434...?” Ci; L .1, ,r‘fa 4:»

A?“ 4:} ‘fi::§§g§~ a;
.: M A

3333‘ 133% 33%“,‘3'. 3.33; «,3
‘_1‘i‘ fig "1‘ ,
3% .: fi'fifi'i'aj; 3?. 0
1 '1.
1.

“.32.

A’i-

‘I

vs:

.4:

. 3- «'i A : I 4 :. 4.“; :N'Jégf“
~34» ' "It *3..." =33"
‘9’ ',‘3E'§f;‘:::,:

, .,..~.;.. :

,-

II‘

(‘fltml
3:335 3'33“ .3“;
‘ HQ"

‘ ,I‘l ‘3'
'1'
IN
W 31:31,. '
1‘31],
u§ ‘

II {31"1' ' ‘. “if;

3333‘.

‘ia'firé-aw
W. .

A.» ’3’}
“1&-

 

‘M‘z

a!
:55.

I“:

1'

521:“

' -r’ 1'
i?“

JAA‘NI

E3313 3'1 "3;
WI

_,...‘,‘

,zza.__,. ,r: 1-}: 3 1'. ' _;', “if:

,
aw,
=9"

P?

"-9.731“
x 4..

‘3‘”

f5“

7’

's'wxz; ' - is???
Eu ~L

‘ .95:
N‘s-J

3::-

(2-57“,

~ " A ’ I» ‘85?"
. . . ' ‘2’ i- v":"‘ 'e‘ Jw . __ ' . ‘ - #35; f;-
3 A :.V -. I. - _ ‘7» .. .ai- -,_: - . gm: Kg

3.

.I a f a
. -44.... mun-‘5'.“ A
.. 4-2233“ .416

THESIS

 

Michigan Stave
A University

 

This is to certify that the

thesis entitled

SELECTED CELLULAR AND SUBCELLULAR INTERACTIONS OF TOXIC
AND DETOXIFIED ENDOTOXIN WITH EUKARYOTIC CELLS

presented by
HAS SAN TAVAKOL I

has been accepted towards fulfillment
of the requirements for

(PIM‘I) degreein WI ? I\I

 

Date H/“/@
I 7

0-7 639

 

‘f- . u
\ ‘5‘;
\IC‘ 3r-..II “.23
!\"I.~ ;::.“\“",/”~ 1:97;; I VJ ,4

 

OVERDUE FINES:

25¢ per day per item

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

 

 

SELECTED CELLULAR AND SUBCELLULAR INTERACTIONS OF TOXIC

AND DETOXIFIED ENDOTOXIN WITH EUCARYOTIC CELLS

BY

Hassan Tavakoli

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Microbiology and Public Health

1980

Haas?“

\-

ABSTRACT

SELECTED CELLULAR AND SUBCELLULAR INTERACTIONS OF TOXIC AND
EXPERIMENTALLY MODIFIED ENDOTOXIN WITH EUCARYOTIC CELLS

BY

Hassan Tavakoli

Organ distribution and cell association of toxic and
experimentally modified endotoxin in_zivg and in_yi3£g was
compared. More than 50% of intravenously (i.v.) injected
toxic endotoxin or poly L-a-ornithine mixed endotoxin (each
200 ug/mouse) became associated with the liver and spleen.
The pattern of distribution did not change between 1 and
5 h. Alkaline detoxification of endotoxin, which changed
the physical and chemical pr0perties of endotoxin, also
changed the organ distribution in that significantly less
detoxified endotoxin (180 ug/mouse) was recovered from liver
and spleen. Once fixed, the cellular and subcellular
distribution of detoxified endotoxin did not significantly
change between 1 and 5 h.

Association of endotoxin with hepatoma tissue culture
cells (HTC) cells differed in several ways from in ziyg
organ distribution. For example, HTC-cell association and
nuclear transfer of toxic and poly L-d-ornithine mixed
endotoxin was a gradual and time dependent process. Poly

L—a—ornithine increased in vitro cell association of

 

Hassan'Tavakoli
endotoxin by almost 10-fold. HTC-cell association of
alkaline-treated detoxified endotoxin was 3- to 8-fold
higher than toxic endotoxin and increased with time.
Cumulatively, these observations indicate that while tissue
culture cells could conceivably provide a more controllable
experimental system in which to study the fate and patho—
genic mechanisms of endotoxin at the cellular and
subcellular level, HTC-cells, under the conditions employed
herein did not yield binding data which compared favorably
to in yiyg results.

The present study also examined the in vitro inter-
action of toxic and experimentally detoxified endotoxin with
chromatin, DNA and steroid hormone—receptor complexes.

Toxic endotoxin was not able to bind to DNA alone, but
relatively higher amounts could slichtly inhibit the binding
of glucocorticoid hormone—receptor complexes binding to DNA
(14.6% inhibition/300 ug endotoxin). Toxic endotoxin was
capable of interacting with HTC-cell chromatin (1 ug
endotoxin/20 ug DNA) almost four times more efficiently than
alkaline—treated detoxified endotoxin (0.24 ug endotoxin/

20 ug DNA). Interaction of glucocorticoid-receptor steroid
complexes with chromatin was inhibited in a dose dependent
manner by both toxic and detoxified endotoxin, but again
relatively high concentrations of endotoxin were needed.
Under certain conditions endotoxin can block hormone-receptor

complex binding to the DNA and chromatin. The biological

 

Hassan'Tavakoli
significance of this observation is complicated by the need
of high concentrations of toxic endotoxin and by the fact
that biologically detoxified endotoxin works even more

efficiently than toxic endotoxin.

DEDICATION

To my wife Zohreh Tavakoli and my parents,
Mr. and Mrs. Tavakoli, without whose love, encouragement
and support, my education and this thesis would not have

been possible.

ii

ACKNOWLEDGMENTS

I wish to express my sincere appreciation to my major
advisor, Dr. R. J. Moon, for his unfailing inspiration,
intuitive guidance and kind patience.

I would also like to thank other members of my Committee,
Dr. N. B. McCullough, Dr. P. T. Magee and Dr. A. J. Morris.

I also wish to thank Dr. Betty Werner, Dr. Richard Sawyer,
Dr. Rick Friedman, Bub Leunk, Ellen Keitelman, Ruth Vrable,
Bryan Petschow, Lee Schwocho and Dace Valduss for their
friendship, discussion and encouragement throughout my
graduate study.

Finally, I would like to thank my sister, Shekoh Tavakoli,
for her enthusiasm and emotional support throughout the past

several years.

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . . . . . . . . . .
LIST OF FIGURES . . . . . . . . . . . . . . . . .

INTRODUCTION . . . . . . . . . . . . . . . . . . .

LITERATURE REVIEW . . . . . . . . . . . . . . . .
Chemical Structure of Endotoxin . . . . .
Modification of Biophysical and Endotoxic

Properties of Endotoxin . . . . . . . . .
Radiolabeling of Endotoxin' . . . .

Effect of Polycationic Polymers on the Entry

of Macromolecules Into Mammalian Cells . .
Interaction of Endotoxin With Cells . . . . .
Fate of Endotoxin In vivo and In vitro . . .

 

Pathogenesis of Endotoxemia . . . . . . . .
Protective Role of Adrenocortical Hormones
Against Endotoxin Poisoning . . . . . .
The Mediation of Endotoxemic Effects . . . .
Mechanism of Action of Steroid Hormones . . .
Rationale for the Present Study . . . . . . .
LITERATURE CITED . . . . . . . . . . . . . . . . .
ARTICLE I . . . . . . . . . . . . . . . . . . . .
APPENDIX I . . . . . . . . . . . . . . . . . . . .
ARTICLE II . . . . . . . . . . . . . . . . . . . .
APPENDIX II . . . . . . . . . . . . . . . . . . .

iv

80

108

LIST OF TABLES

Table Page
Article I
1 Comparison of selected biological and physical
properties of toxic and modified endotoxin . . . . 50
2 Organ and tissue distribution of toxic 51Cr-
labeled endotoxin (200 ug) in mice . . . . . . . . 52

3 Organ and tissue distribution of 51Cr— labeled
toxic endotoxin (200 ug) mixed with poly-L— a-
ornithine (48 ug) in mice . . . . . . . . . . 53

4 Organ and tissue distribution of 5ICr—labeled
0.25 N NaOH.treated endotoxin (180 ug) in mice . . 54

5 HTC—cell association and cell distribution of
toxic endotoxin (92 ug/5 ml cells), in Vivo
mouse plasma-treated (28.5 pg LPS/5 HI celI)
and in vivo mouse plasma- -treated centrifuged
(14 ug LPS75 ml cell) endotoxin after 1,3 and
5 h incubation at 37° C . . . . . . . . . . . . . 56

6 HTC—cell association and cell distribution
of endotoxin mixed with poly—L—d-ornithine
(92 ug/S ml cell) after 1, 3 and 5 h incubation
at 37°C . . . . . . . . . . . . . . . . . . . . . 58

7 HTC-cell association and cell distribution of
0.25 N NaOH treated-endotoxin (63 ug/S ml cell)
after 1, 3 and 5 h incubation at 37°C . . . . . . 60
Appendix I

1 HTC- -Eell assocgition and cell distribution of
Cro Na Cro mixed with endotoxin and

NaiSlCroj—labezaled efidotoxin at 37°C . . . . . . . 73
2 Limulus lysate test on HTC whole cell and nuclear
lysate incubated with endotoxin for 3 h . . . . . 75

List of Tables (Continued)

Table

Appendix If(continued)

3 TC-cell association and cell distribution of
Cr—labeled toxic endotoxin after 1, 3, 5 and
24 h incubation at 4°C . . . . . . . . . . . . .
4 Organ and tissue distribution of (180 ug)
Cr-labeled 0.25 N NaOH treated endotoxin
in mice . . . . . . . . . . . . . . . . . . . .
Article II
1 Hydrocortisone induction of PEPCK in the presence
and absence of toxic and detoxified endotoxin .
2 Hepatic association and subcellular distribution
of toxic and detoxified endotoxin 5 h after iv
injection into mice . . . . . . . . . . . . . .
3 Binding of 51Cr-labeled toxic and detoxified
endotoxin to chromatin . . . . . . . . . . . . .
4 Inhibition of AHRC binding to chromatin in the
presence of toxic and detoxified endotoxin . . .
5 Binding of 51Cr-labeled endotoxin to cellulose
or DNA-cellulose . . . . . . . . . . . . . . . .
6 Binding of AHRC to DNA—cellulose in the presence

and absence of endotoxin . . . . . . . . . . . .

Appendix II

1

Binding of activated hormone—receptor complex to
sonication-prepared and nuclease—extracted
chromatin . . . . . . . . . . . . . . . . . . .

vi

Page

77

79

90

91

93

94

96

97

115

LIST OF FIGURES

Figure ' Page
Appendix I

l CsCl ultracentrifugation of untreated
endotoxin . . . . . . . . . . . . . . . . . . . . 69

2 CsCl ultracentrifugation of normal mouse
plasma treated endotoxin . . . . . . . . . . . . 7o

3 CsCl ultracentrifugation of 0.25 N NaOH
treated-endotoxin . . . . . . . . . . . . . . . . 71

Appendix II

1 Sephadex G—25 chromatography of activated
hormone- -receptor complex . . . . . . . . . . . 110

2 Binding of increasing amounts of 3H—dexamethasone
activated-receptor to constant amount of
chromatin . . . . . . . . . . . . . . . . . . . . 112

3 Binding of constant amount of 3H-dexamethasone
activated-receptor to increasing amounts of
chromatin . . . . . . . . . . . . . . . . . . . . 113

INTRODUCTION

The chemical and biological properties of bacterial
endotoxin have been under investigation for more than a
century. Thousands of papers dealing with chemical struc-
ture, purification and biological effects of endotoxin have
been published during the past two decades. Unfortunately,
the time and effort devoted to endotoxin has not produced
the same level of understanding of its biological action as
exists for many protein toxins. Endotoxin manifests a
surprising multiplicity of biological effects in animals,
but it is still not clear which of these are important in
lethality.

Endotoxin has a wide range of biological activities
that affects all systems of the host. Some of the major
effects of endotoxin are shock, lethality, cardiovascular
changes, pyrogenicity, immunogenicity, antitumor activity,
Shwartzman reaction, mobilization of interferon, metabolic
changes, cytotoxicity, interaction with complement, release
and sensitization to histamine, abortion, changes in blood
clotting, development of tolerance, effect on the reticulo-
endothelial system (RES), protection against irradiation,
leukopenia and leukocytosis, effects on properdin or natural

antibody levels and adjuvant effects (7,61).

Despite the wide range of the biological changes which
have been associated with bacterial endotoxin it elicits few
cytological or morphological changes. The varied biological
effects of endotoxin make it clear why so much attention,
time and effort has been devoted to study of this bacterial
poison.

One major school of thought on the significant factors
in the pathophysiology of endotoxemia lies in its ability to
induce host metabolic alterations, particularly in liver
parenchymal cells (12-l4,l6,52-55). These changes range
from depletion of carbohydrate reserves to the inhibition of
gluconeogenesis and hormonal induction of certain liver
enzymes (43,72,73). Whether endotoxin exerts its effects on
parenchymal cells directly (90) or through soluble mediators
(57) has not been definitively established. A previous
report from this laboratory (90) indicates that endotoxin
becomes associated with liver parenchymal cells in 3129 and
tissue culture cells in 21332. If the interaction of
endotoxin with tissue culture cells mimics its interaction
in ziyg, then tissue culture cells would provide a more
controlled environment for studying endotoxin-cell interactions.
In the present study, cell and tissue association and distri—
‘bution of toxic and modified endotoxin in whole animals and
in hepatoma tissue culture cells is compared. The second
goal of this study is to gain insight into the nature of
the inhibition of hepatic enzyme induction and the mechanism
of the protective activity of cortisone against endotoxin

lethality.

The results of these experiments show that caution
should be exercised in using tissue culture cells as a model
to study endotoxin cell interaction and the fate of endotoxin
in the cell. This study also indicates that 51Cr-labeled
endotoxin can interact with chromatin but not with DNA in
the form of DNA-cellulose. Endotoxin in able to inhibit the
binding of activated-hormone—receptor-complex (AHRC) to
chromatin and DNA in a dose dependent manner, but the amount
of endotoxin which is needed for significant inhibition of

binding of AHRC to DNA and chromatin is far higher than is

required to elicit the biological consequences of endotoxin.

LITERATURE REVIEW

Chemical Structure of Endotoxin

 

It is well established that endotoxin is an integral
part of the outer membrane of gram negative bacteria (29).
Endotoxin is a lipopolysaccharide (LPS) having a molecular
weight ranging from 1 to 20 million daltons depending upon
the method of purification. It forms aggregates easily
which explains its very high and variable molecular weight
(61). Chemically endotoxin consists of three parts: the
lipid moiety, the R—core and the O—polysaccharide. The
basic structure of the lipid moiety, known as lipid A, is a
diglucoseamine unit that contains fatty acids attached to
amino and hydroxyl groups (19). A fatty acid which seems to
be unique to lipopolysaccharide (LPS) and which is usually
found in the highest concentration is B-hydroxymyristic acid
(46). The glucose-amine units are bound together by 1—6
glycosidic linkages and 1,4—phosphodiester bridges (1,32,64).
The R-core region links lipid A to the O-polysaccharide
moiety. VThe existence of a group of mutant bacteria termed
R or rough forms which lack the O—polysaccharide has
facilitated the investigation of the R—core structure.
Chemical analysis of the R—core has shown five sugars,
phosphate, and O-phosphorylethanolamine (37). The O—poly-
saccharide structure of bacterial endotoxin determines the
O-antigenic specificity of the bacterium. Biochemical
lanalyses of this portion of LPS molecule have demonstrated a

repeating sequence of sugars (38).

Modification of Biophysical and Endotoxic Properties of

 

Endotoxin

Chemical and biological modifications of endotoxin have
been used in the search for an explanation of endotoxin
toxicity. Goodman and Sultzer (33) studied the effects of
mild alkaline hydrolysis on bacterial endotoxin. They
obtained partial alkaline hydrolysis of LPS by dissolving
50 mg of LPS in 3 ml of 0.25 N NaOH and heating the solution
at 56°C for 60 min as originally described by Neter et al.
(60). Their results indicated that while the treatment
reduced the lethality of endotoxin for mice by about 100 fold,
there were no significant changes in the fatty acids of lipid A.
Particle size was significantly reduced, and the material was
more homogeneous and soluble than untreated LPS. Mild
alkaline hydrolysis was found to enhance lps mitogenicity for
murine B lymphocytes.

Skidmore et a1. (78) prepared detoxified endotoxin by
incubating LPS at a concentration of 1 mg per ml in 1.0 N
NaOH at 56°C for 60 min. The solution was then neutralized
with HCl and dialyzed against phosphate buffered saline.

This type of treatment greatly reduced in vitro mitogenic
activity and toxicity of LPS without affecting the chemical
structure of the antigenic polysaccharide (65). Skidmore et
al. (78) established that l N NaOH treated LPS was non-
mitogenic for spleen cells and also showed a positive
correlation between the in vitro activity of LPS as a

lymphocyte mitogen and its in vivo activity as adjuvant and

as an immunogen. Tauber et a1. (84) reported that hydroxyl—
aminolysis is an effective method for detoxifying endotoxin
and for removing almost all of its fatty acid ester groups.

While endotoxin-blood interactions have been intensively
studied, the mechanism of LPS detoxification in the vascular
compartment remains a subject of controversy. The mechanisms
proposed to account for LPS detoxification involve either the
cells of the reticuloendothelial system or humoral factors.
Skarnes (77) believes that circulating plasma represents a
principal site of detoxification and that plasma esterases of
the nonspecific, carboxylic types are of major concern in
defense against circulating endotoxins. His results indicate
that within the first few hours following intravenous injection
of endotoxin a decrease in ionized calcium, a three—fold
increase in heat-stable esterase levels and a striking increase
in the endotoxin detoxifying capacity of plasma all occur in
the circulating plasma of the recipient.

Ulevitch and Johnston have recently reported (85) that
normal rabbit, human, or mouse serum (or plasma) reduces the
buoyant density of endotoxin from Escherichia 391: Olll:B4

(d = 1.44 g/cm3) and Salmonella minnesota R595 (d = 1.38 g/cm3)

 

to a value <l.2 g/cm3. This density shift was associated
with the inhibition of pyrogenic activity, the ability to
produce an immediate neutropenia, and the anticomplement
activity of endotoxin. Preliminary results from their
laboratory indicated that the density shift did not occur as

a result of the degradation of the glycolipid backbone of the

LPS. These data suggest that the interactions of LPS with
plasma (or serum) components leading to reduction in buoyant
density without degradation of LPS may account for a major
pathway of LPS detoxification. In another report, Ulevich et
a1. (86) indicate that this reduction in buoyant density of
endotoxin requires a plasma (or serum) lipid. Delipidation
of plasma (or serum) prevented the conversion of the parent
entotoxin to a low density form. Addition of purified high
density lipoproteins (HDL) to delipidated plasma (or serum)
restored both the ability to reduce the buoyant density of
endotoxin and the capacity to abrogate the activities of
endotoxin.

