HEAVY METAL AND RADlOACTWE
LABELiNG 0F CELLS REACTING T0 V
BRAIN. TRAUMA. IN MICE

Thesis for the Degree of M. S.
MECHiQAN STATE UNIVERSETY
HEATHER M. MURRAY
197 1

W

”‘5 LIBRARY
Michigan Scam
University

 
 
     

 

HEAVY METAL AND RADIOACTIVE LABELING OF

CELLS REACTING TO BRAIN TRAUMA IN MICE

BY

Heather M. Murray

A THESIS

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

MASTER OF SCIENCE

Department of Anatomy

1971

ACKNOWLEDGEMENT

I wish to express my gratitude to Doctor Bruce E. Walker for his
guidance and inspiration as major professor and advisor during the graduate
program and the preparation of the thesis.

I would like to thank Doctor Thomas W. Jenkins and Doctor Clifford
W. Welsch for serving on my committee and offering their suggestions in
the thesis preparation.

I would also like to thank Joanne Burwick for her technical assistance,
Peter Carrington for preparation of the graphs, Mary Grace for typing the
thesis, and Robert Paulson for the preparation of the photomicrographs.

Finally, I wish to express my warm thanks to Joanne Burwick, Diane

Jordan and Patricia Patterson for their friendship and encouragement.

ii

TABLE OF CONTENTS

Page
INTRODUCI‘ION O O O 0 O 0 O O O O O O O O O O O O O O O O O 1

LITERATUREREVIEW.. ...... 7
Reactiontolnjury...... 7
Experimental Allergic Brain Inflammation. . . ...... l8
Transplantation of Tumor in the Brain. . . . . ..... . 20
The Use of Radioautography. . . . . . . . . . . . . . . . 21
The Homograft Reaction. . . . . . ..... . . . . . . 27

METHODSANDMATERIALS. . . . . . . . . . . . . . . . . . 29

RESULTS.... ...... 3.7

DISCUSSION......................... 82

SUMMARY AND CONCLUSIONS. . ....... . . . . . . . 87

BIBLIOGRAPHY ............. ...........89

iv

LIST OF TABLES

Table

l.

20

10.

ll.

12.

13.

Frequency of labeled nuclei in 0. 04mm2 of uninjured areas of

operated an finals.

2
Frequency of labeled nuclei in 0. 04mm of the cortical gray
matter and of the corpus callosum in the unoperated
hemisphere.

2
Frequency of gold stained cells in 0. 0845mm of corpus callosum

of the operated hemisphere.
Statistical results for gold stained cells.
Statistical results for gold stained cells.

Frequency of radioactive nuclei in 0. 09mm2 of the corpus
callosum of the operated hemisphere.

Statistical results for radioactive nuclei.
Statistical results for radioactive nuclei.

A 2
Frequency of radioactive nuclear subgroups in 0. 09mm of
corpus callosum in the operated hemisphere.

Statistical results for radioactive nuclear subgroups.
Statistical results for radioactive nuclear subgroups.

2
Frequency of radioactive nuclei in 0. 04mm of cortical gray
matter in the operated hemisphere.

2
Frequency of radioactive nuclei in 0. 04mm of cortical gray
matter along the wound and of the corpus callosum in the
operated hemisphere.

Page

39

4O

46

49

49

SO

51

51

52

55

56

58

59

LIST OF FIGURES

Figure Page

1.

10.

11.

12.

13.

2
Percent labeled nuclei in 0. 04mm of the cortical gray matter 42
and of the corpus callosum of unoperated animals and of the
uninjured hemisphere of operated animals.

2
Percent large gold stained cells in 0. 0845mm and radioactive 48
nuclei in 0. 09mm2 of the corpus callosum of the operated
hemisphere.

Percent labeled nuclei of each nuclear type in 0. 09mm2 of corpus 54
callosum of the operated hemisphere.

2
Percent radioactive nuclei in 0. 04mm of the cortical gray matter 61
along the wound and in 0. 09mm2 of corpus callosum in the
operated hemisphere.

Percent radioactive nuclei of each nuclear type in 0. 04mm2 of 63
cortical gray matter along the wound in the operated hemisphere.

Small gold stained cells in unoperated hemisphere. 67
Large gold stained cells in operated hemisphere. 67
Large gold stained cells in the corpus callosum of the operated 69

hemisphere 3 days after stab injury.

Radioactive nuclei in the cortical gray matter. 71
Radioactive nuclei in the cortical gray matter. 71
Large gold stained cells in the corpus callosum of the operated 73

hemisphere 8 days after tumor allografting.

Large gold stained cells in the corpus callosum of the operated 75
hemisphere 8 days after tumor allografting.

Large gold stained cells in the corpus callosum of the operated 77
hemisphere 16 days after stab injury.

vi

List of Figures continued

Figure Page

14. Large gold stained cells in the corpus callosum of the operated 79
hemisphere 32 days after stab injury.

15. Large gold stained cells in the corpus callosum of the operated 81
hemisphere 64 days after stab injury.

vii

INTRODUCTION

Repair of brain injuries has been the subject of neuropathological
studies for nearly one hundred years. Gross examination of brains which
had previously been injured revealed that scars had formed in the injured
areas. The characteristics of the scars which formed in central nervous
tissue depended not only on the extent of the injury; the length of time
from injury to examination of the area also influenced the appearance of the
scars. With time, scar tissue filled in the damaged area and appeared to
wall off the area from the rest of the brain. The process of repairing the
brain seemed, therefore, to be similar to the repair of injured areas noted
in more easily accessible parts of the body. Thus, with the formation of a
protective scar, the body could lessen or avoid further damage to the very
vital nerve cells. The same mechanism that protected other areas of the
body, namely scar formation, could protect the brain from infection or
further loss of fluids.

Microsc0pic examinations of injured central nervous tissue revealed
a changing picture within the damaged area too. At a short time after injury,
great numbers of cells were to be found both in the wound areas and in the
surrounding, seemingly unaffected, areas. During the period when scar
formation was seen grossly, cells were noted to fill in and wall off the
injured areas. Many of these cells persisted at long time intervals after an
injury. Where had these cells come from? Had they been present in the
brain prior to the damage and migrated to the injured area? Had precursor
cells in the brain given rise to the great number of cells present at short
time intervals?

Within the past decade these questions and others have been answered
by means of the use of radioisotopes. Several investigators found that white
blood cells, one of the body's defense systems, invaded the area surrounding

l

2
an injury to central nervous tissue. Great numbers of these cells accumulated
in the area for some time after the injury. Many of these cells persisted in
and around the wound for a long time. The white blood cells which persisted
were the mononuclear leukocytes, the lymphoid cells. These findings suggested
that the role the lymphoid cell plays in central nervous tissue injury repair
should be further investigated. Clarification of this role was deemed necessary
due to the limitations of conventional hist010gical and special neuropathological
techniques.

With conventional histological stains, it was determined that there were
elements other than neurons comprising central nervous tissue. These other
elements, the interstitial cells, were designated as neur0glia E}. neuron,
nerve, + glia, glue] . The neuroglial cells were further classes as astrocytes,
oligodendroglia and microglia, each having a characteristic nuclear morphology.
Interest in these cells was stimulated by the study of brain injury repair, and
the cellular processes which lead to repair after injury were studied by many.
Early investigators, using conventional histologic stains, noted that in pathological
conditions of central nervous tissue the neuroglia reacted by "proliferation;"

i. e. , there were an increased number of neuroglial nuclei located around the
damaged area. Indeed, for many years neuroglial proliferative response was
thought to be the primary reaction to central nervous injury (Adams, 1958).
However, due to the limited number of mitotic figures seen in such areas, it
was postulated that this proliferation took place by amitotic, or direct, division
(Penfield, 1932).

With the development of special neuropathologic stains, the cellular
morphology of the glia could selectively be studied. The glial cells were
stained by the salt solutions of such heavy metals as silver and gold. Ram6n y
Cajal (1913) demonstrated the astrocytic cell body with gold chloride
sublimate. Del Rio-Hortega (1919) described the cellular morphology of
oligodendroglia and microglia with silver carbonate. Such staining techniques
enabled neuropathologists to investigate the role of the neuroglia in scar
formation. The astrocytes, stained with gold, were the neuroglial cells

shown to take part in scar formation. Some investigators thought this response

3
to be a secondary reaction in brain injury repair (Hicks, 1947). Other authors
proposed that astrocytic participation in the reparative process was more than
a secondary reaction (Penfield and Buckley, 1928), and possibly, along with
connective tissue elements, formed the protective scar which was seen grossly.

The major participants of the acute reaction following brain injury were
considered to be brain phagocytes. In response to experimental punctures of
the brain, these cells accumulated at the injury site and were responsible for
ridding the area of necrotic debris. The reaction began shortly after injury
and the period of most intense activity was during the first week after injury.
The reaction then gradually subsided with some cells persisting for long periods
of time after the injury. By means of conventional techniques, the brain phagocytes
were variously described as leukocytes, mononuclear leukocytes, histiocytes,
macrophages, compound granular corpuscles, gitter cells and microglia.

Using the neuropathologic silver technique, some of the phagocytes were
classified as microglia which, with time, were considered to form gitter

cells (del Rio -Hortega, 1932). Thus, two neuroglial cell types were implicated
in repair of brain injuries. The microglia phagocytized necrotic debris and
astrocytes formed, or helped form, a protective scar.

The origin of the brain phagocytes still remained in question, however.
Early investigators postulated a source extrinsic to the brain. Somewhat later
investigators held that the phagocytes were derived from cells resident in the
brain. During the past decade the issue has been resolved by the use of
radioisotopes. Tritiated thymidine radioautography permitted the differentiation
of cells undergoing premitotic synthesis of DNA from cells which were in a
non -synthetic stage. With this technique, Smart and Leblond (1961) demonstrated
a slow rate of DNA synthesis in the neuroglia of the normal adult rat brain.
These authors classified the neuroglia undergoing DNA synthesis as astrocytes,
oligodendroglia, microglia and spongioblasts on the basis of their nuclear
morphology. Labeling hemopoietic tissues with tritiated thymidine prior to stab
injury of the spinal cord, Adrian and Walker (1962) demonstrated that cells
responding to the injury were mononuclear leukocytes. By means of the same

labeling technique, Konigsmark and Sidman (1963) were able to show that at

4
least two -thirds of the phagocytes reacting to stab injury of the brain came
from mononuclear leukocytes. Walker (1963) demonstrated in spinal cord that
following tritiated thymidine injection, there was an increase with time of
radioactive nuclei at injury sites. There was, however, no corresponding
decrease in the grain count. This indicated an infiltration of cells into the
area rather than reutilization of label already present at the site. Reutilization
of tritiated thymidine in previous work had led to much weaker labeling of the
new cells than was seen in the original population (Walker, 1963; Walker and
Goldman, 1963). Also, Walker (1963) noted that the nuclear morphology of
the radioactive cells was indistinguishablefrom glial nuclear morphology. From
his investigations he concluded that in the wound, large lymphoid cells slowly
hypertrophied into cells with nuclei resembling large oligondendroglia and
astrocytes. Furthermore, astrocytes were known to aggregate around areas
of central nervous tissue injury. This seemed to correspond to the layering
of hypertrophied, elongated lymphocytes seen in response to skin homografts
(Walker and Goldman, 1963) and tumor homografts (Walker, Yates and Duncan,
1964; 1965). The characteristics exhibited by lymphocytes, namely nuclear
hypertrophy in central nervous tissue injury, and elongation and hypertrophy
of cytoplasm in allograft beds, led to a further consideration. On the basis
of nuclear morphology, lymphocytes were shown to hypertrophy and to resemble
large oligondendroglia and astrocytes. Did these same hypertrophied lymphocytes
exhibit a cytoplasmic affinity for the heavy metal glial stains? Were they, in
fact, responsible for the astrocytic aggregation seen in repair of brain injuries?