Freudenberg et a1. (30) used the method of crossed
immunoelectrophoresis to investigate early changes in plasma
proteins of rats treated with LPS. Their results indicated
that intravenous injection of either a smooth (S) or rough
(R)-form of endotoxin leads to alterations in the HDL of
plasma. The changes were dose dependent and identified as
being due to the formation of a complex of LPS with HDL.
They followed the fate of the complex in the plasma of
injected rats, and showed that the R-form LPS complex
disappeared after several hours, whereas the S-form LPS
complex was still partly present after 2 days. Their results
indicate that the time endotoxin circulates in the blood may
depend on several factors, one of which may concern the type
(S or R) endotoxin. Freudenberg et al. suggested that HDL
may represent a transport protein for LPS in plasma to

organs of clearance or to other cellular targets.

 

Radiolabeling of Endotoxin
To study the mode of endotoxin action, it is necessary
to study the fate of endotoxin in the hoSt and in the cell.

The ability to radiolabel endotoxin has facilitated this type

of study in recent years. Radioactive iodine (71), 32P

14

(40) and C (48) have all been used to label endotoxin.

Braude et al. (17) devised a method of firmly attaching

51chromium in the form of sodium chromate or chromic chloride

to the endotoxin of Escherichia coli. The toxicity of 51Cr—

labeled endotoxin was not altered as a result of the labeling
procedure. Both the sodium chromate and the chromium chloride
were able to bind to endotoxin firmly, however the acidic
condition in the case of chromium chloride resulted in the

loss of toxicity of endotoxin.' His observation also indicated
that serial dilutions of labeled endotoxin resulted in a pro-
portional decrease in both radioactivity and toxicity.

Braude et a1. concluded that the close quantitative correlation

between toxicity and radioactivity justifies the use of

51Cr—labeled endotoxin. In a subsequent study Braude et al.

(18) compared the organ distribution of the free 51Cr to the

51Cr bound endotoxin. Their results indicated that there is a

significant difference in the distribution of free 51Cr

51

and Cr-labeled endotoxin. These data support the stable

51

nature of the Cr-labeled endotoxin and reveal that the

51Cr does not break away from endotoxin upon injection;

Chedid et a1. (21) undertook a series of experiments to

51

further clarify the character of Cr-labeled endotoxin.

Their results indicated that centrifugation of a freshly

prepared 51

Cr-labeled endotoxin mixture yielded a heavy,
highly labeled pellet containing the toxic, labeled moiety.
The supernatant contained a weakly labeled product which was
nontoxic. Further studies showed that 9% of an injected dose
(50 micrograms) of 51Cr-labeled endotoxin was still recover—
able from the plasma 6 hours after injection. Recovered
endotoxin proved to be lethal for adrenalectomized mice,
which had been previously shown to be susceptible to as
little as 0.05 micrograms of endotoxin. Furthermore,
autoradiographic studies demonstrated that the recoverable,
toxic material was still labeled. Chedid‘s et a1. work

51 51

clearly demonstrate that Cr does not dissociate from Cr-

labeled endotoxin after injection of toxic material.

Effect of Polycationic Polymers on the Entry of Macromolecules

Into.Mammalian Cells

 

~

Uptake of a macromolecule is the first necessary step
toward a biological effect in a host cell. Ryser et al.

(69,70) used albumin—Il3l

and monolayers of sarcoma S—l80II
cells as a model to study penetration of foreign proteins
into mammalian cells. Their results indicated that transport
proceeds at very low rates, requires little energy, and is
markedly enhanced by polycationic polymers and certain
polynucleotide aggregates. This enhancement phenomenon
increases with the molecular size of the polymers. The
stimulation of albumin uptake by 3 ug of poly—D—lysine or
poly—L-ornithine per milliliter is approximately ten fold.
Ryser et a1. (70) suggested that polycationic polymers

initiate the entry of macromolecules into the cells by

9

making multiple attachments with the negatively charged cell
surface. While the sites of cell surface interactions of
polynucleotide aggregates might not be the same as for
polycationic polymers, it is likely that multiple contacts
occur between polynucleotides and the cell membrane to
trigger the entry of macromolecules. This enhancement
phenomenon should be useful to investigators concerned with

the intracellular functions of foreign macromolecules.

Interaction of Endotoxin with Cells

 

The problem of how such a large and complex molecule
interacts with cells or membranes has not been resolved.
Anderson et al. (3) and Dumont (26) have suggested that
lipid A inserts into the lipid bilayer of cell membranes.
Endotoxin has been shown to rapidly adhere to the surface of
red blood cells (RBC). Coated RBC can be used to detect
antibodies against endotoxin by indirect hemagglutination.

In 1970, Springer et al. (83) described an inhibitory substance
from the ghosts of human RBC which prevented attachment of

LPS to RBC. The inhibitory material had no apparent effect

on the binding of Salmonella Vi (virulence) antigens or

binding of group and common antigens from the gram positive
bacteria investigated. Springer et al. demonstrated the
interaction between the chemical groups of LPS that attach

to RBC and the inhibitory material. This inhibitory fraction
was named LPS receptor. The receptor can displace the
endotoxin after it has attached to RBC leading the authors

to suggest that its blocking is physical and not enzymatic.

10

The receptor was shown to be sensitive to pH changes on

either side of pH 7 and to be heat labile.‘ In later publica—
tions (81,82) Springer et al. isolated a lipoglycoprotein
(receptor) from smooth and rough strains of all gram negative
bacteria tested. This "receptor" was able to block endotOxin's
attachment to red cells. Receptor activity was destroyed by
proteases and was found to be rich in N-acetylneuraminic

acid, galactose, and hexosamines. They estimated that the

LPS receptor molecule weighed about 256,000 daltons. Another
report from Springer and Adye (80) indicated that they had
been able to extract the LPS receptor with n-butanol water
from isolated platelets, granulocytes, and mononuclear
leukocytes of humans. The extracts contained 40 to 52%
glycerophosphatides, 15 to 22% sphingomyelin, 17% cholesterol,
lessthan 2 to 5% triglycerides and 7 to 13% inactive peptides.'
According to Springer and Adye (80) lipid A inhibited the.
binding of LPS to RBC. Studies of this type may elucidate
how endotoxin gets into cells;

Alkaline-treated LPS had a greater affinity for RBC-than
toxic endotoxin (22). Hemolysis results when excessive
quantities of alkaline-treated LPS are added to RBC.
Benedetto et al. (6) used artificially produced phospholipid
bilayer and monolayer membrane models. Their results
indicated that alkaline treated LPS was at least ten times
more efficient than the native product in penetration.
Benedetto et al. concluded that the interaction of LPS with
phospholipid monolayers was by a combination of penetration

and adsorption to the undersurface.

11

Fate of Endotoxin In vivo and In vitro

 

The results of numerous investigations on the fate of
endotoxin in the host indicates that endotoxin rapidly
accumulates in the reticuloendothelial—cell-rich organs after
iv injection. The spleen and especially the liver appear to
be primarily involved (44). Other sites of considerable endo—
toxin accumulation are the endothelium of blood vessels and
the lung alveoli. The reaction of endotoxin with white blood
cells is very rapid. The results of Rubenstein et a1. (68)
indicate that within 10 min after injection of radiolabeled
endotoxin, polymorphonuclear leukocytes are labeled with
endotoxin. Herring et al. (39) found that platelets absorb
endotoxin. Most endotoxin became fixed to tissues or organs
by 30-60 min and little change was seen in this distribution
even after several days (41,90). However, the Musson et al.
report (59) indicated a significant change in the organ
distribution of endotoxin with time. Inconsistancy in
observations may reflect differences in animal species,
in the type of endotoxin or perhaps, most importantly, the
amount of particulate material in endotoxin preparations.

In this regard, it is worth noting the nonparticulate nature
of endotoxin used by Musson et al. (59).

In order to gain insight into the pOtential target tissues-
of endotoxin the distribution of LPS in LPS-responsive (C3H/st)
and LPS-unresponsive (C3H/HeJ) mice was studied and compared
(59). The results of this study demonstrated that C3H/st

spleens accumulated significantly more LPS than C3H/HeJ

12

 

spleens after intravenous injection of either an immunogenic
or a toxic dose of LPS. After a toxic dose of LPS, there is
more LPS in C3H/st lymph nodes, adrenals, lungs, kidneys, and
heart than in the corresponding C3H/HeJ tissue. Musson et
al. (59) believe that the differential accumulation of LPS
in the tissues of these two strains may be the reason for the
decreased responses of C3H/HeJ mice to LPS.

The liver contains a number of cell types. While the
macrophages (Kupffer cells) are thought to be the main
reservoir of in viva clearance (41,47) there is evidence that
liver hepatocytes are also important for the clearance of
endotoxin (87,90). With autoradiography, Willerson et al. (87)
demonstrated the presence of l4C-labeled endotoxin in liver
hepatocytes as well as Kupffer cells.

Zlydaszyk and Moon (90) found that over 80% of intra-
venously injected 51Cr—labeled endotoxin was trapped in livers
of normal mice 1 hr after the injection. Liver fractionation
studies revealed that nearly 45% of the labeled toxin was
associated with cell nuclei within the liver, 20% with the
mitochondrial-lysosomal fraction and approximately 30% with
the cell sap. To distinguish the distribution of labeled
endotoxin between parenchymal and Kupffer cells, a population
of mixed liver cells was exposed to 1% pronase for 1 hr at
37°C to lyse the parenchymal cells. After this treatment all
the counts that were not sedimentable were assumed to be
parenchymal cell associated. It was found that over 75% of

the counts were parenchymal cell associated and the Kupffer

l3

cells contained only 25% of the labeled endotoxin. In an

in vitro study they tested the endotoxin uptake of six non-
reticuloendothelial system tissue culture cell lines. All
cell lines were able to internalize endotoxin after 3 hr
incubation at 37°C. Once internalized, the labeled endotoxin
was predominantly associated with the nuclear fraction.
Zlydaszyk and Moon (90) believe that both in 3139 and in zitrg
nonreticuloendothelial system cells, as well as reticuloendo-
thelial cells, can sequester endotoxin.

Pathogenesis of Endotoxemia

 

 

Several investigators have tried to link the pathogenesis
of endotoxin to its effects on mitochondria. Pioneer work of
Fonnesu and Severi (27,28) has shown that mitochondria
isolated from livers of endotoxin-poisoned rats had a depres—
sed rate of oxidative phosphorylation compared with those from
normal rats. Harris et al. (36) isolated mitochondria from
beef heart and added endotoxin directly to mitochondria. They
observed a loss of respiratory control, inhibition of res-
piration of NADH-linked substrates, inhibition of coupling of
phosphorylation to terminal electron transport, and a change
in morphology of the inner membrane of mitochondria. De Palma's
et al. (24) results indicated that endotoxin altered the
structure of hepatic mitochondria and depressed their energy
production.

One of the characteristics of endotoxin shock is lowered
blood flow to vital organs such as brain and kidney. Mela et al.

(50) investigated the effect of tissue ischemia on the

14

mitochondria isolated from guinea pig brain and kidney after

18 to 24 hr of Escherichia 9911 endotoxemia. They found that
at 18 hr brain mitochondrial Ca4-+ transport was inhibited by
35%, while kidney mitochronria transport function was still
normal. At 24 hr, both brain and kidney mitochondria exhibited
severe inhibition of Ca++ transport and ATP synthesis.

Mela et al. (50) believe that because of the brain's sensi—
tivity to tissue isochemia, it becomes an important metabolic
target organ in septic and endotoxic shock.

Several metabolic changes after endotoxin poisoning have
been detected in liver parenchymal cells (12-14,l6,52-55).
These metabolic changes may be significant contributing factors
to the pathophysiology of endotoxemia.

Bacterial endotoxin has been known for many years to
lower the carbohydrate reserves in the mammalian host (51).
Abnormal carbohydrate metabolism, characterized by an
initial hyperglycemia, followed by a prompt decrease in
blood sugar to hypoglycemic levels and a concomitant
decrease in liver glycogen (l6), contributes to the patho—

genesis of endotoxemia.

Protective Role of Adrenocortical Hormones Against Endotoxin
Poisoning

The importance of adrenocortical (ACTH) hormones in an
animal's response to endotoxin has been established by removal
of the adrenal gland, which causes a thousand fold decrease

in the LD50 of endotoxin. ACTH is used therapeutically in

15

 

 

the.treatment of gram negative septicemic shock. Pioneer
work of Duffy and Morgan (25) established the protective
effects of ACTH and cortisone against the febril response of
endotoxin in rabbits. They found that the dosage and the
pretreatment of animals with ACTH or cortisone are two
critical factors with respect to the ability of the hormone
to alter the toxic effects of endotoxin. Geller et al. (31)
supported Duffy's et al. results by demonstrating that
pretreatment of mice with 5 mg cortisone acetate 1 hr prior
to a lethal dose of endotoxin consistantly protected 60 to
90% of the mice. Berry (7) found that the injection of
endotoxin and 5 mg cortisone acetate simultaneously protected
78% of mice from the lethal effect of endotoXin. If corti-
sone acetate was give 4 hr after endotoxin injection only 16%
of the animals survived. These results support the importancew
of time of administration and dosage of cortisone in altering
the toxic effect of endotoxin.

Several investigators have attempted to elucidate the
protective mechanism of corticosteroids in lethal endotoxemia.

SlCr-

Ribble et al. (63) investigated the distribution of
labeled endotoxin in mice which were pretreated with 5 mg of
cortisone acetate 8 hr prior to the injection of endotoxin.
Results from this study indicated that there was no significant
difference in the tissue distribution of endotoxin in cortisone
pretreated animals compared to untreated controls." Cremer and

Watson (23), used fluorescent antibody techniques to detect

the alteration in the tissue distribution of intravenously

16

 

injected endotoxin in cortisone acetate-pretreated animals.
Their results indicated that the distribution of endotoxin in
the RES of the rabbits studied was unchanged following
cortisone acetate pretreatment. There was no significant
difference in the rate of phagocytosis by steroid treated
animals compared to untreated controls; however, degradation
and elimination seemed impaired. Agarwal (2) found that
endotoxin inhibits the binding of tritiated cortisone to

nuclei in liver and spleen of Swiss albino mice. He suggested
that cortisone and endotoxin may compete for the same receptor
or_proreceptor entities. Cumulatively, these data do not
support the idea that the antagonistic nature of glucocorticoids
in endotoxin poisoning is related to the alteration in the
tiséue distribution of endotoxin. The activity of the

steroid hormones in the maintenance of near-normal carbohydrate
levels in endotoxin-poisoned mice (see above) suggests a

possible action at the metabolic level.

The Mediation of Endotoxemic Effects

 

There are uncertainties as to whether the parenchymal
cells are the direct targets of the endotoxin or whether the
toxic effects of endotoxin on parenchymal cells are mediated
by factors within the circulatory system.

It has been known for several years that endotoxin
antagonizes the glucocorticoid induction of hepatic trypto-
phan oxygenase (TO) and phosphoenolpyruvate carboxykinase
(PEPCK) (7). Endotoxin has a similar effect on glycogen
synthase (49) but does not have any effect on tyrosin amino-

transferase (TAT). _By developing and using specific antiserum

17

 

to T0 and PEPCK Rippe and Berry (66,67) and Berry and Rippe
(ll) established that endotoxin inhibits the synthesis of
these enzymes rather than inactivating them in the liver.
Shtasel and Berry (74) determined the effect of cortisone'and
endotoxin, singly and in combination on the ribonucleic acid
(RNA) synthesis in livers of mice by measuring the incor-

14

poration either of inorganic 32P or of C-orotic acid into

the RNA. The effect of these agents on protein synthesis was

l4C-leucine. The results of this

evaluated with the use of
study indicated that bacterial endotoxin increased the
synthesis of RNA and protein in the liver. Cortisone increased
the uptake of isotOpe into liver RNA but depressed the
incorporation of leucine into hepatic protein. Shtasel and
Berry (74) assumed that since there was a net increase in
hepatic protein, and mRNA synthesis, the block in the hepatic
enzyme induction in response to endotoxin must be selective.
They have suggested that the block in the hepatic enzyme
induction is mediated. Berry (8) considered interferon as a
possible mediator of endotoxin action but their results
indicated that mice injected with either living or ultraviolet—
irradiated Sindbis virus were unable to induce TO, PEPCK and
TAT in response to hydrocortisone injection. Berry et al.

(15) and Moore and Berry (56) found that synthetic double
stranded RNA poly I:C, an inducer of interferon, increased

the sensitivity of mice to LPS and mimicked the antagOnism of

endotoxin on.glucocorticoid induction of PEPCK and TO.

18

 

Recent reports have suggested the existance of a plasma
mediator called glucocorticoid—antagonizing factor (GAE) (10).
Serum from endotoxin-poisoned mice was able to inhibit the
hormonal induction of PEPCK in endotoxin-tolerant mice, whereas
endotoxin itself did not, suggesting that endotoxin—tolerant
animals fail to produce GAF in response to LPS but if GAF is
provided exogenously they are able to respond normally.

Moore et al. (57) used antimacrophage' serum to demonstrate
the role of macrophages in the production of GAF. Moore et

al (57) injected mice with mouse antimacrophage serum and
observed a full induction of T0 and PEPCK in response to
concurrent administration of hydrocortisone and endotoxin.
Additional data (57) also indicate that the supernatant

fluid from an overnight culture of macrophages incubated in
the presence of endotoxin, also inhibited the hormonal induction
of PEPCK in endotoxin‘tolerant mice. ‘Because of these results,
Moore et al. have suggested that macrophages are the source

of GAF. Since endotoxin fails to block the glucocorticoid
induction of T0 and PEPCK in nu/nu (nude) mice, Moore et al
(58) used these animals to demonstrate an apparent need for

T cells in the release of CAP. When serum or peritoneal

cells from zymosan-pretreated animals was combined with

endotoxin in_vitro and introduced into nude mice, PEPCK

 

induction by glucocorticoids was blocked. From these obser—

vations they concluded that either T-cells are needed for GAF

formation or that macrophages of athymic mice are deficient in

their ability to produce GAF.' Goodrum et al. (34) found that

19

 

indomethacin, a nonsteroidal, anti-inflammatory agent suppresses
the appearance of GAF in serum as well as its production by
peritoneal exudate cells. Indomethacin did_not interfere

with the animal's response to GAF. It is not clear how
indomethacin prevents the production of GAF. Berry et al.

(9) tried to analyze the mode of action of GAF by examining

its effect on PEPCK induction in mice and hepatoma cell

cultures. They tentatively suggest that GAF acts at the

level of transcription. GAF is able to inhibit the induction

of only PEPCK, not T0 in both conventional and athymic mice (79).

Whether GAF and interferon are identical is not clear.’

Mechanism of Action of Steroid Hormones

 

Steroid hormones play a central role in development and
physiological regulation in animals ranging from arthropods
to primates. A large body of circumstantial evidence suggests
that all steroid hormones act via basically the same mechanism.
Because of their lipophilic nature, steroids readily pass
into the cytoplasm of the target cells where they bind to
specific receptors (88). This binding process results in an
"activated" state of the steroid receptor which now has an
increased affinity for chromosomal binding sites in the
nucleus. The consequence of the presence of steroid-receptor
in the nucleus is determined by the particular biological
response characteristic of the target tissue. It has been
widely assumed that because of their lipophilic properties,

the steroids enter the cells by diffusion, and the cell

membrane provides little or no barrier to them. Target

20

tissue contains receptors that avidly bind the hormones.

This observation has led to the assumption that the greater
accumulation of hormone in target tissues results from their
ability to retain steroids rather than a difference in the
uptake of the hormone (35). When steroids enter the cells
they bind both to receptors as well as many nonspecific
proteins. The binding of steroid to receptor is specific and
characterized by a high affinity. It is believed that a
conformational change in the receptor ocdurs when the steroid

is bound. By using sucrose density gradients a good body of

 

evidence indicates a gross change in the size of the receptor
after estrogen binding and exposure to temperatures of 20°C or
above. The estrogen-receptor complex may take a variety of
forms (48, SS, GS, or 88) in the cytoplasm, and it is not
clear in which form the receptor exists prior to estrogen
binding or entry into the nucleus (35). Translocation of
steroid-receptor complexes to nuclei occurs via nuclear

pores and does not seem to be an energy-dependent process
because various metabolic inhibitors have no effect on the
process (35).

One major unanswered question is the chemical nature of
the chromatin binding site for the receptor steroid complex.
Several lines of evidence suggest that glucocorticoid-receptor
complexes bind directly to DNA. Baxter et al. (5) found that
receptor—steroid binding to isolated hepatoma tissue culture
(HTC)-cell nuclei was greatly diminished by pretreatment of

nuclei with DNase, but if complexes were bound to nuclei prior

21

to DNase treatment, they remained bound during the digestion
process. Their results also indicated that receptor-steroid
complexes bind to purified HTC—cell DNA with characteristics
similar to the binding of receptor-steroid complexes to
isolated nuclei. Tomkins and his group (89), by a new genetic
approach, provided evidence that the glucocorticoid-receptor
binding site on chromatin is DNA. They developed a gluco—
corticoid-resistant mutant cell from mouse lymphoma tissue
culture cells. Normally, these cells were killed by the
addition of glucocorticoids (75). Further studies (89)
indicated that the mutant cells fell into three classes:

one class of mutants had 8-26% of the receptors in the
nucleus, a second class of mutants had 85—93% of the receptors
in the nucleus and a third class had no receptors, whereas

the wild type cells had 50% of the receptors in the nucleus.
Those mutants with fewer nuclear receptors were called nt-
(nuclear transfer deficient) and those with more nuclear
receptors than the wild type were called nti (increased
nuclear transfer). Yamamoto et al. (89) examined the binding
of the receptors from these mutants to DNA on DNA-cellulose
cOlumns. They found that nt— mutant glucocorticoid-receptors
had little affinity for the DNA, but the nti showed approxi-
mately the same affinity as the wild type. The authors
concluded that the critical interaction of the receptors in the
nucleus occurs with the DNA. Another report from Tomkins's
(76) laboratory indicated that chromosomal proteins reduce

the number of acceptor sites while increasing the affinity

of DNA for glucocorticoid-receptors. Other reports indicated

22

 

chromosomal proteins (62), ribonucleoprotein particles (45)
and the nuclear membrane (42) as potential nuclear acceptor.
.sites for steroid-receptors. Even though several reports
(75,76,89) strongly suggest that DNA is the primary nuclear
binding site for steroid-receptors, firm conclusion in this
area may have to await more definitive biochemical assays.

The demonstration that nuclear translocation of receptors
in_zizg is a nonsaturable process, was in conflict with reports
of saturable binding to nuclei (35,88). Chamness et al. (20)

resolved this problem by demonstrating that apparent saturation

 

of nuclear sites in zitrg_was due to competition of excess
nonreceptor components for receptor binding sites. In their
experiments they added increasing concentrations of receptor
to nuclei in the presence of constant protein concentration.
By eliminating the differential effect of the competing
nonreceptor components in this way, they observed that a
constant number of the receptors bound to nuclei regardless
of the receptorznucleus ratio. These results clearly demon-
strated that nuclear binding of receptors was not saturable,
thus conforming more closely with in_vizg observations. The
observation of Chamness et a1. (20) led the way for the
discovery of an inhibitory material which is present in
cytosol and which inhibits the binding of steroid-receptor
complexes to nuclei, chromatin and DNA. The experiments of
Simons et al. (76) showed that HTC-cell cytosol contains a

material which inhibits the binding of homologous receptor—

 

steroid complex to HTC-cell nuclei or chromatin and DNA.

The inhibitors appear to be macromolecular and can be

23

removed without destroying the receptor-steroid complex.
Simons et al. (76) suggested that the most likely mechanism
of inhibitor action involved association of inhibitor with
the receptor to diminish the affinity of the complex for
acceptor sites.

Baily et al. (4) have described another type of steroid-
receptor inhibitor which is a low molecular weight, thermo-
stable (resistant at 100°C for 30 min) molecule. Their primary
results have indicated that the inhibitor acts on the

activation process of steroid-receptor complexes.

 

Rationale for the Present Study

In summary, the mechanism by which steroid hormones cause
specific gene transcription and exert their biological activity
is not well advanced, but the basic features of steroid action
have been outlined and are similar for all steroids. The
initial steps in hormone action involve the rapid entry of the
steroid into the cells followed by binding to cytoplasmic
receptor proteins. The receptor-steroid complexes eventually
become concentrated in the cell nucleus and this association
apparently results in an increase in specific mRNA. The product
of translation of mRNA leads to the biological reactions (76).
Whatever the precise mechanism of steroid action may be, it _
is important to note that in endotoxin shock early glucocorti-
coid therapy has been shown to be protective against endotoxin
lethality (7,25,31). The ig_yitrg_model of the steroid-
receptor system is a much more controllable model than an in
2122 system. In theory the in yitrg_system provides an

excellent model from which to gain insight into the pathogenic

24

mechanisms of bacterial endotoxin and the protective nature
of glucocorticoids in endotoxin shock. Studies that
critically analyze the effect of endotoxin on glucocorticoid
binding might also be of value in elucidating the mechanism
of endotoxin action at the molecular level.

Despite extensive effort in the past two decades, the
mechanism(s) by which bacterial endotoxin exerts its patho-
physiological changes has not been definitively explained.
Several metabolic changes in liver parenchymal cells after
endotoxin poisoning (12-14,16,52-55) have been considered
significant contributors in the overall pathophysiology of
endotoxin shock. Whether these changes are mediated through
soluble products released from macrophages or other cells after
phagocytosis of endotoxin (8,9) or whether liver parenchymal
cells are directly affected by endotoxin is not clear (90). It
is known that substantial amounts of endotoxin are trapped in
the liver in_ziyg and become associated with parenchymal cells
(87,90). If endotoxin is directly affecting liver parenchymal
cells, the in give and in_zitrg interaction of endotoxin with
parenchymal cells should be similar. Beginning from the
hypothesis that interaction of endotoxin with non-RES hepatic
cells is basically similar in 2122 and in_zitrg, experiments
were designed to gain insight into the nature of this inter-
action. Experimentally, toxic and a variety of modified
endotoxin were used to test the validity of this hypothesis.

In early preliminary experiments attempts were made to
make primary cell culture preparations of parenchymal cells

directly from the liver to use in in_vitro studies on endo—.

25

toxin uptake. Cell viability was very poor and in vitro
uptake studies yielded unreliable data. Since no permanent
non-transformed cell lines were available, the decision was
made to do the bulk of our studies on the well characterized
hepatoma tissue culture (HTC) cell line. This parenchymal-
derived cell line seemed particularly well suited for our
studies for a number of reasons. First, HTC-cells are a
parenchymal-derived hepatoma cell line and in our study we
were particularly interested in the direct interaction of
endotoxin with non—phagocytic hepatic cells. Several reports
(87,90) have indicated that substantial amounts of endotoxin
trapped in the liver in zivg_become associated with parenchymal
cells. Parenchymal cells exhibit most of the metabolic changes
associated with endotoxin shock thus making hepatic derived cell
lines an excellent candidate in which to evaluate and compare the
basic interactions of endotoxin with cells 1E.XEE£2' Second,
this cell line is the most extensively characterized hepatoma
cell line with respect to interaction with steroid hormones in
21353. The cells are steroid-responsive (5) and hence provide
an excellent cell line to explain endotoxin's ability to inhibit
the induction of selected hepatic enzymes by glucocorticoids (8).
The cell line does not behave exactly like hapatic cells
in vgzg_in all respects. For example, two glucocorticoid
inducible enzymes namely TO and PEPCK are not found in these
cells. Despite these selected drawbacks, the fact that these
are a suspension culture cell line and demonstrate selected

responses consistent with those in_vivo made them by far the

 

best choice of cells to use at the inception of our study.

26

LITERATURE CITED

Adams, G. A. and P. P. Singh. 1970. Structural
features of lipid A preparations isolated from,

Escherichia coli and Shigella flexneri. Biochim.

 

 

Biophys. Acta 202:553-555.

Agarwal M. K. 1972. Receptor activity and cortisone
binding following reticuloendothelial functional
alteration in mice._ Int. J. Biochem.‘ 3:408-412.

Anderson, J., F. Melchers, C. Galanos and O. Lfideritz.

1973. The mitogenic effeCt of lipopolysaccharides on

bone marrow-derived mouse lymphocytes. Lipid A as the

mitogenic part of the molecule. J. Exp. Med. 137:943-953.
Bailly, A., N. Sallas and E. Milgram. 1977. Low
molecular weight inhibitor of steroid receptor acti—
vation. J. Biol. Chem. 252:858-863.

Baxter, J. D., G. G. Rousseau, M. C. Benson, R. L.
Garcea, J. Ito and G. M. Tompkins. 1972. Role of

DNA and Specific cytOplasmic receptors in glucocorti-
coid action. Proc. Natl. Acad. Sci. USA 69:1892-1896.
Benedetto, D. A., J. w. Shards and D. o. Shah. 1973.
The interaction of bacterial lipOpolysaccharide with
phospholipid bilayers and monolayers. Biochim.

Biophys. Acta 298:145-157.

Berry. L. J. 1975. Bacterial toxins. Lloydia 38:8-20.
Berry, L. J. 1977. Bacterial toxins. Critical

reviews in Toxicology. 5:239-318.

27

10.

11.

12.

13.

14.

Berry, L. J., K. J. Goodrum, C. W. Ford, I. G. Resnick
and G. M. Shackleford. 1980. Partial characterization
of glucocorticoid antagonizing factor in hepatoma
cells, pp. 77-81. In Schlessinger (ed), Microbiology.
American Society for Microbiology, Washington, D.C.
Berry, L. J., R. N. Moore, K. J. Goodrum and R. E.
Couch, Jr. 1977. Cellular requirements for enzyme
inhibition by endotoxin in mice, pp. 321-325.

In Schlessinger (ed), Microbiology, American Society
for Microbiology, Washington, D.C.

Berry, L. J..and D. F. Rippe. 1973. Effect of endotoxin

on induced liver enzymes. J. Infect. Dis. 128:8118-8121.
Berry, L. J. and D. S. Smythe. 1959. Effects of bacterial
endotoxin on metabolism. II. Protein carbohydrate
balance following cortisone. Inhibition of intestinal
absorption and adrenal response at ACTH. J. Exp. Med.
110:407-418.

Berry, L. J. and D. S. Smythe. 1963. Effects of
bacterial endotoxin on metabolism. IV. The role of
tryptOphan pyrrolase in response of mice to endotoxin.

J. Exp. Med. 118:587-603.

Berry, L. J. and D. S. Smythe. 1964. Effects of
bacterial endotoxin. VIII. Enzyme induction and

cortisone protection. J. Exp. Med. 120:721-732.

28

 

15.

16.

17.

18.

19.

20.

Berry, L. J., D. S. Smythe, L. S. Colwell,

R. J. Schoengold and P. Actor. 1971. Comparison of

the effects of a synthetic polyribonucleotide with the

effects of endotoxin on selected host responses.

Infect. Immun. 3:444-448.

Berry, L. J., D. S. Smythe and L. G. Young. 1957.

Effects of bacterial endotoxin on metabolism. I.

Carbohydrate depletion and the protective role of

cortisone. J. Exp. Med.

110:389-405.

Braude, I. A., F. J. Carey, M. Zalesky and D. Sutherland.

1955. Studies with radioactive endotoxin. I. The use

51

of Cr to label endotoxin of Escherichia coli. J.

 

Clin. Invest. 34:850-857.

Braude, I. A., F. J. Carey and M. Zalesky. 1955.

I. Studies with radioactive endotoxin. II. Correlation

of physiologic effects of radioactivity in rabbits

injected with lethal doses of E. coli endotoxin labeled

with radioactive sodium chromate. J. Clin. Invest.

34:858-866.

Burton, A. J. and H. E. Carter. 1964. Purification

and characterization of the lipid A component of the

lip0polysaccharides from Escherichia coli. Bio-

chemistry 3:411-418.

 

Chamness, G. C., A. W. Jennings and W. L. McGuire.

1974. Estrogen receptor binding to isolated nuclei.

A nonsaturable process.

29

Biochemistry 13:327-331.

 

21.

22.

23.

24.

25.

26.

27.

28.

Chedid, L., R. C. Skarnes and M. Parant. 1963.
Characterization of 51Cr-labeled endotoxin and its
identification in plasma and urine after parenternal
administration. J. Exp. Med. 117:561-571.

Ciznar, I. and J. W. Shands. 1974. Effect of alkaline
treated lipopolysaccharide on erythrocyte membrane
stability. Infect. Immun. 4:362-367.

Cremer, N and D. Watson. 1957. Influence of stress on
distribution of endotoxin in RES determined by fluores—
cein antibody technic. Proc. Soc. Exp. Biol. Med.
95:510-513.

De Palma, R. G., Y. Harano, A. V. Robinson and

W. D. Holden. 1970. Structure and function of hepatic
mitochondria in hemorrhagie and endotoxemia. Surg.
Forum 21:3-6.

Duffey, B. J. and H. R. Morgan. 1951. ACTH and corti-
sone aggravation or suppression of the febrile response
of rabbits to bacterial endotoxin. Proc. Soc. Exp.
Biol. Med. 78:687-689.

Dumont, F. 1975. Effect of stimulation with bacterial
lipOpolysaccharide on the surface-charge of mouse
B-lymphocytes. Ann. Immunol. (Paris) 126C:453-456.
Fonnesu, A. and C. Severi. 1956. Lowered p/o ratios
with mitochondria isolated from livers showing cloudy
swelling. Science 123:324-324._

Fonnesu, A. and C. Severi. 1956. Oxidative phosphory-
lation in mitochondria from livers showing cloudy

swelling. J. BiOphys. Biochem. Cytol. 2:293.

30

29.

30.

31.

32.

33.

34.

35.

Freer, J. H. and M.R.J. Salton. 1971. The anatomy and
chemistry of gram negative cell envelopes, pp. 67-126.
In. G. Weinbaum, S. Kadis and S. J. Agl (ed), The
microbial toxins, vol. 4. Academic Press Inc. NY.
Freudenberg, M. A., T. C. Bog-Hansen, U. Back and C.
Galanos. 1980. Interaction of lipopolysaccharides
with plasma high-density lipoprotein in rats. Infect.
Immun. 28:373-380.

Geller, P., E. Merril and E. Jawetz. 1954. Effects of
cbrtisone and antibiotics on lethal action of endotoxins
in mice. Proc. Soc. Exp. Biol. Med. 86:716-719.
Gmeiner, J., O. Luderitz and O. Westphal. 1969.

Biochemical studies on lipopolysaccharides of Salmonella

 

R mutants. VI. Investigations on the structure of the

 

lipid A component. Eur. J. Biochem. 7:370-379.
Goodman, G. W. and B. M. Sultzer. 1977. Mild alkaline
hydrolysis of lipopolysaccharide endotoxin enhances its
mitogenicity for murine B-cells. Infect. Immun.
17:205-214.

Goodrum, K. J., R. N. Moore and L. J. Berry. 1978.
Effect of indomethacin on the response of mice to
endotoxin. J. Reticuloendothel. Soc. 23:213-221.
Gorski, J. and F. Gannon. 1976. Current models of
steroid hormone action: a critique. Annu.Rev.

Physiol. 38:425-450.

31

36.

37.

38.

39.

40.

41.

Harris, R. A., D. L. Harris and D. E. Green. 1968..

Effect of Bordetella endotoxin upon mitochondrial

 

respiration and energized processes. Arch. Biochem.
Biophys. 128:219-230.

Hellerquist, C. G. and A. A. Lindberg. 1971. Struc-
tural studies on the O-specific side chains of the cell

wall lipOpolysaccharide from Salmonella typhimurium.

 

Carbohydr. Res. 16:39-48.
Hellerquist, C. G., B. Lindberg, S. Svensson, T. Holme

and A. A. Lindberg. 1969. Structural studies on the

 

O-specific side chains of the cell wall lipopolysac-

charide from Salmonella typhimurium LT2. Carbohydr.

 

Res. 9:237-241.

Herring, W. B., J. C. Herion, R. I. Walker and J. G.

Palmer. 1963. Distribution and clearance of cir-

culating endotoxin. J. Clin. Invest. 42:79-87.

Homma, J. Y., T. Katsura, S. Hosoya, Y. Miyazaki, K.
32

Kimura and N. Saito. 1951. P -labeled endotoxin of

Pseudomonas aeruginosa. Preliminary report-isolation

 

of radioactive endotoxin from culture filtrate on
synthetic media containing P32-labeled phosphate and
its distribution in mice. Jap. J. Exp. Med.
21:381-393. I

Howard, J. G., D. Rowley and A. C. Wardlaw. 1958.
Investigations on the mechanism of stimulation of non-
specific immunity by bacterial lipOpolysaccharides.

Immunology 1:181-203.

 

32

42.

43.

44.

45.

46.

47.

48.

49.

Jackson, V. and R. Chalkley. 1974. The binding of
extradiol-17B to the bovine endometrial nuclear mem-
brane. J. Biol. Chem. 249:1615-1626.

La Noue, K. P., A. D. Mason, Jr and J. P. Daniels.