The following experiment is concerned with the role lymphocytes play
in repair of brain injuries and the morphological relationship between lymphocytes
and neuroglia, primarily, the astrocytes.

The experimental design used to obtain information about these roles was
simple stab wound in the cerebrum or allografts of spontaneous mammary
tumor to the cerebrum in conjunction with tritiated thymidine injections at
various time intervals. Portions of each injury or transplanted area would
subsequently be processed either by radioautography and nuclear staining or

heavy metal staining for neuroglial cellular morphology. This design was chosen

5
on the basis of the considerations dis cussed previously. Stab injury was shown
by heavy metal stains to produce aggregations of astrocytes surrounding the
injured area. Using radioautographic techniques in combination with nuclear
stains, it has been postulated that lymphocytes slowly hypertrophy and, with
time, come to resemble large oligodendroglia and astrocytes. In addition, the
layering of hypertrophied lymphocytes seen in allograft beds in areas outside
the cerebrum resembled the "felting" of astrocytes in brain repair described
by Linell (1929). Spontaneous mammary tumor was used because allografts
to areas outside the central nervous system evoked a lymphocytic response.
This appeared to be the same reaction which had previously been described
in homologous transplantation of spontaneous mammary tumor to the cerebrum
(Murphy and Sturm, 1923). Although Smith and Walker (1967) were not able
in tumor allografts to cerebrum to demonstrate cytoplasmic basophilia in the
responding labeled cells by the usual histologic techniques, the features of
allograft rejection seen in the investigation were similar to those seen in
radioautographic experiments on grafting skin (Walker and Goldman, 1963)
and grafting tumor subcutaneously (Walker e_t_ a1. , 1965). Furthermore, by
injecting tritiated thymidine into juvenile mice, the glial cell population was
labeled. Two months later, there was no decrease in grain count of the labeled
cell types. This indicated that these cells form a fairly stable population and
do not undergo repeated divisions. This population produced no detectable
proliferation or migration in response to tumor allografts (Smith and Walker,
1967). Infiltrating lymphocytes from the blood were presumed to be responsible
for the tumor allograft rejection. In electron microscopic examination of
tumor homografts outside the central nervous system, the lymphocytes were
seen to have elongated cytoplasmic expansions. The same mechanism could
be expected in lymphocytes responding to tumor allograft rejection in the cerebrum
with its very dense neuropil. Since infiltrating lymphocytes probably hypertrophy,
their elongated cytoplasmic expansions might stain with the heavy metal stains,
thus making their cellular morphology indistinguishable from the reactive

astrocytes seen in neuropath010gical conditions. This experiment is concerned

6
with the role of lymphocytes in brain injury repair. The use of heavy metal
glial stains in conjunction with radioautOgraphy should further define this role
and the morphological relationship between reactive astrocytes and lymphocytes.
If the two populations are indistinguishable on the basis of staining characteristics,
a more plausible solution to the problem of the increased numbers of astrocytes
noted in repair of brain injury would thus be presented, without resorting to

the theory of amitotic division.

LITERATURE REVIEW

Reaction to injury

 

Neuroglial hyperplas ia has long been considered the most characteristic
response of nervous tissue to a great variety of injurious stimuli (Adams, 1959).
The three types of neurOglia have all been implicated as responsible for the
increased numbers of cells noted in and around areas of central nervous tissue
damage.

Other cell types have also been commonly described as responding to
central nervous tissue damage. Among these are fibroblasts, connective
tissue elements, adventitial cells and vascular cells.

Tschistowitsch (1898) claimed that the glia cells swelled up and divided
actively after the introduction of sterile, small Icelloidin tubes into the brains
of mice. He felt that there was no evidence from his experimentation that
the glia "added to the host of phagocytic cells which digest and remove the
destroyed tissue, " but he did not conceive it impossible. The phagocytic cells
were thought to be derived from connective tissue cells of the pia and from
the adventitia of the penetrating vessels, partly from the endothelial lining
of blood vessels and leukocytes. Repair of the lost parts was considered to be
made almost entirely by connective tissue. Nor did the glia take part in the
formation of the cicatrix which replaced the destroyed elements. Rather, glial
cells were reported to be merely concerned with the elaboration of a secondary
zone of sclerosis around the scar.

Essick (1919), studying experimental puncture wounds in dogs, reported
similar results. On gross examination two months after injury, there was solid
replacement of destroyed nervous tissue by connective tissue. Microscopic
examination of damaged areas of 2, 3, and 4 months duration showed that
leukocytes had disappeared from the area, but phagocytic cells were still
present in considerable numbers. Connective tissue had filled in the defect,

7

8

and between this replacement tissue and normal nervous tissue was a zone
of reactive cells, which Essick termed a zone of gliosis. Healing was
reported to be by the ingrowth of connective tissue because no glia cells
could be found within the scar tissue itself.
Macklin and Macklin (1920) lesioned rat brains with hot and cold needles,
then injected dye into the animals two days later to study vital staining
of macrophages responding to the injury. Fully developed macrophages were
found in the lesioned area from time intervals of 2 to 74 days after dye
injection. The period of most intense activity was from 2 to 6 days after
injection when vitally stained mononuclear cells containing much lipoid material
were found in the inflamed region. However, when these cells were compared
with macrophages of wounds of other parts of the body, they noted that the cells
took up a lesser amount of stain and the duration of the staining period was
shorter. The macrophages were thought to multiply by mitosis at the injury
site and no transitional forms from lymphocytes were found. It was believed
that the vitally stained cells were recruited in part from neuroglia, because
hypertrophied neuroglial cells containing dye granules and other material
were found in the area of inflammation. (This observation was made with
hematoxylin and eosin stain.) In the healing process, macrophages were
incorporated into the scar which had been formed by fibroblasts. There was
no evidence in the investigation that the macrophages had actually proliferated
to form the scar. When foreign tissue such as muscle, liver or spleen was
introduced into the lesion in an effort to attract macrophages similar to those
found in other areas of the body, negative results were obtained. The
responding cells did not show transitional forms from lymphocytes, there were
no greater numbers of cells, there was no increase in dye -uptake by the cells,
nor were there more mitotic figures than observed in hot or cold needle lesions.
Penfield (1925) studied microglia and the process of phagocytosis in
gliomas. He concluded that the microglial activity surrounding a glioma was
similar to that described by del Rio-Hortega in other destructive processes
of the central nervous system. They were seen by the author to act as

scavengers of products of degeneration and cerebral destruction. In an area

9
surrounding a tumor, microglia took on ameboid forms; in areas of cerebral
softening, more spherical and reticulated forms (Gitterzellen) were found.
Regions of very active mitotic division were seen. The author stated that
the function of microglia was to transport substances to outer surfaces of
blood vessels, where they lost their granular and vacuolar appearance,
decreased in size, and finally left the vicinity in characteristic form "for
renewed phagocytos is. " Penfield concluded that the presence of scavenger
cells about blood vessels was indicative of the transfer and delivery of
ingested substances rather than new formation of these cells from fibroblasts.

Bender (1925a, 1925b) noted "moderate to marked gliosis" in dog brains
which had been experimentally punctured and injected with various extracts
of nervous tissue. The observations were made in wounds of 4 weeks duration.
The effect on the blood vessels was the same in all the experimental animals.
Those vessels in and adjacent to the lesion were prominent due to a slight
increase of endothelial and fibroblastic elements. Phagocytic reactions varied
directly with the amount of hemorrhage and tissue destruction and inversely
with the amount of encapsulation. Formation of a fibrous capsule was seen
in three of five experimental groups: the injection of cerebrin, cholesterol
and solid residue. Gliosis of the spider cell, fibril-forming type and cytoplasmic
type, was noted adjacent to the lesion in all groups and ranged from "moderate
to marked" reaction.

In a study of human puncture wounds, Wilson (1926) found connective
tissue cicatrices in injuries 7 and 11% months old. Punctures made with a
needle 1 mm in diameter showed no "gliosis" per se. In none of the needle tracks
did the neuroglia appear proliferated at the edge, forming a limiting membrane;
rather, the connective tissue elements seemed to blend with the neuroglial
fibers at the edge. The reaction appeared more prominent in the gray matter
than in white, but the author concluded that neuroglial elements played only
a small part in repair of the damaged area.

Formation of a cicatrix in rabbit brain experimentally injured with a
blunt needle was studied by del Rio -Hortega and Penfield in 1927. They felt

that repair of lesions could be divided into 3 stages. The first stage was thought

10

to be characterized by microglial phagocytic activity which continued for a
long time. Following this, the astrocytes were believed to become swollen
and to undergo degeneration by "clasmatodendrosis. " This activity was
thought to set in motion the third stage which the authors described as being
characterized by amitotic division of astrocytes. The cells produced by this
hyperplasia were reported to become fibrous and also to arrange themselves
in a radial manner around the wound. Astrocytes were stained with Cajal's
gold chloride sublimate stain. Hortega's silver carbonate stain was used to
demonstrate micrOglia. These cells were described as passing through transitional
forms to become mature macrophages, compound granular corpuscles. At the
center of the tracks, a strong connective tissue core was noted.

Because of the findings that occasional open tracks were found, without
a connective tissue core and with little or no gliosis in the surrounding area,
Penfield and Buckley (1928) injured dog brains with blunt and open needles.
When blunt needles were used, all the injured tissue remained behind. This
resulted in a closed track with a core of connective tissue which was firmly
attached to the dura. The track was surrounded by a moderate gliosis in the
cortical gray matter, while gliosis did not occur or there was even a decrease
in astrocytes of the white matter. On the other hand, when punctures were
made with a cannula of the same diameter as the blunt brain needle used, a
gaping track resulted. This defect contained little or no connective tissue.
It was surrounded by a moderate gliosis in the cortical gray matter and no
gliosis or even a decrease in astrocytes was noted in the white matter. The
authors contrasted the wounds made by both types of needles. About closed
tracks, the astrocytes sent in their large expansions concentrically; about open
tracks, these expansions were tangential to the canal. In general, the presence
of damaged brain tissue in a wound produced a contracting cicatrix, gliosis in
the contiguous areas and cerebral distortion. Clean removal of cerebral
tissue resulted in little, if any, cicatrization, moderate gliosis in the gray
matter and none in the white matter, slight superficial adhesions, and

practically no distortion of the brain.