1968. The impairment of gluconeogenesis by gram
negative infection. Metabol. Clin. Exp. 17:606-611.
Levy, E., F. C. Path and B. H. Rueliner. 1967. Hepatic
changes produced by a single dose of endotoxin in the
mouse. Amer. J. Pathol. 51:269-284.

Liang, T. and S. Liao. 1974. Association of the

 

uterine l7B-estradiol-receptor complex with ribonucleo-
protein in zitrg and in XEZE- J. Biol. Chem.
249:4671-4678.

Lfideritz, O, K. Jann and R. Wheat. 1968. Somatic and
capsular antigens of Gram-negative bacteria, pp. 105-221.
IE_M. Florkin, E. H. Stotz (ed), Comprehensive .
biochemistry, vol 26A. Amsterdam:Elsevier.

Mathison, J. C. and R. J. Ulevitch. 1979. The
clearance, tissue distribution, and cellular localization
of intravenously injected lipopolysaccharide in rabbits.
J. Immunol. 123:2133-2143.

Mayne, Y. C. and R. Jones. 1957. Tissue distribution

of 14

C-labeled bacterial polysaccharide in guinea pig,
rat, mouse and rabbit. Proc. Soc. Exp. Biol. Med.
98:734-737.

McCallum, R. E. and L. J. Berry. 1973. Effects of

endotoxin on gluconeogenesis, glycogen synthesis and

 

liver glycogen synthase in mice. Infect. Immun.
7:642-654.

33

50.

51.

52.

53.

54.

55.

56.

57.

Mela, L., A. Burrows, A. Knee and K. Wrobel-Kuhl. 1979.
Ischemic injury of brain and kidney mitochondria in'
endotoxin shock. Circ. Shock 6:183 (Abstract).

Menten, M. L. and H. M. Manning. 1924. Blood sugar
studies on rabbits infected with organisms of the
enteritidioparatyphoid B.group. J. Med. Res. 44:675-
693.

Moon, R. J. 1971. Carbohydrate metabolism and sur-
vival of endotoxin-poisoned mice given tryptophan.

Infect. Immun. 5:288-294.

 

Moon,-R. J. 1971. Tryptophan oxygenase and tryptophan
metabolism in endotoxin-poisoned and allopurinol treated
mice. Biochim. Biophys. Acta 230:342-348.

Moon, R. J. and L. J. Berry. 1968. Effect of trypto-
phan and selected analogues on body temperature of
endotoxin-poisoned mice. J. Bacteriol. 95:764-770.
Moon, R. J. and L. J. Berry. 1968. Role of tryptophan
pyrrolase in endotoxin-poisoning. J. Bacteriol.
95:1247-1253.

Moore, R. N. and L. J. Berry. 1976. Inhibited hormonal
induction of hepatic phosphoenolpyruvate carboxykinase
in poly I:poly C treated mice, and endotoxin-like
glucocorticoid antagonism. Experientia 32:1566-1568.
Moore, R. N., K. J. Goodrum and L. J. Berry. 1976.
Mediation of endotoxin effects by macrophages. J.

Reticuloendothel. Soc. 19:187-197.

34

 

58.

59.

60.

61.

62.

63.

64.

Moore, R. N., K. J. Goodrum, L. J. Berry and J. R.
McGhee. 1977. An.abnormal response of nude mice to
endotoxin. J. Reticuloendothel. Soc. 21:271—278.
Musson, R. A., D. C. Morrison and R. J. Ulevitch.

1978. Distribution of endotoxin (lipopolysaccharide)
in the tissue of lipopolysaccharide-responsive and
unresponsive mice. Infect. Immun. 21:448-457.

Neter, B., O. Westphal, O. Lfideritz, E. A. Gorzynski
and E. Eichenberger. 1956. Studies of enterobacterial
lipopolysaccharides. Effects of heat and chemicals on
erythrocyte modifying, antigenic, toxic and pyrogenic
prOperties. J. Immunol. 76:377-385.

Nowotny, A. 1969. Molecular aspects of endotoxic
reactions. Bacteriol. Rev. 33:72-98.

O'Malley, B. W. and A. R. Means. 1974. Female steroid
hormones and target cell nuclei. The effects of
steroid hormones on target cell nuclei are of major
importance in the induction of new cell functions.
Science 183:610-620.

Ribble, J. C., M. Zalesky and A. I. Braude. 1959.

51Cr-labeled endotoxin in cortisone

Distribution of
treated mice. Bull. Johns Hopkins Hosp. 105:272-283.
Rietschel, E. T. 1975. Chemical structure and

biological activity of endotoxins (lipopolysaccharides)

and lipid A. Arch. Pharmacol. 287:73-84.

35

 

 

 

65.

66.

67.

68.

69.

70.

71.

Rietschel, E. T., H. Gottert, O. Lfideritz and

O. Westphal. 1972. Nature and linkages of the fatty
acids present in the lipid A component of salmonella
lipopolysaccharides. Eur. J. Biochem. 28:166-173.
Rippe, D. F. and L. J. Berry. 1972. Study of
inhibition of induction of phosphoenolpyruvate
carboxykinase by endotoxin with radial immunodiffusion.
Infect. Immun. 6:766-772.

Rippe, D. F. and L. J. Berry. 1973. Immunological
quantitation of hepatic tryptophan oxygenase in
endotoxin poisoned mice. Infect..Immun. 8:534-539.
Rubenstein, H. S., J. Fine and A. H. Coons. 1962.
Localization of endotoxin in the walls of the peri-
pheral vascular system during lethal endotoxemia.
Proc. Soc. Exp. Biol. Med. 111:458-467.

Ryser, H.J.P. 1968. Uptake of protein by mammalian
cells: An underdeveloped area. The penetration of
foreign proteins into mammalian cells can be measured
and their functions explored. Science 159:390-396.
Ryser, H.J.P., T. E. Termini and P. R. Barnes. 1976.
Polynucleotide aggregates enhance the transport of
protein at the surface of cultured mammalian cells.
J. Cell. Physiol. 82(2):221-228.

Seligman, A. M., M. J. Shear, J. Leiter and B. Sweet.

1948. Chemical alteration of polysaccharide from

 

Serratia marcescens. I. Tumor-necrotizing polysac-
charide tagged with radioactive iodine. J. Natl.

Cancer Inst. 9:13-18.

36

72.

73.

74.

75.

76.

77.

78.

Shands, J. W., Jr., V. Miller and H. Martin. 1968.
The hypoglycemic activity of endotoxin. I. Occurrence
in animals hyperreactive to endotoxin. Proc. Soc. Exp.
Biol. Med. 130:413-417.

Shands, J. W., Jr., V. Miller, H. Martin and

V. Senterfitt.~ 1969. Hypoglycemic activity of endo-
toxin. II. Mechanism of phenomenon in BCG infected
mice. J. Bacteriol. 98:494-501.

Shtasel, T. F. and L. J. Berry. 1969. Effect of
endotoxin and cortisone on synthesis of ribonucleic
acid and protein in livers of mice. J. Bacteriol.
97:1018-1025.

Sibley, C. H. and G. M. Tomkins. 1974. Isolation of
lymphoma cell variants resistant to killing by
glucocorticoids. Cell 2:213-220.

Simons, S. S., H. M. Martinez, R. L. Garcea,

J. D. Baxter and G. M. Tomkins. 1976. Interactions
of glucocorticoid receptor. Steroid complexes with
acceptor sites. J. Biol. Chem. 251:334-343.

Skarnes, R. J. '1970. Host defense against bacterial
endotoxemia mechanism in normal animals. J. Exp. Med.
132:300-316.

Skidmore, B. J., M. J. Chiller, D. C. Morrison and

W. O. Weigle. 1975. Immunologic properties of bacterial
lipOpolysaccharide (LPS): Correlation between the
mitogenic, adjuvant, and immunogenic activities. J.

Immunol. 114:770-775.

37

79.

80.

81.

82.

83.

84.

85.

Snyder, I. S., M. Deters and J. Ingle. 1971. Effect
of endotoxin on pyruvate kinase activity in mouse
liver. Infect. Immun. 4:138-142.

Springer, G. F. and J. C. Adye. 1975. Endotoxin-
binding substances from human leukocytes and platelets.
Infect. Immun. 12:978-986.

Springer, G. F., J. C. Adye, A. Bezkorovainy and

B. Jirgensons. 1974. Properties and activity of the
lipopolysaccharide receptor from human erythrocytes.

Biochemistry 13:1379-1389.

 

Springer, G. F., J. C. Adye. A. Bezkorovainy and

J. R. Murthy. 1973. Functional aspects and nature of
the lipOpolysaccharide-receptor of human erythrocytes.
J. Infect. Dis. 128:S202-8212. I

Springer, G. F., S. V. Huprikar and E. Neter. 1970.
Specific inhibition of endotoxin coating of red cells
by a human erythrocyte membrane component. Infect.
Immun. 1:98-108.

Tauber, H., H. Russel and W. J. Guest. 1961. Nature
of polysaccharides obtained from endotoxins by hydroxyl-
aminolysis. Proc. Soc. Exp. Biol. Med. 107:964-965.
Ulevitch, R. J. and A. R. Johnston. 1978. _The modifi-
cation of biophysical and endotoxic properties of
bacterial lipopolysaccharides by serum. J. Clin.

Invest. 62:1313-1324.

38

 

86.

87.

88.

89.

90.

Ulevitch, R. J., A. R. Johnston and D. B. Weinstein.
1979. New function for high density lipoproteins.
Their participation in intravascular reactions of
bacterial lipopolysaccharides. J. Clin. Invest.
64:1516-1524. I

Willerson, J. T., R. L. Trelstad, T. Pincus, S. B. Levy
and S. M. Wolff. 1970. Subcellular localization of
Salmonella enteritidis endotoxin in liver and spleen of
mice and rats. Infect. Immun. 1:440-445.

Yamamoto, K. R. and B. M. Alberts. 1976. Steroid

 

receptors: Elements for modulation of eukaryotic
transcription. Annu. Rev. Biochem. 45:721-746.
Yamamoto, K. R., M. R. Stampfer and G. M. Tomkins.
1974. Receptors from glucocorticoid-sensitive
lymphoma cell and two classes of insensitive clones:
Physical and DNA binding properties. Proc. Nat. Acad.
Sci. USA 71:3901-3905.

Zlydaszyk, J. c. and R. J. Moon. 1976. Fate of 51

Cr-
1abeled lipopolysaccharide in tissue culture cells and

livers of normal mice. Infect. Immun. 14:100-105.

 

39

ABSTRACT

In vitro and In vivo Association and Subcellular Distribution

 

of Toxic and Experimentally Modified Endotoxin

 

Cell association and organ distribution of toxic and
experimentally modified endotoxin in whole animals and hepatoma
tissue culture (HTC) cells was compared. Mouse organ distri-
bution of toxic and poly L-d-ornithine mixed endotoxin indicates
that most of the endotoxin is associated with the reticuloendo-
thelial (RES) rich organs, and organ distribution of neither
one changes over a.5 hour period. Significantly-less detoxified
endotoxin is recovered from RES rich organs of mice. HTC-cell
association and nuclear transfer of toxic endotoxin is a gradual
and time-dependent process. Poly L-d-ornithine at a concen-
tration of 4 ug/ml significantly increases HTC-cell association
of endotoxin. Nuclear transfer of endotoxin mixed with poly L-a-
ornithine, like toxic endotoxin, is a gradual and time—dependent
process. HTC-cell association of alkaline-treated detoxified
endotoxin is significantly higher than the association of toxic
endotoxin and increases with time. In contrast to toxic and
poly L-a-ornithine mixed endotoxin, nuclear association of
alkaline-treated detoxified endotoxin does not significantly
change during 5 hour incubation. These observations indicate

that while tissue culture cells could provide a more controllable

4O

 

 

experimental system in which to study the fate and pathogenic
mechanism of endotoxin at the cellular and subcellular level,
HTC-cells under the conditions employed here do not yield
binding data which compare favorably to in_yiyg results.
Caution must be exercised when extrapolating in zitrg data to

the actual in_vivo action of endotoxin.

41

INTRODUCTION

A previous report from this laboratory (21) indicates
that non-RES cells such as mouse liver cells and tissue
culture cells as well as cells of the RES system, can
sequester endotoxin. If conditions could be established
whereby binding and internalization of endotoxin by tissue
culture cells would mimic in_zizg realities tissue culture
cells could provide a much more controllable model for
detailed study of the biological action of endotoxin. This
report compares toxic and a variety of experimentally
modified endotoxin preparations with respect to their in
3123 distribution among organs and tissues of mice. A more

detailed comparison of binding to hepatic tissues in vivo

 

and HTC-cells in Vitro is also presented.

 

The various endotoxin preparations include alkaline
detoxified and plasma treated endotoxin. Alkaline hydrolysis
cleaves ester linked fatty acids from lipid A and significantly
reduces lethality and pyrogenicity in 3122. Alkaline detoxifi-
cation also enhances binding to red blood cells and reduces
endotoxin particle size but does not affect its antigenic
specificity (5). Normal rabbit, human or mouse serum (or
plasma) also reduces toxicity and buoyant density of endotoxin
(18,19). The density shift is associated with the inhibition
of pyrogenic activity, the ability to produce an immediate
neutropenia and anti-complementary activity of endotoxin.

A third manipulation of endotoxin was exposure to poly-

cationic polymers. Polycationic polymers and certain

42

polynucleotide aggregates enhance the entry of macromolecules
into cultured mammalian cells (16,17). Ryser has suggested
that polycationic polymers initiate the entry of macro-
molecules into cells by making multiple attachment with the

negatively charged cell surface (16).

43

METHODS AND MATERIALS

Chemicals

 

LyOphilized Salmonella typhimurium lipopolysaccharide

 

(LPS) extracted by the Westphal procedure was obtained from
Difco Laboratories (Detroit, MI). It was stored at 4°C and
dissolved in isotonic, pyrogen free sterile saline.

251Cr04) was purchased from New England

Chromium-51 (Na
Nuclear Corp. (Boston, MA) in 2 mCi lots, with a specific
activity of approximately 300-500 Ci/g. Limulus amebocyte
lysate was obtained from Microbiological Associates
(Walkersville, MD) and stored in a lyOphilized form at -20°C.
Pyrogen-free saline was purchased from Cutter Laboratories

(Berkeley, CA). Poly-L-d-ornithine was obtained from Sigma

(St. Louis, MO).

Alkaline Hydrolysis of Endotoxin

 

Mild alkaline-treated endotoxin was prepared by dis-
solving 50 mg of LPS in 3 ml of 0.25 N NaOH. The solution was
heated at 56°C for 60 minutes (7). The modified endotoxin was
precipitated with cold anhydrous ethanol, washed three times
with ethanol, dialyzed against deionized distilled water and
lyophilized. I

In vivo Plasma-treated Endotoxin .
51

 

In_vivo plasma-treated Cr-labeled endotoxin was

prepared by injecting 15 mice intravenously (i.v.) with

51

400 ug Cr-labeled endotoxin per mouse.‘ Mice were bled into

heparinized tubes 1 minute later and the plasma prepared.

44

Plasma was incubated with excess HTC-cells at 37°C for

20 minutes to adsorb nonspecific agglutinins. After agglu—
tinin adsorption, the whole plasma was used to study HTC-cell
association and cell distribution of plasma-treated endotoxin.

51Cr-labeled endotoxin was

For some experiments,
recovered from plasma by centrifuging at 100,000xg for
10-12 hours. The pellets (CA 50% of the label usually
recovered) were resuspended in 1.8 m1 MEM and 250 ml (about

70 ug LPS) was added‘to 1.25 x 107 HTC-cells.

Preparation of Chromium Labeled Endotoxin

 

 

To 100 mg of LPS, 1.5 mCi of chromium-51 was added. The
mixture was dissolved in 10 ml of normal physiological saline
and incubated at 37°C for 48 hours with constant stirring. This
step was followed by dialysis at 4°C against deionized water
for 3 to 5 days. The labeled endotoxin was centrifuged at
8000xg for 1 hour and the supernatant was recentrifuged at
100,000xg for 8 hour. The latter pellet was suspended in
4.0 m1 of deionized distilled water and kept at -70°C. This
preparation contained high molecular weight, highly toxic
lipopolysaccharide fraction (4), with a specific activity of
CA 7 uCi/ml for toxic endotoxin and 3.5 uCi/mg for 0.25 N NaOH.
treated endotoxin. For individual experiments, the extent

of 51

Cr decay was calculated and incorporated into the final
calculations (20).

51

 

Cr-labeled Endotoxin in Experimental Animals

Mice were injected intravenously with 0.20 mg of 51Cr-

51

Distribution of
Cr-labeled endotoxin mixed with 48 ug

labeled endotoxin,

45

SlCr-labeled 0.25 N NaOH

poly-L-d-ornithine, or 0.18 mg
treated endotoxin. After 1, 3 and 5 hours, mice were killed

by cervical dislocation and the amount of radioactivity determined
in liver, gall bladder, spleen, kidneys, intestine, heart,

lung, brain, carcass, skin and feet. *Subcellular distribution

of 51Cr-labeled endotoxin within the liver was determined by
homogenizing the organ in two ml of TKM-0.25 M sucrOse (0.05 M

Tris HCl, pH 7.5 at 20°C; 0.025 M KCl; 0.005 M MgCl and 0.25

2.
M sucrose) in a glass homogenizing tube with a Teflon pestle.
Two ml of the homogenate was centrifuged at 800 x g for 15
minutes. Counts per min in the pellet (crude nuclear fraction)
after one wash with TKM-0.25 M sucrose and supernatant (non-
nuclear fraction) were determined. Pure nuclei were prepared by
mixing 1.0 ml of liver homogenate with 2.0 ml of 2.3 M sucrose
in TKM. The mixture was then underlaid with 1.0 ml 2.3 M
sucrose in TKM and centrifuged for 40 minutes at 120,000

x g at 0°C (2). After centrifugation the supernatant was

poured off and the wall of the tubes were wiped with a spatula
and Kimwipe. The pellet (pure nuclei) was counted in a gamma
counter. About 50% of the nuclei were recovered in the pellet
and 50% remained in the supernatant as measured by the
distribution of DNA (3). Between 85 and 90% of the injected

counts were routinely recovered from the whole animal.

Cell Association and Intracellular Distribution of 51Cr-

 

labeled Endotoxin by Hepatoma Tissue Culture Cells (HTC)

 

HTC-cells were grown in MEM supplemented with 5% heat

inactivated fetal calf serum. The cells were harvested and

46

 

 

suspended in 25 ml MEM at a concentration of 5x105 cells per
ml. Approximately 1.2-2.5x105 cpm of either 51Cr-labeled

toxic endotoxin (23 ul of 20 mg/ml), toxic endotoxin mixed

with poly-L-a-ornithine (4 ug/ml) or alkaline treated endotoxin
(50 ul of 6.3 mg/ml) was added to the cells. The mixture was
incubated at 37°C with constant.stirring in a suspension
culture flask. At 1, 3 and 5 hour time intervals, 5 ml samples
Were removed, washed twice with Tris buffer [0.04 M Tris
(hydroxymethyl) aminomethane-hydrochloride, 0.1 M KCl, 0.001 M

MgCl pH 7.2]. The pellet was transferred to a fresh tube.

27
to avoid increased counts stuck to the glass and counted in a

 

Packard gamma counter. To determine the extent of nuclear
association of LPS, the counted cells were treated with 0.5%
Nonidet P-40 (NP-40) for 10 minutes on ice and centrifuged at
800xg for 10 minutes. Counts per minute in the supernatant
(non-nuclear fraction) and the pellet (crude nuclear fraction)
were determined.

Limulus Lysate Test

 

 

All glassware was sterilized and made pyrogen-free by
dry heat (180°C for 4 hours). Disposable sterile pyrogen-
free pipets were used for all experiments. The limulus lysate
test was performed on toxic endotoxin, endotoxin mixed with‘
poly L-a-ornithine (20 ug LPS/4 ug ornithine) and 0.25 N NaOH
treated endotoxin. Toxic endotoxin and pyrogen-free saline
were used as positive and negative controls.

Ten-fold serial dilutions of the respective samples were
prepared in pyrogen free saline. An aliquot (0.1 ml) of the

sample was mixed with an equal volume of limulus lysate and

47

 

incubated at 37°C. Changes in the opacity and clotting

were judged visually on a scale of l to 4.

Isopycnic Density Gradient

 

Isopycnic density gradient ultracentrifugation was
performed with CsCl solutions of an average density (d) of
1.40 to 1.50 g/cm3. Samples were centrifuged to equilibrium I
at 40,000 rpm for 60 hours at 20°C in the Spinco model L3-50
ultracentrifuge (Beckman Instruments Co.) with the SW 50.1
rotor. Fractions of 0.1 ml were collected and the densities
calculated from the index of refraction measured with a

refractometer (Bausch and Lomb).

Lethality Studies in Mice

 

Charles River CD-1 mice (18-22 g) sensitized with lead
acetate were used to determine LD50 of the various endotoxin
preparations. The median lethal dosage was determined by
injecting mice with 5 mg of lead acetate in 0.2 m1 saline
intraperitoneally (IP). Animals receiving lead acetate alone
served as controls. For each concentration of endotoxin, at
least six animals were used. Survivors were scored after

72 hours and the LD was calculated by the method of Reed

50
and Meunch (15).

48

 

 

RESULTS

Comparison of Selected Biological and Physical Properties of
Toxic and Experimentally Modified Endotoxin

Both similarities and differences were noted in the various
endotoxin preparations used in this study. There was no
significant difference in the limulus lysate reactivity of
endotoxin and endotoxin mixed with poly L-a-ornithine (20 ug
LPS/4 ug ornithine) (Table l). Alkaline-treated endotoxin was

about l,000-3,000-fold less toxic as indicated by both limulus

 

lysate assay and LD50 analysis. The LD of endotoxin mixed

50

with ornithine was two-fold lower than untreated endotoxin.
The buoyant density of endotoxin and endotoxin mixed with

ornithine was the same (1.51 g/cm3). There was a decrease in

the buoyant density of plasma-treated endotoxin (1.44 g/cm3)

and an increase in the density of 0.25 N NaOH treated endo-

toxin (1.59 g/cm3).

Distribution of Toxic and Experimentally_Modified Endotoxin

 

Among Organs of Normal Mice.

 

Table 2 shows the organ distribution of labeled toxic

endotoxin at 1, 3 and 5 hours after iv injection. At all time
51 '

 

points about 50% of the Cr-labeled endotoxin was found in

the liver. There was no significant difference in the organ

 

distribution of endotoxin throughout the ensuing 5 hour study.
About 7,000-8,000 cpm (12%).of the liver-associated counts

were recovered from crude nuclear fraction. Only a much

49

TABLE 1

Comparison of selected biological and physical properties

of toxic and modified endotoxin

 

 

 

LD50 in
lead acetate Limulus Buoyant
treated mice lysate density
Experimental (ug) (pg/ml) (g/cm3)
Toxic LPS* 0.31 l x 10.3 1.51
Ornithine mixed
LPS 0.15 1 x 10‘3 1.51
0.25 N NaOH .
treated LPS 1000 l 1.59
In vivo plasma
treated LPS NP** NP 1.44

 

*Lipopolysaccharide

**Not performed

 

50

smaller fraction of labeled endotoxin could be recovered from
pure hepatic nuclei which suggests that most of the labeled
endotoxin was actually bound either to the surface of nuclei
or other contaminating cellular debries. There was no gradual
increase in the nuclear transfer of endotoxin In zIzn between
1 and 5 hours as seen in hepatoma cells.

In zIyn whole organ distribution of endotoxin mixed with
ornithine, was not significantly different from endotoxin
alone through the entire 5 hour study (Table 3). Liver nuclear
association of endotoxin mixed with ornithine fluctuated in a
somewhat random manner. The significant increase in the pure
nuclei association of endotoxin mixed with ornithine at 3
hours cannot be explained.

The organ distribution of 0.25 N NaOH-treated endotoxin
differed significantly from toxic endotoxin. After 1 hour
significant differences could be observed in the liver and
spleen, and the carcass, skin and feet. Only 20—26% of 0.25 N
NaOH-treated endotoxin was associated with the liver. Sig-
nificantly less alkaline-treated endotoxin was also recovered
from the spleen and significantly more was associated with the
carcass, skin and feet than for toxic endotoxin. There was no
significant redistribution of 0.25 N NaOH-treated endotoxin
during the 5 hour study. No gradual nuclear transfer of
alkaline treated endotoxin was observed.

Association and Subcellular Distribution of Toxic and Mouse
Plasma-treated Endotoxin by HTC-cells.

Table 5 compares association and subcellular distribution

of toxic and mouse plasma-treated endotoxin within HTC-cells.

51

 

 

 

 

HVwbm N

 

 

 

 

 

OHmm: mum «Hmmcm mwmnfiwocnwos 0m noxwo mHOHLHmUoHoQ osmonowa Awoo toy»: Beam
Hake... ZGOHmHomVS nognm mmH wacnow
. meb m omHH
HCHm macaw bp<mH mpwmms xwmbmw wsnmmnpbm mmde babe meH: omfiommm mood wwmmmmfi
mum «Ham momco wwwm Hmmu qum cm» Nmmw Hmo mqwqm mpww. Nw.
H H H . . H H H H H H H H H H H
Hoe new Emww Heww mom Hmme www no» mm mon. Hhmm Hm
mmm mwem mwwuw Howom wmmu Howqm owe Hmmm Hmo wmwmm mqwm on
wwfi H H H H H H H H H H H H
mm mm» cmHH may Nmo mew mm How no Nmmm me. Hm
mwfi. wmm Ohmm mmmhw qmmw NHBO memo amp wocm Hum mammw mmem mm
H , H H_ H .H H H H H H H H
we now Sumo wmm Mum pow» mo wow Hm Nubm mam OH

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m n Bomb om ops H

mnmsmmnm mHHOH om dam Bows ocnmwsom

mHoS

me mxpmHHmenm

 

52

 

 

 

 

amopo m

Onomb mum mpmmcm upmfinposnpos 0m mp

apxom spew UOpm HIQIOHUpHUHBm “am too p: Spam

OHImeopmm HOXHO osoonoxps Amoo too

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

bdfifi.§kbopops .nocsnm mom apnoemm
. . . mob m. 85
mCHm nmcom bp<mH mppom: xpmbmm podmmnpsm mmmfid bcsm mfimpb omHommm moon wpmmomfi
mmcc mmmm mmqmmo mmmm mpwo meow pwop mmmm map cooqma pommo pm
H.3H H H H. H H H H H H H H H
pg mmp Noom mpm mom pwom poo ham Nmm mmpm mppc m
m .mmm pmmmq mome mmem upon ammo mum wwmm pmm mocpo mmmm mm
:H H H H H H H H H H H H H
mo mom mmqm mom poo mmm om cpm mm qum pmom pm
. mmw pmmpm mpmmm mmom Nppm mmmm Bop Npmm pmm .Nmmqm mmmm an
own. H H H H H H H H H H H H
mm poom mmpq mum pmm mam om mmu pm mpmm mom pm
m u 3mm:.om ops H mambmmfio mHHou om duo Bow: ocfimpbmm mnoa mpx mxpmeBoSHm
c n zon mpmbpmpOHSHpK mpmmmfiosfl mHoB m U
o u zoo meSpmpomSpr mpmmmnmsd mHoa m mum m s
a w zon mpmspmpomsnpm mpmmmfimbw mHoa m mum m 3

53

 

 

qupm m

onows mum npmmsm mpmnnpccwpo: 0m mpowlpmcmpmo o.mm 2 zoom

wnmwnmm mbmonoxp: Apmo :QV p: Spam

 

 

 

 

 

 

bpsqucorfl.opa oocsnm won ZpSCHm
mWHs m ompp
VCHm Oncmm bp<om mppom: mpmbmm ponmmwpbo mmwfid boom wflmp: anommm moon wpmmmmH
pmq pawn mmom mm» ppmm mono mmm mmm pmm Nmmqm ummp mm
p :Hm H H H H H H H H H H H H
mm who goo mm ppo . wpo ppm mm mo pmmq me po
6 Npm pmmm ppmmo mmm pomm mmpm mom mmq ow mmpmw mqpm
m 5H H H H H H H H H H H H o
mm pmp omp pom mm mmm pmo mm pm pogo mpm
m 3H0 pmm ppmo pomum mom mum mmmp mmw mom mm pmmom mom» pm
H H H H H H H H H H H H
mm mo mg» mp ppo mom mm pom pm mmq pmp q

 

 

 

 

 

 

 

 

 

 

 

 

 

m u Emma Om 065 H momsmem oHHOH 0m #50 Emma ocnmpsom whoa mpx mxpmesznm
6.0 u 3mm: om ops H mnmcmmnm mHHOH 0m dam Bomb oodmpbmm mHoB mOCH mxpmmpamsnm

54

 

 

Experimentally, 1.25 X 107 HTC-cells were mixed with about
2.5 x 105 cpm (ca 460 pg) of labeled toxic endotoxin. After
1 hour incubation at 37°C, only 63 cpm of the toxic endotoxin
became associated with 2.5 x 106 HTC-cells. By 3 and 5 hours,
association per 2.5 x 106 cells was 122 and 189 cpm respectively.
After 1 hour very few of the cell-associated counts were
recovered from the crude nuclear fraction. Nuclear transfer
of toxic endotoxin gradually increased after 1 hour so that
by 3 and 5 hours 42 and 65 cpm were associated with the crude
nuclear fraction respectively. The presence of toxic endotoxin
in both whole cells and nuclear fractions was confirmed by the
limulus lysate test (data not shown).

Table 5 also shows binding of uncentrifuged In zIzn

5

exposed endotoxin to HTC-cells. From 1.1 x 10 cpm (ca

142.5 ug/1.25 x 107 cells) of endotoxin added to cells, 82

cpm became associated with 2.5 x 106

cells after 1 hour,
gradually increasing to 255 cpm after 5 hours. Nuclear
association of plasma-treated endotoxin changed only slightly
through the ensuing 5 hours. When In nIzn plasma-treated
endotoxin was recovered by centrifugation and then incubated

4

with HTC-cells, of the 5 x 10 cpm (ca 70 ug/l.25 x 107 cells)

of endotoxin added to cells, 167 cpm became associated with

2.5 x 106

cells after 1 hour, gradually increasing to 277 cpm
after 5 hours. Relative to the starting counts this preparation
more than doubled in binding compared to uncentrifuged In KIZQ
exposed endotoxin. After 1 and 3 hour incubation there was

no significant change in the nuclear association of plasma-

treated centrifuged endotoxin. After 5 hours of incubation,

55

 

 

emopm m

manloopp mmmoopmnpo: mom ompp mpmdnpccnpos om doxpo oomowoxp: Aom :©\m¢5p omppv.

p: <p<o 305mm

 

ppmmamlnfimmwmm Amm.m to bmm\m 3p omppv mam mm.<p<o 305mm ppwmamlnummnmo

 

omwapmamomApm to hwm\m 3p omppv momonoxp: mmde p. m mum m : puocomwpo: me when

 

 

 

 

 

 

 

 

. . m Mm.<p<o ppmmam memdmo Hm <p<o ppmmam wummwmm
aoxpo mumonoxpu mamowoprm omSWHHmcmmm msmowoxpom
mxpmeSmbnmp p m m p m m . p m m
nocsnm wo.mmen mmmmo mpuom mpmmm upmup Npmmp Npqum pommq pomom pomom
SHH: H H ,H H H H H H H
mmmm pooo pomp mqmm qum wmmq mph mpm. mmm
mm pmm pmo mm poo mum pmo mom who
2:0pm ompp H ,H H H H H H H H
c H» mm pm we .cm Ho m c
po mm mm mm mm mp no mp. pom
zcopmp H H H H H H H H H
p m pw. q po po m m pp
mm mm mm mp pom pom ppm mom mom
zonincopmmfi H H .H H H H H H H
Hnmceeoc w m. m m. we we Hp m we

 

 

 

 

 

 

 

 

 

 

m u 3mm: om ova

H mdmoamnm

mHHOH 0m firm Bomb oowmpnmm

mHoB pm mxpmflpamsflm

 

 

 

105 cpm of the cell associated counts were recovered from
crude nuclear fraction.

The decrease in the amount of plasma-treated endotoxin
relative to toxic endotoxin used in these experiments reflects
the difficulty in recovering substantial amounts of endotoxin
from the blood of mice. Endotoxin is cleared very rapidly
from the circulation. The amounts used per experimental
flask reflects the total counts recoverable from pooling

plasma from at least 4 mice.

HTC-cell Association and Cell Distribution of Endotoxin Mixed

With Poly L-a-ornithine

 

Table 6 shows HTC-cell association and subcellular
distribution of endotoxin mixed with poly L-d-ornithine.

7 HTC-cells were mixed with about

Experimentally, 1.25 x 10
2.5 x 105.cpm (ca 460 ug) of labeled endotoxin mixed with

poly L-d-ornithine. Mixing endotoxin with ornithine signifi-
cantly enhanced endotoxin binding to HTC-cells. After 1 hour
incubation at 37°C, 1399 cpm of endotoxin were cell associated

6 cells and this increased after 3 and 5 hours

with 2.5 x 10
indubation to 1618 and 1850 cpm respectively. After 1 hour
451 cpm of the cell associated counts were recovered from
crude nuclear fraction and after 3 and 5 hours 932 and 1106
cpm of the cell associated counts were bound to crude nuclear
fraction. At the concentration used (4 ug/ml), ornithine
(which did not markedly change the physical structure of
endotoxin, Table 1) increased the cell association and

subcellular distribution of endotoxin at least 10 times

compared to toxic endotoxin.

57

Table 6

HTC-cell association and cell distribution of
endotoxin mixed with poly L-a-ornithine
(92 ug/Sml cell) after 1, 3 and 5 h incubation

at 37°Ca

 

 

 

 

 

 

 

 

Incubation Time in Hour

Experimental 1 3 . 5
40300 42135 42893

Counts to start i i i
With 537 418 617
1399 1618 1850

Whole cell i i f
49 104 129
451 932 1106

, + + +

Nuclei “ ‘ 7
38 67 95
949 T ' 723 917

Non-nuclear f t :
fracthn 37 21 49

 

 

 

 

 

 

a = Mean of cpm 1 standard error of the mean obtained from

at least six experiments

58

 

 

HTC-cell Association and Cell Distribution of 0.25 N NaOH

Treated Endotoxin

 

Alkaline-detoxified endotoxin showed a third-type of '
binding pattern (Table 7). After 1 hour incubation of 1.25 x
107 HTC-cells with about 1.25 x 105 cpm (CA 315 ug) of labeled
0.25 N NaOH treated endotoxin, 507 cpm became associated with
2.5 x 106 cells. Binding did not significantly increase after
1 hour, since only 600 and 594 cpm were observed after 3 and
5 hours respectively. After 1, 3 and 5 hours incubation 172,

205 and 175 cpm of the cell associated counts were recovered

from crude nuclear fraction respectively. The extent of

 

crude nuclear fraction transfer of alkaline—treated endotoxin
was significantly lower than toxic endotoxin at 5 hours and

endotoxin mixed with ornithine at 3 and 5 hours.

 

59

 

Table 7

HTC-cell association and cell distribution of
0.25 N NaOH treated-endotoxin (63 ug/S ml cell)

after 1, 3 and 5 h incubation at 37°Ca

 

 

 

 

 

 

Incubation Time in Hour
Experimental 1 3 h 5
23770. 23420 23567
Counts to start i i i
With ’244 260 384
507 600 594
I + + +
. Whole cell ‘ ~ ‘ ‘
13 24 27
172 205. 175
i i i
Nuclei
5 16 10
384 485 531
Non—nuclear + + i
fraction 6 6 21

 

 

 

 

 

 

a = Mean of cpm 1 standard error of the mean obtained from

at least six experiments

60

DISCUSSION

The pattern of toxic endotoxin distribution among organs
and tissues of normal mice has been reported previously (9,11,
21). To a large extent this distribution follows the distri-
bution of the RES system and the clearance rates from the
circulation are very rapid. Most toxic endotoxin becomes fixed
to tissues or organs by 30-60 minutes (11) and there is little
change in this distribution with time as seen by results up‘
to 5 hours presented in this manuscript and longer (9,21) by
others.

Alkaline detoxification changes not only the chemical
and physical properties of endotoxin but also the rate of
clearance and distribution of endotoxin in the host (6) (see
Table 4). Once the detoxified endotoxin is fixed in tissues
after 1 hour it does not significantly change location up to
5 hours. The bulk of the detoxified endotoxin is found in
the liver, kidney, intestine and carcass. Since the mice
were not perfused free of blood it is difficult to determine
whether the relatively increased carcass counts reflect
increased binding to the host muscle mass or just free
detoxified endotoxin in the blood. It was not surprising
that ornithine, which has dramatic effects on uptake of
endotoxin in HTC-cells (see Table 6) did not alter distribution
In_zIzn. This compound appears to have little effect on the
physical or biological properties of endotoxin per se (see
Table 1) but instead affects permeability of the cells In

vitro. In all likelihood most of these effects in the complex

61

 

In zIzn_environment would be neutralized by normal physio-
logical processes of metabolism and excretion.

The increasing intereSt in studying the biological
actions of endotoxin 12.31332 (8,12,13,21) makes it incumbent
upon the researcher to establish that In_ZIE£n actions approxi-
mate Inerzn realities. Despite experimental modifications
in the endotoxin molecule all of our whole animal studies
reaffirm the notion that the liver is the major organ
responsible for endotoxin clearance from the blood. This

fact alone makes hepatic derived cell lines an excellent

 

candidate in which to evaluate and compare the basic inter-.
actions of endotoxin with the cell In yIEnn and In 2129' As
will be seen below, this comparison is not favorable.