11

Globus (1928) reported that in chronic vascular disease of the brain, the
changes in the astrocytes were progressive, with generalized hyperplasia,
hypertrophy and transformation of the protoplasmic type into the fibroblastic
variety. Oligodendrocyte changes in the investigation were mainly expressed
in generalized hyperplasia and mild hypertrophy. The microglia were described
as undergoing regressive changes to the final formation of compound granular
corpuscles, gitter cells.

Cone (1928) concluded that there was a marked histologic difference in
the acute responses which neuroglia and microglia underwent in central nervous
damage. Astrocytes and oligodendroglia reacted in acute disease of the central
nervous system by degenerating as did the nerve cells. Microglia responded
in acute destructive lesions of the central nervous system by forming rod cells
and then compound granular corpus cles. The author thought that metamorphosis
from one cell group to another did not occur in acute changes. This study was
done with the special glial stains.

Ferraro and Davidoff (1928) studied the reaction to removal of a portion of
the cerebral cortex by Hortega's silver carbonate technique. They suggested
that the oligodendroglia reacted by "acute swelling, atrophic degeneration,
hypertrophy, hyperplasia and occasionally by the fusion of several cells into
clusters. " It was believed that oligodendroglia transformed into compound
granular corpuscles.

In 1929, Linell investigated the neuroglial changes following cerebral
trauma. In this experiment dog brains were injured with an open cannula and
the neuroglial reaction was studied over a 2-month period by neuroglial stains.
This reaction was reported to be a dual response, a microglial reaction and a
macroglial reaction. The microglial reaction consisted of the transformation of
micrOglia into compound granular corpuscles which in their fully developed
condition, were large, round, probably ameboid cells, heavily loaded with
lipoid material. This change took place rapidly, being well developed by 3 days
after injury and reaching its maximum at 6 days after injury. Linell commented
to the probable function of microglia and compound granular corpuscles: to

phagocytize damaged nervous tissue and transfer of this degenerated tissue to the

12
blood stream for disposal. The macroglial reaction was characterized by later
onset. Slight hypertrophy of astrocytes was reported in the wound area by 3
days, more marked hypertrophy surrounding the area of microglial reaction at
6 days after injury. At 14 days there was marked hyperplasia and hypertrophy
of astrocytes reported by the author. He reported a definite orientation of
strong astrocytic processes in a direction at right angles to the wound margin
at this time. These processes reached the edge of the wound track. At 3 weeks,
the astrocyte reaction was described as reaching its maximum. Some "felting"
of astrocytic processes in the long axis of the wound to form scar tissue
comprised of glial fibers was noted at 4 weeks. By 2 months, "the glial feltwork
had thickened considerably and few cell bodies were observed in the glial

' 9 9|

mat . The author considered the purpose of the reaction to be the formation
of a protective scar. "The scar of a healed wound of the brain is the terminal
result of the mirroglial reaction. "

Russell (1929) studied experimental cerebral puncture reaction with trypan
blue and the silver carbonate stain for microglia. Rabbits were used due to
the ease with which microglia were demonstrated in this species by the silver
carbonate stain. Intravital staining of the microglia was observed while there
was "no evidence of the formation of Gitterzellen from neuroglia or that the
neuroglia was able to take a vital stain. " The ability of microglia to phagocytize
dye granules led to the conclusion that microglia were a part of the reticulo-
endothelial system.

Wells and Carmichael (1930) studied microglia in tissue culture. The
silver carbonate method for staining the microglia of the brain also was
effective in demonstrating cells present in cultures of tissue such as peritoneum
and embryonic limb buds. These authors concluded from their in vitro ex-
periments that microglia were of mesoblastic origin. This was in agreement
with the conclusion of Hortega (1919) that micrOglia were a mesodermal element.

Rand and Courville (1932a, l932b) described the histologic changes in the
brain in cases of fatal injury to the head. Microglia, oligodendroglia and astrocytes

were all considered by these investigators to undergo amitotic division. Mitosis

was found only in "fully developed compound granular corpuscles. " Microglia

13
were thought to react to local destruction and disintegration of tissue;
oligodendroglia altered in response to general or widespread processes, par-
ticularly edema; and astrocytes reacted to local destruction and disintegration
of tissue. The authors believed that gliosis only occurred as a result of tissue
destruction arising regionally; generalized gliosis following head injury was
thought not to occur. Microglia transformed to fully developed phagocytes,
compound granular corpuscles. Oligodendroglia became enlarged and
vacuolated almost immediately after injury. The authors claimed that astrocytes
beyond the zone of necrosis were seen to actively proliferate. In the formation
of glial scars, the astrocytes were considered to compensate for loss of tissue
and to fill up the defect. At the margins of lacerations and in severely
contused areas, microglia, oligodendroglia and astrocytes were thought to
undergo destruction along with nerve cells.

Greenfield (1932), in describing the cellular response to nervous tissue
injury, wrote that a response by mononuclear leukocytes was extremely common.
He felt that these cells were not derived from the circulating blood, but rather,
were formed through multiplication of the lymphoblasts already present in the
perivascular space.

Dunning and Furth (1935), in studying the relationship between microglia,
histiocytes and monocytes, demonstrated that histiocytes of chick embryo liver
and peritoneal membrane possessed an affinity for the silver carbonate stain
which revealed the microglia of the brain.

Hassin (1936) studied the reaction to implantation of a foreign body into
central nervous tissue. He introduced cotton plechets into dog brains and
followed the ensuing reaction for up to 6 months. The author concluded that
there were 2 types of gitter cells: 1) ectodermal, which phagocytized destroyed
nerve tissue and were formed by oligodendroglia and 2) micrOglia cells, which
phagocytized non -nervous material. The types of cells which were seen responding
to cerebral injury depended on the character of the injury and varied according
to the type of foreign body present in the damaged area. Microglia were not
the only source of gitter cells. Gitter cells were formed from microglia in

acute, severe neuropathologic conditions in which not only nerve tissue, but

14
also the glia, succumbs. Gliogenous gitter cells occurred only in cases of
isolated destruction of nerve fibers, the glia remaining unaffected. "When
destruction of nerve tissue is not massive, but selective and limited to single
nerve fibers as in multiple sclerosis, the reaction is also selective and is
confined to the glia. " In areas of secondary and primary degeneration, such
as that seen in some cerebral lesions, the neuroglial type of gitter cells were
reported. This observation "speaks against a vascular factor in their histogenesis,
as a vascular factor produces massive destruction of the nerve tissue, with
formation of microglial (mesodermal) types of such cells. "

Carmichael, Kernohan, and Adson (I 939) wrote that the primary response
of nervous tissue to injury was that of phagocytes and the proliferation of
fibrous tissue around the damaged area. The glial response was thought to be a
secondary reaction.

Baggenstoss, Kernohan, and Drapiewski (1943) studied wounds made with a
trochar in human material. The authors described changes seen in wounds of
several hours to 7 years duration. In wounds of less than 5 days duration, there
was acute swelling of oligodendroglia noted by the authors. Compound granular
corpuscles were first seen 3 days after injury. Transitional forms from microglia
to compound granular corpuscles were not found. The numbers of compound
granular corpuscles were seen to increase through the second week, and many
persisted in the area through 6 months duration. Hypertrophied astrocytes
were noted at 4 days and persisted throughout 7 years. Scavenger cells
appeared to be derived largely from the endothelial cells and adventitia of larger
vessels as well as from blood mononuclear cells. In wounds of 8 to 10 days,
significant numbers of fibroblasts and collagen fibrils were first described.
Lymphocytes were seen in perivascular spaces in increased numbers; numerous
transitional forms between lymphocytes and fully developed compound granular
corpuscles were described by the authors. By 3 weeks, 3 zones were seen: a
central zone of necrosis, partly filled with compound granular corpuscles; a
capillary zone with fibroblastic proliferation and numerous lymphocytes; and
a peripheral zone of hypertr0phied astrocytes. An increased number of astrocytes

over the previous period was not noted, nor did they appear to have a particular

15
arrangement or orientation to the wound as had been previously reported by
Penfield 9:31; (1928) and Linell (1929). There was an increase in the number
of astrocytes in the cortex around a wound of over 6 months duration. At 7
years, the wound center was loosely filled with collagen fibers. No phagocytes
or lymphocytes were seen, the capillary zone had disappeared, however, the
astrocytes persisted. The authors concluded that astrocytes play a minor role
in the reparative processes.

Dougherty (1944) concluded that the first cells reacting to experimental
brain wounds in the adult animal were lymphocytes which transformed to
macrophages. Compound granular corpuscles arose from microglia and in
imprint preparations were typical histiocytes which could not be differentiated
from lymphogenous macrophages.

Dublin (1945) studied the histogenesis of compound granular corpuscles in
response to heat cautery in the brain of rabbits. He concluded from his experiments
that compound granular corpuscles appeared to arise principally by proliferation
and differentiation of the adventitia of blood vessels, and less frequently from
vascular endothelium. Microglia commonly showed some degree of swelling
and loss of processes, but transformation to mature compound granular corpuscles
were only infrequently seen.

In 1947, Silver and Walker studied thermal lesions of the cerebral cortex.
Five days later with the use of colchicine, many mitotic figures were noted in
cells associated with capillaries. They thought it possible that perivascular
cells were the origin of compound granular corpuscles. There was no evidence
noted of microglial proliferation over a 6 month period following injury.

Hicks (1947) studied experimental wounds of mouse brains with blunt, or
occluded needles, open needles and punctures with an open needle through which
was injected foreign material: bone particle suspension, various extracts of
nervous tissue, and a suspension of avian tubercle bacilli. The gold chloride
sublimate stain was not used in this study to demonstrate gliosis, defined by
Hicks as a proliferation of fibrillary astrocytes and a laying down of their fibrils.
In all experimental groups except that of the longest time interval of puncture

with the introduction of avian tubercle bacilli, the results were the same. After

16
the first 48 hours, mononuclear phagocytes accumulated at the track margin and
ingested red blood cells and cellular debris. This reaction reached its maximum
by the end of the first week. Some phagocytes with debris persisted in the
wound for as long as 270 days. Fibrous tissue proliferated along with the acute
phagocytic response and there was conspicuous fibroblastic response from the
injured meninges. By the second and third week, the track was filled with a loose
fibrous core. Throughout the length of the experiments there was almost no
proliferation of astrocytes and glial fibrils could be seen only extremely rarely.
The narrowing of the track was attributed to the gradual disappearance of
phagocytes and the shrinking and disappearance of fibrous tissue. On the other
hand, the introduction of avian tubercle bacilli produced a different picture. At
48 days after injury, fibroblasts were in the track. Just at the outer edge of this
zone, there were fairly numerous astrocytes with proliferated fibrils arranged
in a somewhat radial fashion toward the wound track. The gliosis was attributed
to the release of toxins into the tissue causing sublethal injury, but little necrosis.
Hicks concluded that direct injury of astrocytes was necessary to stimulate
gliosis, and that proliferation of astrocytes was a secondary healing mechanism.

Swank and Hain (1952) reported adventitial proliferation 1 week after
injecting emboli into the arterial system of the brain. Experimental transient
ischemia was studied in dogs by Lewis and Swank (1953). The authors reported
an increased number of perivascular oligodendroglia.

Collias and Manuelidis (1957) described the histopathological changes
observed in cat brains with implanted electrodes. In wounds of less than 1 week
duration, the primary characteristic of the response was considered to be
microglial proliferation and formation of gitter cells. From 1 to 2 weeks,
hypertrophy of astrocytes, streaming of fibroblasts, and perivascular "round-
cell infiltrates" were seen. At 1 month the microglial reaction had abated,
but the astroglial reaction had greatly increased over previous stages. In wounds
of 2 to 6 months duration, there was scar tissue formation "with an obvious

H

tendency to encapsulate the implanted electrode. In general, they reported

regressive changes in all types of responding cells.

l7

Klatzo, Piraux and Laskowski (1958) described the reaction of central
nervous tissue to edema caused by low temperature injury. Endothelial
proliferation followed by an intense astrocytic gliosis was observed in these
experiments. Studying stab wounds with the use of the electron microscope in
1959, Schultz and Pease were unable to identify the responding cells as lymphocytes.
Following d-particle irradiation to produce lesions of the cerebral cortex,
Maxwell and Kruger (1965, 1966), using the electron microscope, found hypertrophy
and hyperplasia of pericytes, which were considered to develop into phagocytes.
Vaughn (1965) also reported that phagocytes develop from pericytes and adventitial
cells.

Dublin (1967, p. 214-217) discussed the stages of cellular reaction to
traumatic intracerebral hemorrhage. The first response, during the first 2 days,
was considered to be the infiltration of polymorphonuclear leukocytes. Lymphocytes
appeared about 2 to 3 days after damage, which was thought to be followed by
the entrance of macrophages into the area. In lesions of large proportions,

" cells proliferate in great streams from the vascular adventitia, accompanied

by mononuclears from the blood stream. " Blood vessels and glia were considered
by the author to enter the area more slowly than macrophages. "Especially in
lesions which don't reach the cortical surface, healing which is in addition to
vascular proliferation is mainly glial. " If the cortical surface were broken in
injury, the meninges entered the reaction, and connective tissue was extensively
deposited.

Fuentes and Marty (1970) described the reaction to electrolytic lesions in
the cerebral cortex of the cat by the use of the light and electron microscopes.
They reported that at all time intervals after damage, reactive oligodendroglia
were seen as hypertrophic and increased in numbers; there was also hyperplasia
of astrocytes. The authors concluded that the neuroglial proliferation was
produced by amitotic division.

Cavanagh (1970) studied the proliferation of astrocytes around a needle
wound in the rat brain. Adult rats received experimental puncture wounds and
groups were killed by perfusion at l, 2, 3, 4 and 6 days after injury. Tissue

sections were not stained with silver or gold neuroglial stains. Mitotic activity

18
was based on the number of mitotic figures seen in white and gray matter, and
mitotic figures were classified according to size and phase. "Small" mitotic
figures, measuring less than 8» in diameter, were considered to be macrophages,
micr0glia, oligodendroglia and other cells. "Large" mitotic figures, measuring
more than 8» and usually 12-15}1. in diameter, were considered to be astrocytes.
Some astrocyte nuclei in metaphase measured as much as 20-25}; in diameter.
The author reported dividing astrocytes up to 120%} from the wound edge, but
that the number of "large" mitoses were always maximal between 200 -40(hi.
from the wound. The number of mitotic figures seen in immersion -fixed material
was reported to be less than half the mean mitotic rate visible in perfusion-fixed
material. This suggested that delay in the fixation of brains may have led to
earlier reports of amitotic proliferation of astrocytes. When colchicine was
injected directly into the brain wound, greater than twice as many colchicine-
arrested metaphases were seen than when colchicine was injected intraperi-
toneally. The author concluded that there is no inherent resistance to colchicine
in small and large glial cells of the brain; rather, colchicine given systemically
does not have the ability to reach the dividing cells in the brain. There was an
exponential decline of mitotic rates /unit area with distance from the wound
edge. Away from the immediate vicinity of the injury, peak proliferation of all
non -neuronal cells was reported to occur on the second and third day after
injury. It was suggested that this may have been a response to the diffusion of
materials in the escaped plasma from the region of vasular damage. Mitoses
in astrocytes were thought to form up to one -fifth of the total cell divisions,

and no evidence of an amitotic process of division could be detected.

Experimental allergic brain inflammation

 

Good (1950) induced acute non -allergic and allergic inflammations in the
rat brain. He reported formation of phagocytes in central nervous tissue from
hematogenous lymphocytes. In comparing allergic with non -allergic inflammatory
states, it was found that allergic inflammation was more severe and the
hist010gical and characteristics differed. Numerous plasma cells, giant cells
and eosinophils differentiated the allergic from the non -allergic state. A second

challenge in a previously sensitized animal produced a heightened reaction.

19

Paterson (1960) demonstrated that transfer of allergic encephalomyelitis
can be accomplished by injection of lymph node cells obtained from donor rats
sensitized to spinal cord into rats pretreated neonatally with normal rat spleen
cells. The recipients acquired immunological tolerance to donor cells by
pretreatment and, thus, the donor lymph node cells could survive and function
longer after transfer.

Studying lesions of experimental allergic encephalomyelitis in rats,
Waksman and Adams (1962) observed that the lesions appeared to start with
perivenous infiltration by mononuclear cells. These cells invaded the nervous
parenchyma and seemed to destroy myelin directly. Plasma cells began to appear
in the vascular cuffs, choroid plexus and meninges 2 to 3 days after the lesions
first developed.

Levine, Hirano and Zimmerman (1965) studied hyperacute allergic
encephalitis in rats with the use of the electron microscope. They described
altered vascular permeability and leukocytic invasion of the parenchyma by
escape through the walls of venules.

Lampert (1967) investigated "ordinary" and "hyperacute" experimental
allergic encephalomyelitis (EAE) of rats with the electron microscope. He
described the method of inducing lesions of both types of EAE and the characteristics
of the resulting lesions. Early lesions of both ordinary and hyperacute EAE
showed perivascular accumulations of mononuclear cells. The lesions appeared
to be initiated by a few sensitized mononuclear cells which penetrate the walls
of venules and destroyed oligodendroglia and myelin sheaths. Many non-
specific mononuclear cells followed, stripped and digested the remnants of partially
lysed sheaths and perhaps also the sheaths related to necrotic oligodendroglia.

Cammermeyer (1967) studied microglial cells in diffuse and granulomatous
encephalitis in the rabbit. The author concluded from silver impregnated
material that the increase in number of microglia is associated with mitotic
division of a "juxtavascular histiocyte. " There was no evidence suggestive of
migratory activity or amitotic division of microglial cells. Transmissible mink
encephalopathy was studied by Marsh and Hanson (1969). Intense astrocytosis

was demonstrated with Cajal's gold chloride sublimate stain. Duckett and Pearse

20
(1969) described periventricular astrocytosis and acute necrotizing encephalopathy
in rat brains. In brains with foci of necrosis, there was perivascular cuffing.
The cells present in the cuffs and in the foci were mononuclear, and were con-
sidered to be small lymphocytes. The authors reported the areas adjacent to
the ventricular system free from histologically appreciable evidence of inflam-
matory processes. However, there was marked periventricular gliosis. The
"cellular response in the periventricular region was not evident on examination
of sections stained with traditional methods, yet it was remarkably evident
with the use of histoenzymatic techniques. " The authors suggested that some of
the cells of the choroid plexus contributed to the cells which became periventricular

astro cytes.

Transplantation of tumor in the brain

 

Following Shirai's report (1921) that grafted heteroplastic tumor grew as
well in alien host brain as in homologous host brain, Murphy and Sturm (1923)
investigated the brain as a locus for transplantation. These authors found
that transplanted mouse tumor grew well in rat, guinea pig, and pigeon brains if
the tumor did not come into contact with a ventricle. A graft lying in the
ventricle, or even coming into contact with the ventricle, led to a reaction in
which the graft remnant was found embedded in a mass of reaction tissue similar
to that which appeared "about an heterologous graft in subcutaneous tissue. "
This rejection phenomenon was characterized by the graft being surrounded by
a zone of cells in which the lymphoid variety of cell was dominant. The intensity
of the cellular reaction was seen .to be roughly proportionate to the degree of
relationship existing between the two species concerned in the experiment. The
perivascular spaces of the blood vessels in the immediate vicinity of the
heterologous transplant showed a pronounced collar of round cells, and the
lumina of small vessels were often blocked with lymphocytes. As the tumor
grew and reached the ventricle, the choroid plexus became enormously swollen
and engorged with lymphocytes. The contact with this collection of lymphocytes,
which frequently invaded the tumor first, caused necrosis of the portion of the

tumor lying near the ventricle. This was followed by necrosis of the entire tumor.

21
When adult autologous spleen was introduced into a non -resistant animal in
conjunction with heterologous tumor tissue, the tumors failed to grow. In a
later study, Medawar (1948) found that homotransplants of the skin survived in

the brain, if and only if the transplants failed to become vascularized.

The use of radioautography

 

Hughes fl (1958) reported the use of a radioisotope of thymidine, which
is a specific precursor of DNA, in an investigation of cellular proliferation in
the mouse. They found the specific advantage of tritium for radioautography to
be the very high resolution obtained. This was due to the very weak energy and,
thus, the short range of its 3 radiation. "The maximum range in tissue of a
tritium B-ray is only 6J1. and half of the 3's will travel less than 1 )1. " Tritiated
thymidine was rapidly absorbed from the blood stream. The authors concluded
that all labeling of nuclei occurred during the first hour after injection. Cronkite
gt_a_l_._ (1959) investigated the use of tritiated thymidine in the study of DNA
synthesis and cell turnover in hemopoietic tissues. They reported a very
rapid incorporation of tritiated thymidine into the hemopoietic tissues throughout
the body. Cells were present both in the bone marrow and in the peripheral
blood which were capable of synthesizing DNA and were thus able to divide
at least once again. The effective availability time of tritiated thymidine was
believed to be less than 60 minutes, during which time it was either degraded
or incorporated into DNA. When thymidine was injected subcutaneously, it
remained in the body 30 minutes. The authors reported that in dogs, small
lymphocytes with dense nuclei were rarely labeled as long as 36 hours after
injection. In human beings only a rare labeled small lymphocyte was found in
peripheral blood and then only after several days. However, large and
medium -sized lymphocytes appeared promptly. Leblond, Messier and Kopriwa
(1959) studied tritiated thymidine as a technique for investigation of renewal of
cell populations. They reported the radiation damage to be less than that
produced by phosphate -P32 and adenine -C14. Because of this, it was possible
to use doses of thymidine -H3 which allowed cells to be traced over periods of

months in the body without apparent damage to these or other cells. Messier,

22
Leblond and Smart (1958), using tritiated thymidine, found a large number of
labeled nuclei in the subependymal region of the lateral ventricle. Within 3
hours of injection, labeled neuroglial cells were found scattered throughout the
brain. These cells were located at such a distance from the subependymal layer
that the authors considered it unlikely that the labeled cells had migrated such
a distance in so short a time. These findings indicated that there was another
population of cells able to synthesize DNA in central nervous tissue in spite
of the apparent absence of mitotic figures.