The liver contains a number of cell types. While the
macrophage (Kupffer cell) are thought to be the main reservoir

of In vivo clearance (11) there is also evidence that about

 

75% of endotoxin trapped in the liver is associated with
parenchymal cells (21). What this means is that if 50% of
intravenously injected endotoxin is found in the liver, it
might be assumed that about 35% is associated with parenchymal

cells. Hence, the parenchymal cell binds endotoxin In vivo

 

very rapidly. By contrast, the HTC-cell line used in this

study binds little endotoxin (see Table 5) even after 5 hours.

 

Reasoning that In vivo endotoxin is exposed to plasma before
binding to parenchymal tissue we tried to enhance binding in

culture by using endotoxin exposed to plasma In vivo but

 

again, binding was discouragingly low (see Table 5).

62

Binding of alkaline-treated endotoxin to HTC-cells was
3 to 8 fold greater than toxic endotoxin. This is Opposite
to endotoxin binding In_XIzn. When the HTC-cell membrane was
influenced by ornithine (16) we observed our greatest binding
and subcellular uptake of endotoxin yet ornithine In_!Izn_had
little effect on endotoxin distribution. Furtherevidence of
a lack of correlation between 1E.ZEE£9 tissue culture models
with Insz 9 realities is seen from the data on nuclear
association. With toxic endotoxin nuclear association gradually
increases with time, while In zInn_no such pattern developes.
One consistancy is that nuclear accumulation of detoxified
endotoxin does not seem to increase with time In_zIt£n_or In

Cumulatively, we must conclude that conditions have not
yet been identified by which endotoxin binds to HTC-cells to
the extent it binds to liver In_zIzn. Before meaningful
biological studies on £2.21EEQ toxicity can be performed,
conditions which compare favorably to 22.2122 realities must
be defined.

A preponderance of current literature suggests that endo-
toxin toxicities In zIxn_might be mediated by release of
soluble materials into the circulation (14). Both our labora-
tory (Werner and Moon in preparation) and others (1) have
been able to demonstrate, by addition of plasma from endotoxin
poisoned mice, biological actions on tissue culture cells which
seem to mimic In_zIzn changes. While in the final analysis,
these mediators of endotoxin action might prove to, in fact, be
the key to understanding the pathophysiology of endotoxin

actions in vivo, study on direct toxicities of endotoxin on

63

“—g _._

 

non-RES cells remains a plausable alternative. A prerequisite
to such studies is that In vivo and In_vitro binding should be
similar and, in our opinion we have not yet been able to

achieve these experimental conditions In_vitro.

64

LITERATURE CITED

Berry, L. J., K. J. Goodrum, C. W. Ford, I. G. Resnick
and G. M. Shackleford. 1980. Partial characterization
of glucocorticoid antagonizing factor in hepatoma
cells, pp. 77-81. In Schlessinger (ed), Microbiology.
American Society for Microbiology, Washington, DC.
Blobel G. and V. R. Potter. 1966. Nuclei from rat
liver: Isolation method that combines purity with high

yield. Science 154:1662-1665.

 

Burton, K. 1968. Determination of DNA concentration
with diphenylamine, pp. 163-166. In Grossman, L. and
K. Moldave (ed), Methods in enzymology, vol XII.
Nucleic Acids part B. Academic Press, New York.
Chedid, L., R. C. Skarnes and M. Parant. 1963. Char-

51Cr-labeled endotoxin and its identi-

acterization of
fication in plasma and urine after parenternal
administration. J. Exp. Med. 117:561-571.

Ciznar, I. and J. W. Shands. 1971. Effect of alkaline

treated lipOpolysaccharide on erythrocyte membrane

stability. Infect. Immun. 4:362-367.

 

Golub, S. D., H. M. Grochel and A. Nowotny. 1968.
Factors which affect the reticuloendothelial system
uptake of bacterial endotoxins. J. Reticuloendothel.
Soc. 5:324-339.

Goodman, G. W. and B. M. Sultzer. 1977. Mild alkaline
hydrolysis of lipopolysaccharide endotoxin enhances its
mitogenicity for murine B cells. Infect. Immun.

17:205-214.

10.

11.

12.

13.

14.

Harris, R. A., D. L. Harris and D. E. Green. 1968.
Effect of Bordetella endotoxin upon mitochondrial
respiration and energized processes. Arch. Biochem.
Biophys. 128:219-230.

Howard, J. G., D. Rowley and A. C. Wardlaw. 1958.
Investigations on the mechanisms of stimulation of non-
specific immunity by bacterial lipopolysaccharides.
Immunology 1:181-203.

Johnson, K. J. and P. A. Ward. 1972. The requirement
for serum complement in the detoxification of bacterial
endotoxin. J. Immunol. 108:611-616.

Mathison, J. C. and R. J. Ulevitch. 1979. The
clearance, tissue distribution, and cellular localization
of intravenously injected lipopolysaccharide in rabbits.
J. Immunol. 123:2133-2143.

McGivney, A. and S. G. Bradley. 1979. Effects of
bacterial endotoxin of lysosomal and mitochondrial
enzyme activities of established cell cultures. J.
Reticuloendothel. Soc. 26:307-316.

Mela, L., L. D. Miller, J. F. Diaco and H. J. Sugarman.
1970. Effect of I, nnII endotoxin on mitochondria
energy-linked functions. Surgery 68:541-549.

Moore, R. N., K. J. Goodrum, R. Couch, Jr. and L. J.
Berry. 1978. Factors affecting macrophage function:
glucocorticoid antagonizing factor. J. Reticuloendo-

thel. Soc. 23:321-332.

66

15.

16.

17.

18.

19.

20.

21.

Reed, L. J. and H. Meunch. 1938. A simple method for
estimating fifty percent endpoints. Amer. J. Hyg.
27:493-497.

Ryser, H.J.P. 1968. Uptake of protein by mammalian
cells: An underdeveloped area. Science 159:390—396.
Ryser, H.J.P., T. E. Termini and P. R. Barnes. 1976.
Polynucleotide aggregates enhance the transport of
protein at the surface of cultured mammalian cells. J.
Cell. Physiol. 87:221-228.

Skarnes, R. C. 1970. Host defense against bacterial
endotoxemia: Mechanism in normal animals. J. Exp.
Med. 132:300-316.

Ulevitch, R. J. and A. R. Johnston. 1978. The modifi-
cation of biophysical and endotoxic properties of
bacterial lipopolysaccharides by serum. J. Clin.
Invest. 62:1313—1324.

Wang, C. H., and D. L. Willis. 1965. Radiotracer
methodology in biological science, pp. 25—31. Prentice-
Hall, Inc., Englewood Cliffs, New Jersey.

Zlydaszyk, J. C. and R. J. Moon. 1976. Fate of 51Cr-
labeled 1ipopolysaccharide in tissue culture cells and

livers of normal mice. Infect. Immun. 14:100-105.

67

 

 

APPENDIX I

 

68

 

 

l8 L66
ENDOTOXIN
l6 I62
I4 I58
l2 L54 1
.0
n 5
g \
9.
x IO use >_
I:
z (D
a- E
0 8 L46 o
6 L42
4 L38
2 L34

 

 

 

 

 

o .2 .4 .6 _ .e 1.0
NGRMALIZED FRACTION

FIG. 1. CSCl ULTRACENTRIFUGATION OF UNTREATED ENDOTOXIN

69

 

 

PLASMA TREATED ENDOTOXIN

IO Iso

 

m 2
o 2
" 2°.

6 L42 .,.
x l-
, 9
g ‘d‘
4 I38

2 L34

 

 

 

0.2 0.4 O. 6 0.8 |.O
NORMALIZED FRACTION

FIG 2. CSCl ULTRACENTRIFUGATION OF NORMAL MOUSE PLASMA

TREATED ENDOTOX IN

70

 

 

 

 

.2594 Moon ocroxmso enoomme'
dsl.59
3.0 l:65
L60
"’9.
x 2.0 L55
2
d.
0
l.50
L0 L45
L40
0 .2 .4 .6 .8 IO "35
NORMALIZED FRACTION
FIG 3. CSCl ULTRACENTRIFUGATION OF 0.25 N NaOH TREATED

ENDOTOXIN

71

DENSITY (6/003 )

51

Association and Subcellular Distribution of Cr-sodium

Chromate, 51Cr—sodium Chromate Mixed with Endotoxin, and
51

 

Cr—bound to Endotoxin
51

 

A concern using Cr-LPS is that Cr might become dis-
sociated from the LPS complex. To address this possibility
directly, we incubated free chromium with HTC-cells in the
presence and absence of endotoxin and compared uptake and
subcellular distribution data with that obtained for 51Cr-
LPS complex. The results of this experiment (Table 1)
clearly show significant differences in the cellular as-
sociation and subcellular distribution of the bound vs free
isotope. For example, when NaZSlCrO4 alone or NaZSlCrO4
mixed with endotoxin was incubated for 1 hr at 37°C, 1192
and 998 cpm of the counts respectively were associated with

51

the cells compared to 63 cpm when Cr-labeled endotoxin was

‘used. Cell association of unbound Na 51CrO4 increased with

2
time but nuclear transfer did not. Cell association and
nuclear transfer of 51Cr-labeled endotoxin increased with
51

time. With unbound Na2 CrO the nuclear-cytoplasmic ratios

51

4

are opposite those obtained with Cr-labeled endotoxin.

The significant differences between the cell association and

51 51

cell distribution of Cr-labeled endotoxin and Na2 CrO4,

rules out the possibility that unbound Na251CrO instead of

4
51Cr-labeled endotoxin was taken up by the cells and nuclei.

72

 

 

Hmopm p

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmmOOHmnpoa mam ompp mpmnnpccnpos.om zm

upmapmpomsnpw mpmmmnobn mnoa zm mp

mwcmpmm.

mp

nno».msm zm

N

. mp mp . . .

maonompp mmm00pmnpos mam ompp mpmnnpqcnvos Om zmm ONCE. zm QHOB wamo spa: oSQOHOXH:

.msm zmnmpOHOAIpmcmpmm msdonoxp: ow moon.
. mp

23890 23308 :88 59 rem zcN nHoanannd Hem

.mxanHancHnH H u m H a m H c m
.nocsnm do momma umamm mmmmm mmwmm qumm mpmpm mmmmp hmmmo. mpqom mpamw

H H H H. H H. H H H

spas -

gem mum .mmm ,pwmo pomm pmop. mumm pmoo pomp
SUOpo ompp ppmw puwm Newm mmm puom. wmwm mm pmm pmo

H H . H H H H H H H
pmm mpm mmp poo pmm pmm 0 pm MN
zcopmp .mmm Emu mom Emu eon amp po mm mm

. H H H H H H H H H
pom mm mm mm mm mm p m pm
zoslbcopmmn meo pwom Nwmm mow pmwp Nomp mm mm mm

mnmonpo: H H _ H H H H H H H
_ em HmH New um Hoe Hem c m H

ngmm: om ops H mnmsmmnm mnnon om.nrm ammo ocnmpbmm ”Hos mu pmmmn m mxpmfipamsnmw nopp

nnom Bpxmm spa: msaowoxp: smm

anemlpmcmpom mnmonoxp: on mpp «paw pOHDnm damn ammo

73

Limulus Lysate Test

 

Presence of toxic endotoxin in HTC-cells and nuclei was
further confirmed by the limulus lysate assay. This test
can detect as little as 0.125 nanogram/ml endotoxin (Micro-
biological Associates Lot. No. L09578).

HTC-cells were grown, harvested and resuspended in

10 ml of MEM at a concentration of 5x105 cells per ml.

51Cr-labeled endotoxin (58 ug) was added to the cells and
incubated for 3-4 hr. After incubation the cells were
washed four times with pyrogen-free saline and resuspended
in 1 ml of deionized pyrogen-free distilled water. The
cells were broken, as judged by microscopic examination, by
passing through a 25 gauge needle. A ten-fold serial dilution
series of the broken cells was prepared in pyrogen-free saline.
A sample (0.1 ml) of broken cells was mixed with an equal
volume of limulus lysate and incubated at 37°C. Changes in
the opacity and clotting were recorded on a scale of l to 4
by eye.

For limulus lysate tests on cell nuclei, the nuclei
were prepared by treating HTC-cells with 0.5% Nonidet P-40
for 10 min on ice. Nuclei were harvested by centrifugation
at 800xg for 10 min, washed once with pyrogen-free saline and
resuspended in l.ml pyrogen-free distilled water. Nonidet I
P-40 at the concentration of 0.5% did not clot the limulus
lysate. The nuclei were broken, as judged by microscopic
examination, with a glaSs Dounce homogenizer. Ten-fold I

serial dilutions of nuclei were prepared and the limulus

lysate test was performed as described above.

74

 

TABLE 2

Limulus lysate test on HTC whole cell and nuclear

lysate incubated with endotoxin for 3 hr

 

us 0F LPS 0R DILUTION OF CELLS_0R-NUCLEI,

 

 

EXPERIMENTAL 10'1. * 10'2 10’3 .10‘4 .1055
Toxrc ENDOTOXIN 4+ 4+ 4+ 2+ -
WHOLE CELL LYSATE '4+ _ 3+ 1+ - -
NUCLEI LYSATE 4+ - — — _

 

 

 

 

 

 

\J

U]

Normal HTC-cells and nuclei were processed to serve as
negative controls. Positive controls were ten-fold dilutions
of normal endotoxin ranging from 1 ug to 1x10-6 pg per ml.

Table 2 shows the results of the limulus lysate test on
whole cells and nuclear lysates. Cell lysate from cell
incubated with endotoxin was able to initiate the clotting

reaction at 10-3

1

dilution. A strong clotting reaction was

visible at 10- dilution of the nuclear lysate. Neither

normal HTC-cell lysate nor nuclear lysate was alone able to

clot the limulus lysate.

51

Cell Association and Cell Distribution of Cr-labeled'

 

Endotoxin by Hepatoma Tissue Culture Cells at 4°C

HTC-cells were grown in MEM supplemented with 5% heat
inactivated fetal calf serum. The cells were harvested and
suspended in 25 ml MEM at the presence of toxic endotoxin
with constant stirring at 4°C. At 1, 3 and 5 hr intervals,
5 ml aliquots were removed and cell association and cell
distribution of endotoxin was detected as stated before

51Cr-labeled

(first paper). The accumulation rate of
endotoxin at 4°C was significantly lower than 37°C. After

1 h only 23 cpm of the initial counts became cell associated
(Table 3). At 3 hr and up to 24 hr cell association of
endotoxin at 4°C was significantly lower than cell association
of endotoxin at 37°C. After 1 and 3 hr incubation negligable
numbers of the cell-associated counts were recovered from

tcrude nuclear fraction. Nuclear transfer of endotoxin

increased slightly so that by 24 hr of incubation there were

76

Table 3

 

 

 

 

 

 

 

 

 

 

 

HTC-cell association and cell-distribution of 5J'Cr-labeled
toxic endotoxin after 1, 3, 5 and 24 h incubation at 4°C.
Counts per minuteab
Experimental 1 3' 5 24
Counts to start 51593 51459 51249 47962
. f + + i
With ‘ ‘
904 426 672 843
23 58 69 91
Whole cell i i i i
5 6 7 6
Crude nuclear 13 18 28 40
f f i f
fraction 2 2 7 4
Non-nuclear 21 38 36 34
fraction f f i i
2 8 3 2
a=Mean of cpm f standard error of the mean obtained from at

least 6 experiments.

3b=Background was between 50-60 cpm.

77

 

 

 

 

an average of 40 cpm above background associated with

the crude nuclear fraction. It is not clear whether lower
cell association of labeled endotoxin at 4°C was due to lower
binding of endotoxin to the cells or because of a decrease in
the cell internalization of labeled endotoxin. This obser—
vation is consistent with Ryser's report (26, first paper)
.that monolayers of Sarcoma SlBOII adsorb less albumin-I131
at 2°C than at 37°C.

Organ and Tissue Distribution of 0.25 N NaOH treated 51Cr-

 

Labeled Endotoxin in Mice

 

Mice were injected intravenously with 0.18 mg 51Cr-

labeled 0.25 N NaOH treated endotoxin. After 1, 3 and 5 hr,
mice were killed by cervical dislocation, and blood_was
collected by perfusion of 20 ml saline through the apex of

the heart. .The results of this experiment indicated that,

in contrast to toxic endotoxin, only a small fraction of

0.25 N NaOH treated endotoxin was associated with the liver

and spleen. Most of the counts were in the blood and carcass.
With time less radiolabeled 0.25 N NaOH treated endotoxin was
found in the blood and more recovered from liver and intestine.
Association of detoxified endotoxin with liver and spleen was
significantly lower than for toxic endotoxin. This observation
clearly demonstrates that detoxified endotoxin behaves dif-

ferently from toxic endotoxin In vivo.

 

78

 

onmms mam npmmco upmfifipqcfipo: om Apmo any

P5

SHOO .

E

mp

nHupmqmpmm o.mm 2 zoom nfimmnmm osmonoxps

 

rp<oH ZCOpop own,

Doubnm won Spandm

 

 

 

 

mfitcm omfip
tCHo cacao bp<on mppoos. mudbmw panownpso mound Essa wflmp: ownommm moon mpmaomn mpooa
w N~p pmpm omen mOp amp Nmbo ppp mom mp pmumm mmwm pm pmomw
p SH ppu pwwm omow Hum who nmpu Mpu Nmp p pwmbm mwow po Nommw
MN». pmom popum mmu mmm womb. pmm mmm m mmmpm mwmm pm pumom
w 3H6 pmm pump ppmmm Hum pep» uubm pmm . Nmo mp Nommu ammo pm momm
wmp pmbm pompm mm» mp» . bwmm pow . mom no pwmum ubup mm pwmmm
m can He» Heed SHE. 8H Sm £8 H2. at. c H83 2.8 c 2.3
mm» pobm pombm pmo amp bowu ppu pomm o. pumum moom o ummm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ounce On «Snow

c.0uopa 0m fizo

mopmfiwwm mxponpaonnm.

mmpmnmnm oxpmnpambnm.

79

ABSTRACT

In vitro Interactions of Endotoxin, Chromatin, DNA, and

 

Steroid)Hormone-receptors

 

The ability of toxic and alkaline-detoxified endo-
toxin to bind directly to chromatin and DNA and to inhibit
glucocorticoid—receptor binding to these materials has been
characterized. Toxic endotoxin does not bind to DNA in
the form of DNA cellulose, but was capable of binding to
HTC-cell chromatin (l ug/20 ug DNA) four times more efficiently
than alkaline-treated detoxified endotoxin. Interaction
of glucocorticoid-receptor complexes with chrOmatin and DNA
was inhibited in a dose-dependent manner by toxic endotoxin.
Alkaline-treated detoxified endotoxin was even more effective.
It appears most likely that endotoxin interacts directly
with activated hormone-receptor complex, blocking its ability
to bind to the DNA and chromatin. The biological significance
of these observations is clouded by the need for high con-
centrations of endotoxin and the observation that biologically
detoxified endotoxin in some instances works even better than

toxic endotoxin.