Further investigation of the uptake of tritiated thymidine by the neuroglial
cells of normal central nervous tissue Was reported (Hain, Rieke and Everett,
1960; Messier and Leblond, 1960; Smart, 1960, 1961; Adrian and Walker, 1962;
Noetzel, 1962; Altman, 1963; Noetzel and Rox, 1964; Hommes and Leblond, 1967;
Dalton, Hommes and Leblond, 1969; and Mori and Leblond, 1969a, 1969b, 1970.)

Smart (1961) reported that although there was mitotic activity within the
subependymal layer of the adult rat, very few of these newly -formed cells migrate
from the layer. In general, he proposed that most of the cells degenerated
within the layer. Smart and Leblond (1961) studied neuroglial DNA synthesis
in 3 ages of mice. They found labeled neuroglial cells in the normal brains of
3 gram, 18 gram and 28 gram mice. The cells were classified according to
nuclear morphology. These were described as oligodendroglia, astrocytes,
small dark and medium dark nucleated cells. The number and distribution of
the different types of labeled neuroglial nuclei were then recorded. In both the
3 gram and 18 gram groups, the evolution of labeled nuclei after injection of
tritiated thymidine was similar. On the basis of their work, the authors concluded
that oligodendroglia and astrocytes develop from spongioblasts, that is, the small
dark and medium dark nucleated cells, and that the normal brain neuroglial
population turns over slowly with time. Adrian and Walker (1962) used tritiated
thymidine to study DNA synthesis in glial cells of thenormal adult spinal
cord. They found so few labeled cells that the glial cells were not considered
to constitute a renewal system, nor did the ependymal cells appear to be a stem
cell population for neuroglia in the normal animal. In 1963, Altman used tritiated

thymidine to study cell proliferation of rat and cat neuroglia. He found some

23
labeled neuroglial cells in all parts of the brain, suggesting a low rate of glial
proliferation in the normal animal.

Mitotic division of neuroglia in the normal adult rat was studied by Hommes
and Leblond (1967). They found labeled neuroglial cells throughout the brain.
Labeled mitotic figures were also seen. They concluded from their work that
the labeled cells of the brain arose locally and were not of hematogenous origin.
The relatively small size of the labeled nuclei suggested that these were mostly
oligodendroglia or other spongioblastic precursors of oligodendroglia. The
authors combined radioactive labeling with a silver carbonate stain and found
some labeled cells with cellular morphology characteristic of microglia. It was
suggested that microglial mitoses make up about one -fifth (10-31%) of the dividing
cells of the normal brain. Dalton, Hommes and Leblond (1969) used tritiated
thymidine to study the correlation of glial proliferation to age in the normal adult
mouse. They reported a small degree of glial proliferation in adult mice. They
suggested that the glial population turned over slowly with the production of glial
cells being balanced by the loss or degeneration of an equal number of glial cells.
Continued activity of some of the brain nuclei was postulated to account for the
glial proliferation observed throughout life.

Mori and Leblond (1969a, 1969b, 1970) investigated neuroglial proliferation
in normal young rats using tritiated thymidine. Electron microscopic examination
of the cerebral cortex and the corpus callosum revealed that, under normal
conditions, microglia did not undergo mitosis, but astrocytes and some oligoden-
droglia were able to divide. Some degenerating astrocytes and oligodendroglia
were described by the authors. They suggested that mitosis in 60-80 gram rats
may be associated with growth and with myelination of brain tissue. The ability
of these cells to divide could compensate for cells lost by degeneration and provide
replacement cells in a population with a slow rate of turnover.

The major definitive work that indicated mononuclear cells of the blood
responded to central nervous tissue injury was done by Adrian and Walker (1962).
Tritiated thymidine was injected to label hemopoietic tissue sufficiently in advance
of injury such that no isotope would be available for DNA synthesis by cells pro-

liferating in response to the injury. When the injury area was examined, many

24
labeled nuclei were observed. The radioactive cells were mononuclear and did
not resemble any of the unlabeled glial cells in the adjacent tissues except for
microglia. Since the responding labeled nuclei could not have arisen from cells
resident in the injured tissue, they must have migrated into the tissue. The
only large pool of labeled cells close to the injury was the blood. The invading
cells were mononuclear and, therefore, were lymphoid cells. Walker (1963)
demonstrated that when this same response is evoked and followed over the
course of 32 days, the nuclear morphology of the labeled cells became similar
to glial cells resident in nervous tissue. Radioactive small, spherical nuclei
reached a maximum concentration at 5-9 days after injury and then tended to
decrease. Other cells with radioactive, spherical to oval nuclei tended to
increase with time and the largest were most plentiful at 32 days after injury.
The author suggested that the radioactive cells with irregular nuclei which
resembled microglia were initially monocytes. The other lymphoid cells of
the blood, the large lymphocytes, were initially present in the injury area.
The area these cells occupied at early time intervals "was later occupied by
progressively larger and paler radioactive nuclei until, at 32 days, a large number
of the radioactive nuclei showed morphology typical of neuroglial cells, especially
of the large oligodendroglia and astrocytes. " No evidence was seen for migration
of radioactive cells from surrounding tissue, so the radioactive cells present
at the injury appeared to be the same cells at all time intervals. The author
proposed that nuclei of many of the large lymphoid cells at the wound slowly
underwent pronounced hypertrophy into cells with nuclei resembling large
oligodendroglia and astrocytes.

The technique of labeling hemopoietic tissues prior to injury was subsequently
used to determine the origin of reactive cells in stab injury of the brain
(Konigsmark and Sidman, 1963; Huntington and Terry, 1966; Roessmann and
Friede, 1967, 1968), skin homograft reaction (Walker and Goldman, 1963)
and tumor allograft reaction in the mouse brain (Smith and Walker, 1967).
Konigsmark and Sidman (1963) reported that about two -thirds of the macrophages
appearing about a stab injury to the brain were labeled leukocytes of the blood,

and the authors thought that the cells were probably monocytes. It was concluded

25

that less than 50% of the macrophages appearing in the first 2 days after
injury, and less than 30% thereafter, arose endogenously. The investigators
injected thymidine -C14 after injury and stained sections with Cajal's gold
chloride sublimate solution. The sections were then processed for radioautography.
It was concluded that labeled cells with large nuclei did not take the stain, and
conversely, no astrocytes, as defined by the staining results, were labeled.

Walker and Goldman (1963) demonstrated migration of medium and large
lymphoid cells from the blood to the skin homograft and underlying tissue. Some
round cells entered the donor skin. Most of the lymphoid cells remained in host
tissue in areas under lying the graft. The authors reported that many of these
cells hypertrophied and elongated, and their cytoplasm developed strong,
ribonuclease -labile, basophilia. The number of lymphoid cells migrating to
isografts and autografts was less and the cytoplasmic basophilia was weaker
than that seen in homograft reactions. After all types of grafting, large,
basophilic fibroblasts proliferated and became plentiful. The authors proposed
that mitoses occurred rarely, if at all, in the hypertrophied and elongated
lymphoid cell population.

Smith and Walker (1967) employed tritiated thymidine as a means of labeling
2 populations of cells to differentiate their responses. Tumor transplantation was
used for immunological stimulation and brain injury. Glial cells in the adult
animal were considered to proliferate at a very low rate; in juvenile animals,
however, there was thought to be extensive DNA synthesis. These cells were
labeled in juvenile animals, and most of them retained their label for a long
period of time. Leukocytes, because of their repeated divisions, could be
expected to lose their label over a period of time by dilution below radioauto-
graphically detectable levels. Tumors were allografted into adult animals with
labeled mononuclear cells still present in the blood. An infiltration of these
cells occurred in the area surrounding the tumors, which were rejected by 8 days.
The authors suggested that some of these labeled cells were monocytes, but
many of them were apparently medium -large lymphocytes. Glial cells were
labeled in juvenile animals and tumors were allografted when the anirnals had

matured. There was no evidence that astrocytes and oligodendroglia migrated

26
or proliferated in response to brain injury of immunological stimulation. The
authors concluded that the cells infiltrating the cerebral tissue around tumor
allografts must have come almost entirely from the blood.

Kosunen, Waksman and Samuelsson (1963) studied experimental allergic
encephalitis with tritiated thymidine radioautography. A large portion of the
cells of the perivascular infiltrates were found to be labeled by tritiated thymidine.
Little label was taken up by the brain cells until the appearance of lesions. The
authors concluded that the infiltrating cells were hematogenous and proliferated
actively in the lesions. The cells were characterized cytologically as medium
or large blood lymphocytes which evolved to become histiocytes during the
evolution of the disease.

In a study of cell division occurring in injured spinal cord, Adrian (1968)
used tritiated thymidine radioautography to differentiate 2 populations of cells.
He concluded that although mononuclear leukocytes did proliferate in and around
nervous tissue wounds, a population of cells originally present in the nervous
tissue must also have proliferated in response to the injury. Reznikov (1968)
reported almost no mitotic figures in glial cells after brain stab injury. He
observed that both the intact and injured hemispheres of the brains of ex-
perimentally injured mice had a greater number of labeled glial cells than did
uninjured animals.

Several investigators have dealt with neuroglial DNA synthesis in response
to peripheral or cranial nerve injury (Sjbstrand, 1965a, 1965b, 1966; Watson,
1965; Kreutzberg, 1966; Friede, 1967; Friede and Johnstone, 1967; Olsson
and Sj6strand, 1969; Adrian and Smothermon, 1970).