80

INTRODUCTION
Glucocorticoid hormones stimulate a wide variety of
biological responses in animal tissues including protection
against the lethal effects of bacterial endotoxin (7,8). A
large body of circumstantial evidence suggests that all
steroid hormones act via basically the same mechanism. The
hormones readily pass into the cytoplasm of cells (3,23) and

bind to specific receptors (4,33). After a possible "activation"

step (11,18,19,31,32) receptor-steroid complexes become con-
centrated in the nucleus (11,18,3l,32). The extent to which
the receptor—steroid complexes bind to chromatin (5,25,31,37)

or DNA (2,33,37,41) has not been definitively established.

Whatever the precise mechanism(s), receptor—steroid complexes
are thought to alter the transcription process resulting in
increased levels of specific messenger RNAs (31,37). The
products of translation of these messenger RNAs leads to the
observable biological effects (25,31,37), one of which is the
induction of selected hepatic enzymes (15,16,20,21,24,27,29,
30,34,38,39,40). Bacterial endotoxin inhibits the induction
of some, but not all of these adaptive enzymes (8,9). One
possible molecular mechanism to explain endotoxin's ability
to inhibit enzyme induction is that the lipopolysaccharide
(LPS) directly interferes with glucocorticoid-receptor binding
to cells or subcellular targets. Experiments described
herein were designed to test the hypothesis that endotoxin
can block steroid binding to hepatic cell chromatin and/or

DNA. This hypothesis seems plausible since a previous report

81

 

from this laboratory (42) has shown that endotoxin associates
directly with hepatic parenchymal cells and their nuclei In
EIXQ. To evaluate the biological significance of the
observed In vitro interactions, most experiments described
herein compare particular variables using both toxic and

alkaline detoxified endotoxins.

82

 

 

METHODS AND MATERIALS
All procedures were performed at 0-4°C unless otherwise
specified.
Animals
Female Spartan HA-ICR mice were used for enzyme studies.
Female Charles river CD-1 mice were used for organ distribution
of endotoxin. Food and water were available 3n libitum.

Chemicals

 

Minimum essential medium (MEM) and fetal calf serum (FCS)
were obtained from GIBCO (Grand Island, NY).

[1,2,3H] Dexamethasone (20-40 Ci/mmol) was purchased from
New England Nuclear (Boston, MA) or Amersham/Searle (Arlington
Heights, IL). Micrococcal nuclease, calf thymus deoxyribo-
nucleic acid and hydrocortisone-Zl—sodium succinate was

purchased from Sigma Chemical Company (St. Louis, MO).

Phosphoenolpyruvate Carboxykinase Assay.

 

Phosphoenolpyruvate carboxykinase (PEPCK) activity was
assayed by the method of Lane et al., (22). PEPCK was
induced by injecting 5 mg of hydrocortisone-Zl-sodium succinate
suspended in 0.5 ml of sterile saline subcutaneously into
the interscapular region. Toxic or alkaline-treated de-
toxified endotoxin (50 ug/mouse), dissolved in saline, was
injected intraperitoneally (i.p.) immediately after hydro-
cortisone. PEPCK activity was determined in liver homo-
genates 4 hr later.
Preparation of Radiolabeled Endotoxin and Pure Nuclei

51

Cr-labeled endotoxin was prepared according to the

method of Chedid et al., (14). Blobel and Potter's (10)

83

 

method was used to prepare nuclei free of debrie by sucrose
gradient centrifugation.

Preparation of HTC-cell Cytosol and Chromatin

 

Hepatoma tissue culture (HTC) cells were grown at 37°C
in suspension cultures in MEM supplemented with 5% heat in-
activated fetal calf serum. Cells were harvested at 600xg for
5 min, washed at least twice with an ice cold buffer [0.04 M
Tris (hydroxymethyl) aminomethane—hydrochloride, 0.1 M KCl,

0.001 M MgCl pH 7.2], suspended in one volume homogenization

2;
buffer [0.02 M Tris (hydroxymethyl) methylglycine, 0.002 M
10.4 dithiothreitol, pH 7.4] and homo-

CaCl 0.001 M MgCl

2’ 2'
genized with a Dounce homogenizer. Homogenates were centri—
fuged at 800xg for 10 min and the supernatant solution (non—
nuclear fraction) was separated from the pellet containing
nuclei which was reserved and used as described below. The
non-nuclear fraction was further centrifuged at 35,000xg for

8 M) was added to the supernatant

15 min. Dexamethasone (5x10—
and the mixture centrifuged at 105,000xg for 90 min. The
supernatant (cytosol) was then incubated for at least 2 hr to
allow steroid binding, followed by activation of the steroid
receptors by supplementing the cytOsol with 0.15 M NaCl and
incubating at 20-25°C for 30 min (18). Activated cytosol

was finally passed through 20 ml of Sephadex G—25 equilibrated
with homogenizing buffer containing 0.25 M sucrose to remove
unbound steroid and salt. The macromolecular fraction con—
taining activated hormone-receptor-complexes (AHRC) was used
immediately. AHRC had about 5 mg/ml protein and was inac—

tivated (IHRC) by incubation at 37°C for 1 hr.

84

 

Nuclear fraction was waShed once in 10 vol of 0.01 M
Tris, pH 8.0 at 0°C and allowed to swell for 2 hr in 3.5 vol
of the same buffer. Nuclei were broken by sonication and
centrifuged at 20,000xg for 15 min. The supernatant contained
chromatin and had about 1 mg/ml DNA as measured by the
diphenylamine technique (12).

51

Binding of Cr—labeled Toxic and Alkaline-treated Detoxified

 

Endotoxin to Chromatin

 

Alkaline—treated endotoxin was prepared by dissolving
50 mg of endotoxin in 3 ml of 0.25 N NaOH. The solution
was heated at 56°C for 60 min (17). The modified endotoxin
was precipitated with cold anhydrous ethanol, washed three
times with ethanol, dialyzed against deionized distilled
water and then lyophilized. Both toxic and detoxified endo—

51

toxin were labeled with Na CrO according to the method

2 4

of Chedid et al., (14).
An aliquot (20 ul) of chromatin was incubated for

51Cr-labeled toxic or alkaline-treated

90 min with 10 ul of
detoxified endotoxin. Chromatin precipitation was achieved

by adding 2 ml of 5 mM MgC12/0.01 M Tris buffer to the assay
mixture. After at least 10 min, the mixture was centrifuged
at 8,000xg for 10 min. As a control, 10 ul of 51Cr-labeled
toxic or alkaline—treated detoxified endotoxin was mixed with
20 ul homogenizing buffer and precipitated with MgCl2 solution.
The cpm of endotoxin precipitated by MgCl2 in the control

tubes was considered background and subtracted from the test.

85

 

Binding_of Hormone-receptor to Chromatin in the Presence of

Toxic and Detoxified endotoxin

 

Different concentrations of toxic or alkaline—treated
detoxified endotoxin were mixed with 10 pl of chromatin.
After making the volumes equal with homogenizing buffer, the
endotoxin-chromatin mixtures or chromatin alone were incubated
for 2.5 hr. To each tube 0.4 ml of the macromolecular fraction
of activated cytosol was added. After 90 min incubation,
0.2 ml of the assay mixture was added to 1.8 m1 of 5 mM MgC12/
0.01 M Tris buffer. The tubes were mixed on a vortex mixer
and after at least 10 min, centrifuged at 8,000xg for 10 min.
The pellet was digested in.0.4 m1 of 1 M NaC1/0.5 M NaOH over-
night at room temperature and counted in 3a70B coctail (rpi).
As a Control, 0.2 m1 of activated cytosol was added to 1.8 ml

of MgCl solution and processed as stated. Counts per minute

2

of activated cytosol that were precipitated with MgCl were

2
subtracted from the test. The percent inhibition of binding
of hormone-receptor to chromatin in the presence of endotoxin
was calculated as follows: loo-[cpm of hormone-receptor
bound to chromatin in the presence of endotoxin/cpm of.

hormone-receptor bound to chromatin]x100.

Binding of Endotoxin to DNA

 

DNA-cellulose (340 pg DNA/ml) was prepared according to
the technique of Albert et al., (1). An aliquot of 0.1 ml
or 0.3 ml DNA-cellulose or cellulose alone was centrifuged
in 1.5 m1 disposable conical tubes (Eppendorf) at 600xg for

5 min. The supernatant was removed by aspiration. To each

86

tube, 20 pl of labeled endotoxin was added and the total
volume was made up to 0.6 ml with Tris/glycerol buffer
(0.01 M Tris/10'4 M dithiothreitol/10% glycerol v/v). The
tubes were capped, placed on a roller bottle apparatus at
10 rpm for 2.5 hr and turned 180° every 30 min during the
assay to help keep the cellulose resuspended. The reaction
was stopped by centrifugation at 600xg for 5 min._ The
supernatant was removed and the 51Cr radioactivity was
measured. The pellets were resuspended in Tris/glycerol
buffer and counted. 1

Binding of Hormone—receptor to DNA in the Presence of Toxic

 

Endotoxin

Aliquots of 0.1 ml of DNA-cellulose (34 pg) or cellulose
alone were centrifuged in 1.5 ml disposable conical tubes
at 600xg for 5 min. The supernatant was discarded and the
pellet was suspended in 0.6 ml of Tris/glycerol buffer.
Different amounts of toxic endotoxin (100 or 300 pg) or
Tris/glycerol buffer were added to DNA-cellulose and left
on a roller bottle (10 rpm) for 1.5 hr and turned 180° every
30 min to keep the DNA-cellulose in suspension. After
incubation, some of the tubes which contained 100 pg of
toxic endotoxin were centrifuged, the supernatant discarded,
and the pellet resuspended in 0.6 ml of Tris/glycerol buffer.
The binding of AHRC to DNA in the presence or absence of
excess toxic endotoxin was determined by raising the total

volume of all the tubes to 0.8 ml by adding 0.2 ml of AHRC.

87

 

 

The tubes were capped and left on a roller bottle for
2.5 hr. The reactions were stopped by centrifugation at
600xg for 5 min. The pellets were resuspended in 0.2 ml
of Tris/glycerol buffer for at least 1 hr at room temperature
before being counted. The percent of inhibition of binding
of AHRC to DNA in the presence of toxic endotoxin was
calculated as follows: loo-(cpm of AHRC bound to DNA cel-
lulose which was incubated with endotoxin-cpm of AHRC bound
to cellulose)/(cpm of AHRC bound to DNA cellulose-cpm of

AHRC bound to cellulose)x100.

88

 

RESULTS
Cortisone Induction of PEPCK in Normal and Endotoxin-

Poisoned Mice

 

'The effect of toxic and detoxified endotoxin on the In
KIXQ induction of PEPCK by cortisone is presented in Table l.
Cortisone administration significantly increased the level
of PEPCK in the liver 4 hr after injection and toxic endo-
toxin inhibited this induction. When endotoxin was detoxified
by treating with 0.25 N NaOH, the ability of endotoxin to
inhibit the induction of PEPCK in whole animals was prevented.

Hepatic Association and Subcellular Distribution of Toxic

 

and Detoxified Endotoxin

 

Table 2 shows the extent of binding of intravenously
(i.v.) injected toxic and detoxified endotoxin to liver and
hepatic nuclei. Experimentally, 200 pg (CA 120,000 cpm)

51Cr-labeled toxic endotoxin were injected into mice.

of
After 5 hr about 50% (65643 cpm) were associated with the
liver. Of this number 7462 cpm could be recovered from
washed crude nuclei, and 366 cpm from pure nuclei. Five
hours after injection of 180 pg (45,000 cpm) of detoxified
endotoxin, hepatic association was significantly lower than
with toxic endotoxin. About 50% of detoxified endotoxin
remained in the-circulation and carcass (data not shown) on
a percentage basis. No significant difference in hepatic

nuclear association of toxic and 0.25 N NaOH-treated endo-

toxin was observed.

89

Table l. Hydrocortisone induction of PEPCK in the

presence and absence of toxic and detoxified

 

 

endotoxin.

Experimental PEPCK Activity
Normal 19.7 i 1.4a
Cortisone 34.2 i 3.7b
Toxic endotoxin + cortisone 20.8 i 2
Detoxified endotoxin + cortisone 34.9 i 3

 

=Mean 1 standard error of the mean obtained from
6 separate experimental determinations.
b=Significant1y different from toxic endotoxin +

”cortisone (P < 0.05).

90

 

Table 2. Hepatic association and subcellular distribution
of toxic and detoxified endotoxin 5.hr after i.v.'

injection into mice.

 

 

cpm of toxic “cpm of detoxified
Experimental endotoxina endotoxinb
Liver homogenate 65643 i 4787C 10878 1 674a
Crude nuclear fraction 7462 i 775 1140 r 50
Purified nuclei 366 i 27 139 i 22

 

a=200 pg (120,000 cpm)/mouse intravenously.

b=180 pg (45,000 cpm)/mouse intravenously.

A c=Mean 1 standard error of the mean obtained from 6 separate-
experimental determinations.

d=Mean r standard error of the mean obtained from 4

separate experimental determinations.

91

Ability of Toxic and Detoxified Endotoxin to Bind to

 

Chromatin

 

51Cr—labeled toxic and

Table 3 shows the ability of
detoxified endotoxin to bind to sonicated-chromatin. When
20 p1 of chromatin (ca 20 pg DNA) was incubated at 0°C
with 51.4 pg of toxic endotoxin, 1.9% of the counts were
bound after 90 minutes. Detoxification of endotoxin lowered

chromatin binding approximately four fold.

Effect of Graded Quantities of Toxic and Detoxified Endotoxin

 

on Binding_of AHRC to Chromatin

 

Graded quantities of unlabeled toxic endotoxin could
inhibit binding of AHRC to a constant amount of sonicated-

chromatin (10 pl). In all experiments sonicated-chromatin

was used. No significant binding of AHRC to nuclease-extracted

chromatin (28) was observed (data not shown).

Table 4 shows that, as the concentration of toxic unlabeled

endotoxin increased from 5 to 500 pg, the percent inhibition
of binding of AHRC to chromatin increased up to a maximum of
slightly higher than 50% with 500 pg of endotoxin. There

was about 345 cpm of AHRC bound per 10 pg of DNA in chromatin.
When AHRC was inactivated at 37°C for 1 hr (IHRC) no binding
of IHRC to chromatin was observed. By contrast, alkaline—
treated endotoxin exerted more of an inhibitory effect on

the binding of AHRC than the toxic endotoxin. In the presence
of 500 pg of alkaline-treated endotoxin, there was essentially
100% inhibition in the binding of AHRC to chromatin (Table

4') .

92

 

 

Table 3. Binding of 51Cr-labeled toxic and detoxified

endotoxin to chromatin.

 

 

Experimental pg added pg bound % binding
Toxic endotoxin 51.4 1 1.9
Detoxified endotoxin 46 0.24 0.52

 

 

 

93

 

 

 

amcpo m. purpcpnpo: 0m pawn cpoopbm no ocnoamnps p5 nsm pnmmoooo om noxpo
mpg monoxpmpma msoonopr.
ops 0m mwo 60559
to on msmonoxpo no ocHoSmnps.
anon oH mcoonoxHa c m we Hoe woo woo woo Henna Hmwno mecwo
o oo mH.~ coo aoo ado

eoxpo osmonopr o o.p q.m po.m pm.m

oonoxpmpmm mbmonopr o mw.m . mm.m mm I

mu.» poo mop woo mmm

 

II

arm mo

UHmwO

monp<mnmm conaosmlnmompnon ooappox.

psmonp<mnom conaosmleoopnon ooappmx.

owpmbw n 300p
Q

N + PEWO.

wmnoosn HSSpcpnpo: 0m cpsdpbm ooapmnmm no mmwo cpsmpsa no nanoamnps.

94

 

 

These results do not establish whether the inhibition
of binding is a result of direct interaction of endotoxin
with chromatin (cf Table 3) or whether free endotoxin re—
maining in solution interacts directly in AHRC discouraging
binding.

Binding of Toxic Endotoxin to DNA and its Effect on the

 

Bindirg of AHRC to DNA

 

Two different concentrations of DNA (34 pg and 100 pg)
were used to investigate whether or not toxic endotoxin was
able to bind directly to DNA conjugated to cellulose (DNA-
cellulose). Data in Table 5 shows that toxic endotoxin does
not bind directly to DNA. When toxic endotoxin (88 pg) was

added to either 0.1 or 0.3 ml of cellulose alone 13,605 and

33,531 cpm of sedimentable label was observed respectively.
When DNA was conjugated to the cellulose there was actually 7
less binding than to cellulose alone, suggesting that endotoxin

does not bind to DNA.

Table 6 shows the effect of two different concentrations
of endotoxin on the binding of AHRC to 34 pg DNA in the form
of DNA-cellulose. There was about 9.5% inhibition in the
binding in the presence of 100 pg of endotoxin and 14.6%
inhibition in the presence of 300 pg of endotoxin. When
DNA-cellulose was centrifuged after incubation with 100 pg
endotoxin and resuspended in Tris/glycerol buffer, there

was no inhibition of AHRC binding (line 5).

95'

 

Table 5. Binding of 51Cr-labeled endotoxin to cellulose

or DNA-cellulose.

 

 

Experimental cpm bound
Cellulose (0.1 m1) 13605
Cellulose (0.3 ml) 33531
34 pg DNA-cellulose (0.1 m1) 10358
100 pg DNA—cellulose (0.3 ml) . 26516

 

 

 

 

96

Table 6. Binding of AHRC to DNA-cellulose in the presence

and absence of endotoxin.

 

 

Experimental cpm of AHRC % Inhibition
bound
Cellulose 730 -
DNA-cellulose 3236 -

DNA-cellulose + 100 pg

endotoxin 2999 9.5
DNA-cellulose + 300 pg

endotoxin 2871 14.6
Endotoxin exposed

DNA-cellulose 3300 -

 

97

 

DISCUSSION

Despite the extensive effort devoted to understanding
the pathophysiology of bacterial endotoxin, so far there are
no conclusive data as to whether endotoxin exerts its toxic
effects directly on cells or through release of soluble
mediator(s). In the present study we have explored an In
yIt£n_model which could conceivably yield insight into the
nature of metabolic changes observed in endotoxemia and
the protective mechanism of cortisone against endotoxin
lethality. In this study we have used a line of gluco-
corticoid-responsive, rat hepatoma (HTC)-cells. This cell
line is particularly well suited for such a study because
at least five steroid—induced responses have been defined
(36) in these cells.