Sjbstrand (1965a, 1965b, 1966), Kreutzberg (1966) and Adrian and
Smothermon (1970) all demonstrated differences between the nuclei of intact or
damaged cranial nerves (either the facial or hypoglossal). In the nuclei of
damaged cranial nerves, an increase in mitotic activity of as much as 30 times
that of the intact nerves was observed 4 days after injury (Kreutzberg, 1966).
This acute reaction abated by the end of the first postoperative week (Sjéstrand,
1965a; Kreutzberg, 1966). The nuclei of most of the labeled cells resembled

microglia; however, Ssttrand (1965a) using Hortega's silver carbonate stain, found the

27
maximum response of microglia to be between the third and fifth postoperative
week. He demonstrated a "remarkable hypertrophy of astrocytes in the
regenerating nucleus" beginning on the third postoperative day by Cajal's gold
chloride stain. The maximum astrocytic reaction was observed at 9-14 days
postoperatively. After the third week, astrocytosis rapidly decreased and was
negligible by 90 days. No significant change was observed in the oligodendroglial
cell population. The heavy metal stains were not used on radioautographs in
his investigation. Olsson and Sjc'istrand (1969) found that leukocytes were able to
enter degenerating portions of peripheral nerves in Wallerian degeneration. Since
such cells were considered to be able to transform into macrophages, the authors
concluded that part of the macrophage population seen in degenerating nerves

was presumably hematogenous.

The homograft reaction

 

Within the past ten years, the rejection of allografts has been studied by
radioautography and also by electron microscopy. Walker and Goldman (1963),
using tritiated thymidine, traced labeled medium and large lymphoid cells from
the blood to an area of skin homograft and the underlying tissue. Some of the
cells were seen to invade the donor tissue, but most remained in the underlying
host tissue. The authors reported that many of these cells hypertrophied and
elongated. The cytoplasm of these cells developed strong, ribonuclease -labile,
basophilia. Large basophilic fibroblasts also proliferated in the area. In
electron microscopic studies of skin homograft and mammary tumor homograft
rejection, Walker et_a_l_. (1964, 1965) described the cytoplasm of hypertrophied
lymphocytes as containing many free ribosomes. Fibroblasts, on the other hand,
had cytoplasm which was almost completely filled with granular endoplasmic
reticulum. The authors suggested that the elongated, hypertrophied lymphocytes,
rather than macrophages, were responsible for homograft rejection.

Wiener, Spiro and Russell (1964) studied epidermal homograft rejection by
electron microscopy. They concluded that lymphoid cells were responsible for

graft rejection. These cells were termed graft rejection cells.

28
Smith and Walker (1967) studied mammary tumor allograft to the
cerebrum with radioautography. Although they were not able to demonstrate
cytoplasmic basophilia in the hematogenous cells surrounding the allograft,
medium and large lymphocytes were considered to be the causative agent in

rejection of the tissue.

METHODS AND MATERIALS

Animals

C3H strain mice obtained from Jackson Memorial Laboratory, Bar
Harbor, Maine were used as experimental animals. Adult animals weighing
23-28 g (average 25 g) received either stab injury or tumor transplantation.
C3H strain mice weighing 6 g were obtained from our colony. All mice
were maintained in plastic cages and received water and food ad libitum.
The food was Old Guilford Laboratory Animal Diet, purchased from the
Emery Morse Company, Guilford, Connecticut.

A/J strain mice with spontaneous mammary tumor were purchased from
Jackson Memorial Laboratory. The tumors were maintained by isologous
transplantation to the mammary area of other female A/J mice. Tumor de-

rived from transplantation was used for brain implantation.

Radioisotope

 

The radioisotope used to label the DNA in this experiment was
tritiated thymidine (thymidine -methyl -H3, 1 mc/ ml sterile water, specific
activity 2. 0 c/mM.) It was obtained from New England Nuclear Corporation,

Boston, Massachusetts.

Stab injury technique

 

The animals were anesthetized with sodium amytal (160 mg/ kg) given
subcutaneously in 5% solution. The hair over the scalp was clipped and the
skin swabbed with alcohol to remove any loose hair and debris. A midline
incision was made in the scalp and the animal secured to a mouse board with
the back up. The animal rested on small urethane -foam pads with holes cut
for the abdominal and mouth areas to facilitate breathing. The edges of the
incision were retracted to expose the right side of the calvaria, the underlying

29

30
connective tissue was cleared and the calvaria swabbed with alcohol. Under
a dissecting microscope, a small hole was made in the skull cap by cutting
out a piece of bone approximately 2mm x 2mm using a dental drill (Emesco
90N) with a milling bit. Care was taken not to drill too close to the sagittal
sinus or the lambdoidal suture. After the piece of bone was removed, a No. 11
scalpel blade was used to make a longitudinal incision 2mm x 0. 8 to 1mm deep
in the right cerebrum. The piece of bone was replaced and secured by rotating
it slightly. The scalp was closed using 9mm metal wound clips (Autoclips,

Clay-Adams, Inc. , N. Y.) and the animal allowed to recover.

Tumor transplantation technique

 

The host and donor animals were anesthetized with sodium amytal
(160 mg/ kg) given subcutaneously in 5% solution. The hair over the scalp
of the host animal and over the tumor area of the donor animals was clipped
and the skin was swabbed with alcohol to remove any loose hair and debris.
The donor animal was secured to a mouse board stomach side up. A midline
incision was made over the tumor area and the area was kept moist with a
gauze pad wetted with normal saline. A midline incision was made in the scalp
of the host animal and the animal was then secured to a mouse board as described
in the stab injury technique above. The edges of the incision were retracted to
expose the right side of the calvaria, the underlying connective tissue was
cleared and the calvaria swabbed with alcohol. A piece of bone approximately
2mm x 2mm was removed from the skull as previously described. After the
bone was removed, a No. 11 scalpel blade was used to remove a piece of cerebral
cortex approximately 2mm x 2mm x 0. 8 to 1mm deep. The meninges overlying
the small area were removed with the cortical material. At this time a small
piece of tumor was obtained from the donor animal and used to fill the defect
in the host animal. The piece of skull bone was replaced and secured by
rotating it slightly. The incision was closed as previously described and the

animals allowed to recover.

31

Intracardiac perfusion technique

 

All animals from which analysis of radioautographs were made were
perfused with 10% buffered formalin. A reservoir of 200 -250 m1 of fixative
was placed 80cm above the level of the animal. The perfusate was delivered
by means of a Cutter"28" Saf-T-Set (Cutter Laboratories, Chattanooga,
Tennessee) to which a disposable 18 gauge, 1 inch hypodermic needle was
attached. Perfusions were begun within 15 minutes of the onset of narcosis.
The animal was anesthetized with sodium amytal (160 mg/ kg) given sub-
cutaneously in 5% solution, and laid stomach side up on a suitable surface.

A transverse incision was made in the skin and abdominal muscles just
caudal to the xiphoid process. A curved hemostat was attached to the
xiphoid process and the diaphragm was incised from the lateral and ventral
portions of the rib cage, thus exposing the heart. At this time 0.03 ml of a
dilute heparin solution (1000 units / ml) was injected into the left ventricle
using a 1 ml disposable syringe with a 27 gauge disposable hypodermic needle.
The pericardium was stripped away and the rib cage incised longitudinally on
both sides just medial to the forelimbs. The chest plate was then lifted
rostrally and held by means of the weight of the hemostat. The dorsal aorta
was ligated at the level of the diaphragm. Grasping the right ventricle with a
small forceps, perfusion was begun by forcefully thrusting the needle of the
perfusion apparatus into the left ventricle. Care was taken to stop the thrust
of the needle while it was within the ventricular cavity. Five to 10 seconds
after maximum flow was established, the right atrium was incised to allow
escape of the returning blood and perfusate. Perfusion was maintained for
approximately 20-30 minutes, or until approximately 100 ml of the perfusate had
run through the animal. At the end of this time, the brain was removed and

placed in 10% buffered formalin.

Experimental design

 

All adult animals were given 3 subcutaneous injections of 100 JIC of
tritiated thymidine per injection at 4 hour intervals. Twenty -four hours after

the first injection, the animals received either stab injury or tumor transplantation

32

as previously described. The control animals received the same dosage of
radioisotope as the experimental animals, but did not receive any operative
treatment.

Eight adult animals with stab injury and 4 controls were killed by
perfusion with 10% buffered formalin at 3, 16, 32 and 64 days after injury
(2 experimental animals and l unoperated animal at each time period.) Two
adult animals with stab injury were killed 3 days after injury by perfusion with
Cajal's formalin -ammonium bromide fixative (FAB) by the intracardiac perfusion
technique described above. The brains were removed and put in FAB. Two adult
animals with tumor transplantation and l unoperated animal were killed by per-
fusion with 10% buffered formalin at 8 days after transplantation. Juvenile
animals weighing 6 g were given 6 subcutaneous injections of 30 )1c of
tritiated thymidine per injection at 4 hour intervals. Sixty days after injection,
when the animals weighed 19-23 g, 2 received tumor transplants as previously
described. Eight days later the 2 experimental animals and 1 unoperated animal

were killed by perfusion with 10% buffered formalin.

Histologic technique

 

A. Processing of tissues

After 1 to 4 weeks of fixation in 10% buffered formalin, the brains were
removed from the fixative and transected through the injured area, dividing
them into a rostral portion, which was used for the collection of frozen sections,
and a caudal portion, which was embedded in paraffin.

Transverse sections, 20 )1 in thickness, were cut on a freezing microtome.
Ten alternating serial sections were collected, rinsed twice in double distilled
water and mounted on cleaned albuminized slides. After drying in a 37°C oven
for 36-48 hours, the sections were coated with Kodak NTB2 nuclear emulsion
according to the method of Walker (1959), omitting celloidin coating and subbing
fluid from the procedure. The radioautographs were exposed for 8-12 weeks
at 5°C, developed 5 minutes in Microdol x and fixed 5 minutes in acid fixer.
They were then stained with hematoxylin and eosin (H and E) as described by
Walker (1959), dehydrated in a series of graded alcohols, cleared in cedarwood

oil and mounted in Permount.

33

Sections to be stained for neuroglia were collected in 10% ammonium
hydroxide solution according to the Globus Modification of Cajal's gold chloride
method (Penfield and Cone, 1961). After 24 hours in ammonium hydroxide, the
sections were washed 3 times in double distilled water and then placed in 10%
hydrobromic acid for 2 hours. This allowed the ammonium bromide to act as
a mordant in the formalin fixed material. The stain for which formalin-
ammonium bromide fixation was required could then be carried out. The sections
were rinsed twice in double distilled water and then stained by Cajal's gold
chloride sublimate method for astrocytes. A few sections from each injury
area were stained for oligodendroglia and microglia as follows. The sections
were collected in double distilled water and then pretreated according to the
method of Wolman (1958): 20 minutes in 5% hydrochloric acid, rinsed in
distilled water, placed in 0. DIN sodium hydroxide solution for 4 hours. They
were then stained by Penfield's Second Midification of del Rio-Hortega's Silver
Carbonate Method. The stain gave inconsistent results and was not used for
experimental determinations.

After staining, the sections were mounted on cleaned, albuminized slides.
They were then dehydrated in alcohol, cleared in cedarwood oil and coverslipped
with Histoclad.

Sections, 7 )1 in thickness, were cut from the paraffin embedded material.
These were put on cleaned, albuminized slides, deparaffinized and then coated
with Kodak NTBZ nuclear emulsion according to the method of Walker (1959).
Celloidin coating and dipping in subbing fluid were omitted from the procedure.
The radioautographs were exposed 5 weeks at 50C, developed 5 minutes in
Microdol X and fixed 5 minutes in acid fixer. The slides were then stained with
H and E, dehydrated in a series of graded alcohols, cleared in cedarwood oil
and mounted in Permount. After drying, the slides were cleaned and used for
analysis.

The brains that were perfused and fixed with FAB were treated somewhat
differently than those described above. After 1 week of fixation, the brains
were removed from the fixative. Frozen sections, 20 y. in thickness, were cut

on a freezing microtome and 10 sections were collected for combined radioautographic

34
processing and heavy metal staining for neuroglia. Five sections were rinsed
twice in double distilled water and then stained by Cajal's gold chloride
sublimate method for astrocytes (Penfield and Cone, 1961) and mounted on
cleaned, albuminized slides. After drying in a 370C oven for 36-48 hours, the
sections were coated with emulsion. The slides were exposed for 1 week,
developed, and fixed. The remaining 5 sections were rinsed twice in double
distilled water and mounted on cleaned, albuminized slides. After radioauto-
graphic processing, as described above, developing and fixing, an attempt was
made to poststain the sections by placing them in Cajal's gold chloride sublimate
staining solution.

Through the rest of the injured area, alternating serial sections, 20 )Lthick,
were collected for radioautographic processing or gold staining. Those sections
to be processed for radioautography were treated in the same manner as the
alternating serial sections from material fixed in 10% buffered formalin. The
slides were developed and fixed after either 4, 8, 12 or 14 weeks exposure at
50C. They were then stained with H and E, dehydrated, cleared and mounted
in Permount. The alternate sections were rinsed twice in double distilled water
and stained by Cajal's gold chloride sublimate stain, as above.

B. Cell Classification

Radioautographs stained with H and E were used to count and to classify
nuclear types and to count silver grains. A nucleus which had at least 4 silver
grains over it was considered radioactive. Nuclear counts were made with an
A/O Spencer microscope using a 97X oil immersion objective and a 10X ocular
lens which was fitted with a 1421 A Net Reticule (20mm dia. 15mm sq. ruled
into 1. 0mm sq's - American Optical Company, Buffalo, New York). The reticule
measured 50 )1. on each side when using the above optical system, and areas of
2500 )1? were utilized for counts. Nuclei were counted in gray matter along the
edges of the stab injury or tumor and a corresponding area in the uninjured
hemisphere of operated anirnals, in the corpus callosum of injured and
uninjured hemispheres of operated animals, and in corresponding areas of
unoperated animals. Nuclear counts were made in areas of 90, 000 )12 (0. 09mm2)

2
in the corpus callosum of each operated hemisphere. Areas of 40, 000 )L2 (0.04mm )

35
were counted in the corpus callosum of the uninjured hemisphere of operated
animals and in unoperated animals. Counts were made in the gray matter
along the edges of the wounds or tumors in an area 50;; wide, or a total
area of 40, 000 )3. Areas of 40, 000)}:2 were counted in corresponding gray
matter of the uninjured hemisphere of the operated animals and in unoperated
animals.

The nuclei of all cells other than neurons, ependymal cells, endothelial
cells and polymorphonuclear leukocytes were counted. By excluding these cells,
it was felt that the cells counted were glial cells and mononuclear cells which
could not be distinguished from the glial cells.

The nuclear types were classified in the following morphological groups:

1) small, dark-staining, with nuclear diameter less than 4-5_)1.; 2) medium round,
with nuclear diameter which was 5-9 )1; 3) medium-sized elongated nuclei, whose
smallest nuclear diameter was 5-9_)1.; 4) large, pale, oval-to-round nuclei, with
a diameter of greater than 9/11. and which were lighter-staining than other nuclei.
In addition, there were some nuclei which were entered into an unclassified
category because they did not fall into the above categories.

Gold stained sections 20 min thickness were scanned using a 43X objective
lens and a 10X ocular lens fitted with a 1421 A Net Reticule which measured
0. 0169mm2 in area with this optical system. The 43X objective lens was
used to facilitate viewing the perikaryon and processes of the gold stained cells.
At a higher power the cells appeared "grainy" and were difficult to define. Five
consecutive areas of 0. 0169mm2 (0.0845mm2 total) were scanned in the corpus
callosum. Counts of astrocytes were not made in cortical gray matter due to the
lack of stained cells in this area.

Two cell categories (Figures 6 and 7) were used: 1) large, well-stained cells
with thick processes and 2) small, more lightly-stained cells with more delicate
looking processes. Cell bodies which lacked definite processes and processes
which were not seen to extend from a definite cell body were not classified or
counted.

To determine the frequency of large gold stained cells in 0. 0845mm2 of

corpus callosum, 20 )Lthick sections stained with H and E were used to count the

36
total number of nuclei in O. 0845mm2. The frequency of large gold stained cells

was then expressed as a ratio of these cells per total nuclei in 0. 0845mm2.

RESULTS

A. Unoperated animals and the uninjured hemisphere or operated animals

On gross examination, the brains of all unoperated animals appeared
normal. There was no evidence of infection around the scalp wound of injured
animals. Transverse sections of these brains were scanned and their architecture
compared with reference plates in an atlas of the mouse brain (Kovak and Denk,
1968). Gray matter and white matter could readily be distinguished in sections
stained with H and E, and with Cajal's gold chloride sublimate stain. It was
also possible to identify the corpus callosum and hippocampus as discrete
structural entities with these stains.

Sections which were stained with gold chloride sublimate showed a pre-
dominance of small, lightly-stained cell bodies with lightly -stained processes
(Figure 6). There were from 0 to 3 large gold stained cells per 0. 0845mm2
of the corpus callosum of unoperated hemispheres, whereas there was an average
of 30 small, lightly-stained cells counted in the same area. The corpus callosum
of the uninjured hemisphere from operated animals demonstrated a similar
preponderance of small cells (average: 33) over large cells (average: 2) per
0. O845mm2. In 0. 0845mm2 of the corpus callosum of unoperated hemispheres,

the large gold stained cell population was less than 1. 5 percent.

Analysis of rad ioautographs

 

Nuclear counts were made in the cortical gray matter and the corpus
callosum of unoperated animals and in the uninjured hemisphere of operated
animals injected with tritiated thymidine 1 day before operation. Table 1
shows the data obtained from these counts. These data are also represented
in the bar chart of Figure 1. In addition, 2 animals were injected with tritiated
thymidine 60 days before tumor transplantation. The data obtained in counts
from the uninjured areas of these individuals are listed in Table 2 and are

37

38
represented in the bar chart of Figure 1. An unoperated animal injected at the
same time interval showed numbers of labeled cells which correlated with

counts made previously on unoperated animals of this age (Smith and Walker,

1967) and these data are not included in Table 2 or Figure l.

39

coHuwucmHchmH. SEPT.

 

 

m .N HV omH o .m N wn Hofifiou
o .o H mm H o o cm a
m .o H in H, .v m we a we
0 o mNH o .H H mo Hobcou
m .H m om H H .m H we a
H .H m E: o o co m mm
o o me o 0 mo Hobcou
HV .H N NHVH H .m m we a
o .m N 03 o 0 mm m 0H
m .m o How m .m m we 3.350
m .N Hu ovH N .w v 3. «m w
H .H N wwH o 0 mm Hoficoo
.\. .H m omH HV .H H mm n
H“ .m m now 0 c on m m
“63mm; 33:: 333 we 4 anoan HoHoHE HoHosc Ho Amzmcvcotmuoao
Eocuom ozuoaoHomH Ho .8956 H38. “coouonH min—omega“ Ho H385: H809. note 25B
.8883: HBOH Hones: H309.
83028 35.80 uoBmE Bram H8380 HeEHc<

 

.mHmPHouH—H o8: coumcmeoU 9: Ho some “a 8:3 new
soHumEmHamcmfi .383 no BS2: 83m 383 muse: vm wEHEHwon mHmtoufi
use: H, um moEHu m 0528?: BEETS Ho oéooH EH? 8835 6.83 2953.

2955» coumuoao Ho mmonm BET—Ha: Ho 583 .o 5. 33:: poHoan Ho 5:25on

N
H 2an

4O

 

 

m .w wH mom m. .v m on c
o .0 0H 00H H. .o 5 NH. w
UoHonH 63:8 63:: poHonHeH H208 HoHosa
883m ozuomoHceH Ho Ho “coouom o>HuomoHHcmu Ho Ho
H383: HSorH. poets: H309 HoHED: H309 .3988 H309
EsmoHHm0 98.80 p.888 8% H3880 Hm88<

 

.coHueHoao Eta when w HooHHHvH
whoa >651 80338383“ .883 oHoHon mzmc o0 28888 .80: H. um
88: o @8283“ con—SHE Ho Snow 8:? 838.28 one? 2988... 2895—.

823883 @3338: 65 8. 83828 9.980 o5 Ho new

$3.88 mam HmoHuHoo o5 Ho N883 .o :H HoHoHE @3693 Ho 5:26on

N oHnHmh.

41

Figure l

2
Percent labeled nuclei in 0. 04mm of the cortical gray matter and of the
corpus callosum of unoperated animals and of the uninjured hemisphere of
operated animals.

The animals were injected with 100 )1c of tritiated thymidine 3 times at 4
hour intervals beginning 24 hours before operation. Each pair of bars re-
presents a different mouse. The last pair at each time interval is from an
unoperated mouse unless designated otherwise. Animals designated 8*
received 30)1c of tritiated thymidine 6 times at 4 hour intervals 60 days
before tumor transplantation.

 

'
.
OI
'
o
m
y
n
G

 

7/////,

u
u
I
c
.
u
M
a...
..\
x.\ \
,\
\\ \
\
.

y////////////////%

y///////////////////////n

////////////fl//////¢///////////

 

,//////,////////4

7/////////////////a

 

   

 

IOT

a4

. h 4 r .
12... 3.22 2.88..

 

Time interval (daycldhr operation

FIGURE

43
B. Operated animals

Description of the wound

 

In all groups of animals injured by stabbing, the wound was found to pass
through the cortical gray matter to the corpus callosum. In a few animals,
the wounds extended deeper and were found to pass through the corpus callosum
into the white matter of the hippocampus. All stab wounds contained necrotic
debris. Those which extended into the white matter of the corpus callosum
and hippocampus retained their hemorrhagic character throughout the length
of the experiment. Some wounds had longitudinal expansions which ran parallel
with the fibers of the corpus callosum. These expansions contained necrotic
debris.

In animals which received mammary tumor allograft, the tissue mass
was seen to lie in the cortical gray matter with its deepest portion abutting or
just interrupting the white matter of the corpus callosum. These wounds did
not appear to retain as much hemorrhagic material as did the simple stab wounds.

In gold stained sections, the injured hemispheres of experimental animals
showed a predominance of large, round to somewhat rectangular cell bodies
(Figure 7). The processes of these cells were darker, more heavily stained,
and appeared thicker than those of the large gold stained cells seen in the
uninjured hemisphere.