Since glucocorticoids are believed to increase enzyme
synthesis by elevating the production of specific messenger
RNAs (26,31) it might be postulated that endotoxin suppresses
the hormonal induction of hepatic enzymes through an effect
on RNA synthesis. However, when a sublethal dose of toxic
endotoxin is given to adrenalectomized mice total net RNA
and protein synthesis in the liver increases (35). If
endotoxin acts by suppressing RNA synthesis there must be
a high degree of selectivity in the action of endotoxin.

If hepatic parenchymal cells are among the direct targets
of endotoxin action, endotoxin might be able to inhibit the

hormonal induction of hepatic enzymes by competing with the

98

 

 

hormone for possible binding sites on or within the target

cells. 'Our results tentatively suggest this possibility in
that endotoxin can interfere with AHRC binding to chromatin
1.923239.

The whole cell uptake of 3H—dexamethasone is not
altered in the presence of 350 pg endotoxin (results not
presented). It seems unlikely that endotoxin interferes
with the cell uptake of the hormone. The fact that endo-
toxin binds to isolated chromatin In nIn£n_(Table 3) suggests
that the association of endotoxin to chromatin may block the
ability of hormone-receptors to bind to chromatin. While
this tends to support the idea that endotoxinattaches to
nuclear chromatin and inhibits AHRC binding, two lines of
data suggest caution must be exercised in drawing this con-
clusion. First, endotoxin seems to bind to chromatin by
interaction with proteins rather than DNA. This is suggested
indirectly by the observation that endotoxin cannot bind to
calf thymus DNA associated with cellulose (Table 5).
Secondly, AHRC can bind to DNA (Table 6) and sonicated
chromatin (Table 4) but does not bind to deoxyribonuclease-
treated nuclei (2) or nuclease-extracted chromatin which
might have lost between 25 to 50% of its DNA (28). If the
AHRC is binding to DNA and the endotoxin is binding to.
protein a direct inhibition would not seem likely. ,Instead,
some stereochemical alteration would have to be involved as

an explanation. Even though our data do not offer

conclusive evidence, what would appear most likely is that

99

 

 

 

the endotoxin interacts directly with the AHRC, blocking its
ability to bind to DNA. This explanation is most con-
sistant with our data since the presence of endotoxin in
solution inhibits, in a dose-dependent manner, the binding
of AHRC to both chromatin (Table 4) and DNA-cellulose

(Table 6). It is possible that endotoxin acts like the
inhibitory material which has been detected in the cytosol
of HTC—cells and rat uterus (13,37). This inhibitor inter-
acts with the receptor-steroid complex (37) and prevents its
binding to the acceptor sites in the nuclei.

The biological significance of the ability of endotoxin
to block AHRC binding is further complicated by the need for
relatively high concentrations of toxic endotoxin and by the.
observation that biologically detoxified endotoxin inhibits
binding even more efficiently than toxic endotoxin (Table
4). These observations and the fact that bacterial endo-
toxin cannot inhibit the induction of the hepatic enzyme
phosphoenolpyruvate carboxykinase by cortisol In yIE£g_(6)
probably indicate that endogenous factors (8,9) or a host-
modified form of endotoxin is more likely to be involved in
the inhibition of PEPCK induction than is a direct toxicity

of endotoxin on hepatic parenchymal cells.

100

 

 

LITERATURE C ITED
Alberts, B. and G. Herrick. 1971. DNA—cellulose
chromatography. Meth. Enzymol. 21:198-217.
Baxter, J. D., G. G. Rousseau, M. C. Benson,
R. L. Garcea, J. Ito and G. M. Tomkins. 1972. Role
of DNA and specific cytoplasmic receptors in gluco-
corticoid action. Proc. Natl. Acad. Sci. USA 69:
1892-1896.
Baxter, J. D. and G. M. Tomkins. 1970. The relation-
ship between glucocorticoid binding and tyrosine amino
transferase induction in hepatoma tissue culture cells.
Proc. Natl. Acad. Sci. USA 65:709-715.
Baxter, J. D. and G. M. Tomkins. 1971. Specific
cytoplasmic glucocorticoid hormone receptors in hepatoma
tissue culture cells. Proc. Natl. Acad. Sci. USA 68:
932-937.
Beato, M., M. Kalimi, M. Konstam and P. Feigelson.
1973. Interaction of glucocorticoids with rat liver
nuclei. II. Studies on the nature of the cytosol
transfer factor and the nuclear acceptor site. Biochem.
12:3372-3379.
Berry, L. J., K. J. Goodrum, C. W. Ford, I. G. Resnick
and G. M. Shackleford. 1980. Partial characterization
of glucocorticoid antagonizing factor in hepatoma
cells, pp. 77-81. In Schlessinger (ed), Microbiology.

American Society for Microbiology, Washington, DC.

101

 

 

10.

11.

12.

13.

Berry, L. J. and D. S. Smythe. 1964. Effects of
bacterial endotoxins on metabolism. VII. Enzyme
induction and cortisone protection. J. Exp. Med.
120:721—732. I

Berry, L. J., D. S. Smythe and L. S. Colwell. 1966.
Inhibition of inducible liver enzymes byendotoxin and
actinomycin D. J. Bacteriol. 92:107-115.

Berry, L. G., D. S. Smythe and L. S. Colwell. 1968.
Inhibition of hepatic enzyme induction as a sensitive
assay for endotoxin. J. Bacteriol. 96:1191-1199.
Blobel, G. and V. R. Potter. 1966. Nuclei from rat
liver: Isolation method that combines purity with high
yield. Science 154:1662-1665.

Buller, R. E., D. O. Toft, W. T. Schrader and B. W.
O'Malley. 1975. Progesterone—binding components of
chick oviduct. VIII. Receptor activation and hormone-
dependent binding to purified nuclei. J. Biol. Chem.
250:801-808.

Burton, K. 1968. Determination of DNA concentration
with diphenylamine, pp. 163-166. In_Grossman, L. and
K. Moldave (ed), Methods in enzymology, vol XII.
Nucleic Acids part B. Academic Press, New York.
Chamness, G. C., A. W. Jennings and W. L. McGuire.
1974. Estrogen receptor binding to isolated nuclei.

A nonsaturable process. Biochem. 13:327-331.

102

 

14.

15.

16.

17.

18.

19.

20.

Chedid, L. R. C- Skarnes and M. Parant. 1963. Char-
acterization of 51Cr-labeled endotoxin and its identi—
fication in plasma and urine after parenteral adminis-
tration. J. Exp. Med. 117:561-571.

Feigelson, P., T. Dashman and F. Margolis. 1959. The
half-life time of induced tryptophan peroxidase In_zIzg.
Arch. Biochem. Biophys. 85:478-482.

Feigelson, P. and O. Greengard. 1962. Regulation of
liver tryptophan pyrrolase activity. J. Biol. Chem.
237:1908-1913.

Goodman, G. W. and B. M. Sultzer. 1977. Mild alkaline
hydrolysis of lipopolysaccharide endotoxin enhances its
mitogenicity for murine B cells. Infect. Immun. 17:205-
214.

Higgins, S. J., G. G. Rousseau, J. D. Baxter and

G. M. Tomkins. 1973. Early events in glucocorticoid:
Activation of steroid receptor and its subsequent
specific nuclear binding studied in cell free system.
J. Biol. Chem. 248:5866—5872.

Kalimi, M., P. Colman and P. Feigelson. 1975. The
"activated" hepatic glucocorticoid-receptor complex:
Its generation and properties. J. Biol. Chem. 250:
1080-1086.

Kenney, F. T. and R. M. Flora. 1961. Induction of
tyrosine d-ketoglutarate transaminase in rat liver.

I. Hormonal nature. J. Biol. Chem. 236:2699-2702.

103

 

 

21.

22.

23.

24.

25.

26.

27.

28.

Knox, W. E. and V. H. Auerbach. 1955. The hormonal
control of tryptophan peroxidase in the rat. J. Biol.
Chem. 214:307-313.

Lane, M. D., H. C. Chang and R. S. Miller. 1969.
Phosphoenolpyruvate carboxykinase from pig liver mito-
chondria. Meth. Enzymol. 13:270-277.

Levinson, B. B., J. D. Baxter, G. G. Rousseau and G. M.
Tomkins. 1972. Cellular site of glucocorticoid-
receptor complex formation. Science 175:189-192.
Lin, E.C.C. and W. E. Knox. 1957. Adaptation of rat
liver tyrosine-d-ketoglutarate transaminase. Biochim.
Biophys. Acta 26:85-88.

Marver, D., D. Goodman and I. S. Edelman. 1972.
Relationships between renal cytoplasmic and nuclear
aldosterone-receptors. Kidney Int. 1:210—223.

Mohla, S., E. R. DeSombre and E. V. Jensen. 1972.
Tissue specific stimulation of RNA synthesis by trans-
formed extradiol-receptor complex. Biochem. Biophys.
Res. Commun. 46:661-667.

Nickol, J. M., K. L. Lee and F. T. Kenney. 1978.
Changes in hepatic levels of tyrosine aminotransferase
messenger RNA during induction by hydrocortisone. J.
Biol. Chem. 253:4009-4015.

Noll, M., J. 0. Thomas and R. D. Kornberg. 1975.
Preparation of native chromatin and damage caused by

shearing. Science 187:1203-1206.

104

 

29.

30.

31.

32.

33.

34.

35.

Nordlie, R. C. and H. A. Lardy. 1963. Mammalian liver
phosphoenolpyruvate carboxykinase activities. J. Biol.
Chem. 238:2259-2263.

Nordlie, R. C., F. E. Varrichio and D. E. Holton. 1965.
Effects of altered hormonal states and fasting on rat
liver mitochondrial phosphoenolpyruvate carboxykinase
levels. Biochem. Biophys. Acta 79:214—221.

O'Malley, B. W. and A. R. Means. 1974. Female steroid
hormones and target cell nuclei. The effects of steroid
hormones on target cell nuclei are of major importance
in the induction of new cell function. Science 183:
610-620.

Rousseau, G. G., J. D. Baxter, S. J. Higgins and G. M.
Tomkins. 1973. Steroid—induced nuclear binding of
glucocorticoid receptors in intact hepatoma cells. J.
Molec. Biol. 79:539-554.

Rousseau, G. G., J. D. Baxter and G. M. Tomkins. 1972.
Glucocorticoids receptors: Relations between steroid
binding and biological effects. J. Molec. Biol.
67:99-115.

Shrago, E., H. A. Lardy, R. C. Nordlie and D. 0.
Foster. 1963. Metabolic and hormonal control of
phosphoenol pyruvate carboxykinase and malic enzyme

in rat liver. J. Biol. Chem. 238:3188-3192.

Shtasel, T. F. and L. J. Berry. 1969. Effect of
endotoxin and cortisone on synthesis of ribonucleic

acid and protein in livers of mice. J. Bacteriol.

97:1018—1025.

 

 

36.

37.

38.

39.

40.

41.

Simons, S. S., Jr., J. D. Baxter, R. L. Garcea and

G. M. Tomkins. 1975. Role of glucocorticoid receptors
in steroid-induced protein synthesis, pp. 37-45. In
Slavkin H. C., and R. C. Greulich (eds.), extracellular
matrix influences on gene expression. Academic Press,
New York.

Simons, S. 8., Jr., H. M. Martiney, R. L. Garcea,

J. D. Baxter and G. M. Tomkins. 1976. Interaction

of glucocorticoid receptor-steroid complexes with

acceptor sites. J. Biol. Chem. 251:334-343.

 

Thompson, E. B., G. M. Tomkins and J. F. Curran. 1966.
Induction of tyrosine d-ketoglutarate transaminase by
steroid hormones in a newly established tissue culture
cell line. Proc. Natl. Acad. Sci. USA 56:296-303.
Weber, G. 1963. Study and evaluation of regulation of
enzyme activity and synthesis in mammalian liver. Adv.
Enzymol. Reg. 1:1-39.

Weber, G., S. K. Srivastova and R. L. Singhal. 1965.
Role of enzymes in homeostasis. VII. Early effects of
corticosteroid hormones on hepatic glyconeogenic
enzymes, ribonucleic acid metabolism and amino acid
level. J. Biol. Chem. 240:750-756.

Yamamoto, K. R., M. R. Stampfer and G. M. Tomkins.
1974. Receptors from glucocorticoid-sensitive lymphoma
cells and two classes of insensitive clones: Physical
and DNA-binding properties. Proc. Natl. Acad. Sci. USA

71:3901-3905.

106

42. Zlydaszyk, J. C. and R. J. Moon. 1976. Fate of 51Cr-

labeled lipopolysaccharide in tissue culture cells and

livers of normal mice. Infect. Immun. 14:100-105.

 

 

 

 

APPEND IX I I

 

108

 

Sephadex G—25 Chromatography of Activated-hormone-receptor-
complex (AHRC)

HTC—cell cytosol was incubated with dexamethasone for
2 hr to allow steroid binding to receptor. This was followed
by activation of steroid-receptor by supplementing the
cytosol with 0.15 M NaCl and incubating at 20-25°C for 30 min.
Activated cytosol was passed through 20 m1 of Sephadex G-25
equilibrated with homogenizing buffer containing 0.25 M sucrose
to remove unbound steroid and salt. Fractions of 1.2 ml were
collected. Counts per minute in 0.1 m1 and absorbance at
280 nm of each fraction were measured (Fig. 1). The results
of these observatiOn indicated that hormone-receptor which
is included in the macromolecular fraction of HTC-cell
cytosol is coming off the column with the void volume. Dexa-
methasone and NaCl were held on the column and thus separated
from the macromolecular—fraction. Fractions-number 6 through
9 were used as crude AHRC.

Quantitation of Chromatin

 

For quantitation of chromatin, increasing amounts (0.05
to 0.BIMQ of crude AHRC (5.5 mg/ml protein) were mixed with
steroid free HTC—cell cytosol (5.5 mg/ml protein) for a total
volume of 0.8 m1. As a blank 0.1 m1 of the mixture was
added to 0.9 m1 of 5 mM MgC12/0.01 M tris buffer. For binding
a constant amount of chromatin (10 pg DNA) was added to 0.7
ml of crude AHRC (0.05-0.7 m1) and steroid free cytosol for
90 min at 0C. After incubation and work up (Material and

Methods for second paper), chromatin bound receptor—steroid

109

 

 

ABSORBANCE (280 nm)

2.6
2.4

.2.2

2.0

I.8

l.6

l.4

l.2

 

 

 

 

 

8 l0 l2 l4 l6 I8 20 22 24 26 28 30

TUBE NUMBER

L08 COUNTS

_____ 3H-Doxomothocono

Absorbouco at 260 nm

FIG. 1. SEPHADEX G-25 CHROMATOGRAPHY OF ACTIVATED HORMONE-

RECEPTOR COMPLEX‘

110

 

 

 

complex was calculated as cpm of precipitated assay mixture
znhmm blank . Fig. 2 shows that with constant protein when
the concentration of 3H-labeled AHRC reaches 3 mg per m1
protein, chromatin (10 pg DNA) becomes Saturated, and no more
AHRC can bind to it.

Quantitation of Activated Hormone-receptor Complex

 

For quantitation of AHRC, 0.6 ml of crude AHRC (5.5 mg/ml
protein) was mixed with 0.2 ml steroid free cytosol (5.5 mg/ml
protein). As a blank, 0.1 ml of this mixture was mixed with
0.9 m1 5 mM MgC12/0.01 M tris buffer. Different amounts of
(l pl-40 p1) of chromatin (1 mg/ml DNA) was added to 0.7 ml
of crude AHRC and steroid free cytosol for 90 min at 0°C.
After incubation and work up (Materials and Methods),
chromatin bound receptor-steroid complex was calculated as
(CPM of precipitated assay mixture-blank). Fig. 3 shows that
as the concentration of chromatin increases (1 pg to 40 pg
DNA) more AHRC becomes associated with the chromatin.

In all of the hormone receptor binding experiments,
0.4.ml AHRC (5 mg/ml protein), either in the presence or
absence of endotoxin, was incubated with 10 pl chromatin
(10 pg DNA).

Binding of Activated Hormone-receptor Complex to Sonication-

 

 

and Nuclease-extracted Chromatin
Since nuclease extracted chromatin is supposingly more
close to the native form of chromatin, attempts were made to

use the nuclease-extracted chromatin instead of sonication-

111

 

 

 

 

 

 

9

0

O!

In

If? l20

a:

0 I00

'—

3] .

o 80

u:

a:

3': 60

ND

”5 40

E

U 20
O

I2'3456

3H-RECEPTOR‘STEROID
(mg protein)

FIG. 2. BINDING OF INCREASING AMOUNTS OF 3H-DEXAMETHASONE

ACTIVATED—RECEPTOR TO CONSTANT AMOUNT OF CHROMATIN

112

 

 

 

H'RECEPTOR- STEROID

3

CPM of

FIG.

 

 

 

 

 

IO 20 30 4O

chromatin (09- of DNA)

BINDING OF CONSTANT AMOUNT OF 3H-DEXAMETHASONE

ACTIVATED-RECEPTOR TO INCREASING AMOUNTS OF

CHROMATIN

113

prepared chromatin. Nuclease—extracted chromatin was prepared
by suspending the nuclei at a concentration of 1x108 per ml in
0.34 M sucrose buffer A [15 mM Tris (pH 7.4), 60 mM KCl, 1.5 mM
NaCl, 0.15 mM spermine, 0.5 mM spermidine, 15 mM B-mercapto-
ethanol, 67 mM phenylmethylsulfonyl fluoride]. The suSpension
was made 1 mM in CaCl2 and digested with micrococcal nuclease
(1 pg/ml) for 10 min at 37°C. The reaction was stopped by
chilling on ice and addition of ethylenedinitrilotetraacetic

acid (EDTA) to a final concentration of 2 mM. The nuclei were

centrifuged for 5 min at 4,000xg, suspended in 1 m1 of 0.2 mM

 

EDTA, (pH 7) with a pasture pipette and recentrifuged for
2 min at 4,000xg. The opalescent supernatant was used as
nuclease-extracted chromatin. .
Table 1 shows that while there was about 345 cpm of AHRC
bound per 10 pg of DNA in sonication-prepared chromatin, no
significant binding of AHRC to nuclease-extracted chromatin
was observed. Since nuclease-extracted chromatin might have
lost between 25 to 50% of its DNA, the absence of significant
binding of AHRC to nuclease-extracted chromatin tends to
support the idea that the binding site of glucocorticoid-

receptors in target cell nuclei is DNA.

 

'114 __.___

Table 1

Binding of activated hormone-receptor complex to

 

 

sonication-
prepared and nuclease-extracted chromatin.
AHRC incubated with CPM
Sonicated chromatin 793
MgCl2 for sonicated chromatin 503
Nuclease-prepared chromatin 594
MgCl2 for nuclease-prepared chromatin 541

 

115

 

 

   
  

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

IHIHIIHLIIIJIII[IIIIIIIHIIIIIIIIIH

03146 3 4