In 2 animals sacrificed 3 days after stab injury, a large collection of
erythrocytes and necrotic debris occupied the central part of the lesion. In
sections stained with H and E, many dark-staining pyknotic nuclei were present
in and around the wound. Polymorphonuclear leukocytes were rarely seen.
Mitotic figures were found in both the gray and white matter near the lesion
at this time. Gold stained sections showed large cells in the corpus callosum
around the lesion and its expansions (figure 8).

Four animals received tumor allografts to the brain. Two animals were
injected with tritiated thymidine 1 day before transplantation and 2 were injected
with tritiated thymidine 60 days before transplantation. Eight days after grafting
the tumors to the brain, the animals were killed. The 2 animals injected 1 day

before transplantation will be described first.

44

In one individual, on sections which were stained with H and E, the
tumor appeared necrotic in certain areas. Many cells with medium -round nuclei
were seen within and surrounding the graft. Figure 9 shows labeled glial -like
nuclei in the tissue surrounding the tumor allograft. Figure 10 shows radioactive
medium -round and medium -elongated nuclei in the cortex within 100 )1 of the
tumor allograft. Polymorphonuclear leukocytes and mitotic figures were not
seen at this time or at subsequent time intervals. Hematoxylin and eosin
stained sections from the second individual did not reveal any allografted
material. It appeared that when the brain of this individual was transected
prior to paraffin embedding, the transverse cut was made just caudal to the
allograft rather than through it. For this reason, analyses of radioautographs
were not made on this individual, and data for the tumor allograft reaction
were collected from only 1 individual. Gold stained sections from both individuals
with tumors were obtained however. These sections, as well as those from the
other animals which received tumor allograft, showed many large, well-stained
cell bodies with thick processes. They were located primarily in the white matter
of the corpus callosum and hippocampus. Some of the cells could be seen in the
gray matter surrounding the unstained tumor mass, but these were not as well
stained, nor did their processes appear to be as large, as those found in the
white matter (Figures 11 and 12). Hematoxylin and eosin stained sections from
the animals injected with tritiated thymidine 60 days before tumor allograft to
the brain also showed the tumor to be necrotic in certain areas. Many cells
with medium -round and medium -elongated nuclei were seen within and surrounding
the graft. Polymorphonuclear leukocytes were not seen. One mitotic figure was
noted in 1 of the animals.

Sixteen days after stab injury, necrotic debris was still present in the
central part of the lesion toward the lower end of the wound track. Many large,
pale to medium -dark nucleated cells appeared to fill the upper portions of the
track. These cells seemed to be contiguous with the cells of the meninges. Gold
stained sections showed many large well -stained cell bodies with large processes
in the white matter of the corpus callosum and hippocampus. A few of these
cells were observed around the lesion in the gray matter adjacent to the corpus

callosum (Figure 13).

45
At 32 and 64 days after stab injury, H and E stained sections showed cells
with medium -round nuclei, medium -elongated nuclei and large, pale nuclei
present in and around the wound. Small dark nucleated cells containing
phagocytosed material were present in expansions of the wound track in the
corpus callosum. Gold stained sections looked similar to those of the animals

sacrificed 16 days after injury (Figures 14 and 15).

Analysis of gold stained sections

 

Table 3 shows the number of large, well-stained cells in the corpus
callosum around the lesion area for all experimental animals injected with tritiated
thymidine 1 day before injury. The total number of nuclei in the same area
were counted in 20 )LthiCk sections stained with H and B. These data are also
presented in the bar chart, Figure 2. The 3C2 statistic as described by Croxton
(1953) was used to test the hypothesis that the ratio of the number of large
well-stained cells to the total number of nuclei in 0.0845mm2 of corpus callosum
did not change significantly with the type of injury or with the time of sacrifice
after injury. Tables 4 and 5 list the X2 and p values calculated from these data
for animals within the same treatment groups, and for groups killed at different
time intervals after stab injury or tumor transplantation. No significant
differences were seen between the 2 individuals killed at each of the designated
time intervals or between the groups killed at various time intervals after stab
injury or tumor transplantation. It should be noted here that the designations
"a" and "b" were arbitrarily assigned to the individuals within a group. These
were used to facilitate the collection of data from individuals and the computation

of the 1C2 and p values.

46

:oHHmH::HQm:mH.H 885....

 

e .o m we com a
n .H H v mm 3% c He
m .o m em nwm n
m .H H om we mom m mm
m .w m mm wmm :
m .o nH mm wem m eH
m .H H NH we. owe ..b
e .OH H we 3m .8 w
m .0 2 mm wem n
v .H H mH em eH m m. m
mHHoo mHHoo mHHoo HoHoHE Hen—out Hm8H:< $83
H888? Bow eofimumubucwz 888m- Bow :oHuwuoao
omumH 888m .HHw8m embed not: oEHH

 

8033:3885 .888 .8 33.8 swam
8t: 2:888 o8: eoSFmeoe on... we coco u: 8on o8? mHmEH:<

68:82:88 H8888 65 Ho 83828 95.80 Ho 88mwwo .o :H mHHoo 88:8 Bow Ho 8:02Hvoum

N

m 03:5

47

Figure 2

Percent large gold stained cells in 0. 0845mm2 and radioactive nuclei in
0. 09mm2 of the corpus callosum of the operated hemisphere.

Animals were injected with 100 }1c of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours before operation. Each pair of bars represents

a different mouse. Animals designated 8* received 30 )1c of tritiated thymidine
6 times at 4 hour intervals 60 days before tumor transplantation.

54!

521

.JI.
28i—

26

24-

- Radioactive nuclei

22“

W Large gold stained cells

20-1

IB‘

I6-

I44

              
 

 

,//////////////////////////////////////////////////////////%

22%

gig:

7///////////// //x/// ///////////////////////// //////////.////////

y////////////////////////////////////////fl////////////// «x

%///////////////////////////%/////////////////////

/////////////////////////////// ////////////////////////////////// a

iii???
%////////////////////////////////////////////////////////28

 

/./////// /é//////////////////////// // ///////////////////.///////4

 

gig/88%:

1 d d

d 1
2 O 8 6 4

2.
O

3.2:. ozone. 2.2:.

Time interval (doyeI after operation

FIGURE 2

Statistical results for gold stained cells

3C2 and p values of differences between the two individuals killed
at each of the designated time intervals after stab injury or tumor

 

 

transplantation.
Time after operation (days) 3C2 value p value
3 O. 6467 > . 30
8 0. 0783 > . 75
16 0. 1998 > . 50
32 1. 1567 > . 25
64 1. 1071 > .25

Statistical results for gold stained cells

3C2 and p values of differences between the groups killed at each
of the designated time intervals after stab injury or tumor trans-

 

 

plantation.

Time after operation (days) 3C2 value p value
3a, 8a 0.1192 >.70
3b, 8b 0.5934 >.30
8a, 16a 0.2407 >.50
8b, 16b 1.4377 >.20
16a, 32a 0.9079 >.30
16b, 32b 0.1276 >.70
32a, 64a 0.0147 >.90
32b, 64b 0.0225 >.80

Gold stained sections from 2 additional animals which received tumor
allografts were also scanned for large gold stained cells. These animals had
received different pretreatment (injection of tritiated thymidine 60 days before
allografting) than the animals listed in Table 3. In both groups the time interval
between allografting and killing was 8 days. The ratios of large gold stained
cells to the total number of nuclei in 0. 0845mm2 of corpus callosum were 47 per
433 (10. 8%) and 53 per 452 (ll. 7%). No significant differences were seen between
the individuals of this group (p >. 80), or between this group and the animals listed
in Table 3 (p >. 50).

50

Analysis of radioautographs

 

2
Table 6 shows the total number of radioactive nuclei counted in 0. 09mm

of the corpus callosum around the lesion for all experimental animals injected

1 day before injury. These data are presented as a bar chart in Figure 2. The
X2 statistic showed that in 1 case there were significant differences between the
2 individuals that were killed at each time interval and in 3 instances between the
groups that were killed at different time intervals after stab injury or tumor
transplantation. Tables 7 and 8 list the X2 and p values calculated from these

data.

Table 6

Frequency of radioactive nuclei in 0. 09mm2 of the corpus callosum
of the operated hemisphere

Animals were injected with 100 )1c of tritiated thymidine 3
times at 4 hour intervals beginning 24 hours prior to operation.
They were killed at each of the designated time intervals after
stab injury or tumor transplantation.

#1

 

Time after Total number
operation Total number of radioactive Percent
(days) Animal of nuclei nuclei labeled
3 a 246 67 27. 3
b 278 70 25. 2
8 * 512 268 52. 5
16 a 328 28 8. 5
b 373 30 8. 0
32 a 535 85 15. 9
b 568 57 10. O
64 a 612 89 14. 5
b 563 64 1 1. 4

*Tumor transplantation

DC? and p values of differences between the two individuals

5 1
Table 7

Statistical results for radioactive nuclei

killed at each of the designated time intervals after stab injury

or tumor transplantation.

 

2
X value

 

Time after operation (days) p value
3 0. 2740 > . 50
16 0. 0561 > . 80

32 10. 9674 < . 001
64 2. 6097 > . 10

Table 8

Statistical results for radioactive nuclei

3C2 and p values of differences between the groups killed at each
of the designated time intervals after stab injury or tumor trans-

plantation.

Time after operation (days)

A

33, 8
8, 16a
16a, 32a
16b, 32b
32a, 64a
32b, 64b

3C2 value?

42.47+
168.1+
11.7729
1.0651
1.0665
0.5254

p value

<.001
<.001
(.001
>.30
>.30
>.30

Table 9 lists the number of each nuclear type counted in 0. 09mm2 of

corpus callosum of the operated hemisphere in animals that were injected 1 day

before injury. These data are presented as a bar chart in Figure 3. Tables 10

and 11 give the 3C2 and p values calculated from these data. Significant differences

were found in each nuclear category between the 2 individuals killed at each of

the designated time intervals and between the groups killed at each of the

designated time intervals after stab injury or tumor allograft.

52

830803085 .8821?.

 

H ww eH w o mem n
e mw w wm o NHe m we
w 0H mN NH w wee n
o Hw mN wH H mmm m Nm
H NH m OH O mum. n
H w n mH o Nm 0 eH
NH NH nw NwH mm NHm a. m
N o 0N N 3 wnN n
e o HN mm m ewN m m
8888000 0:20.: 088N008 8H8: H0 888w. A083
1 0: 1 :80 88082 0:50.: 88082 8.80 HHmEm 88:: H809. :88800
.200 .omuwd 80.8 885.

 

masouwndm 820:2

 

888800 88m. 838:8 0:8: 038980 85 H0 :08 :0 08:8 083 85:.
8033:8885 .883 .8 8:88 830 0: :88 0.80: N 888on 28.808
.80: w 8 0883 m 088810.: 00:88:: hH0 3:2: :83 03038 0.83 80:89.44

0880880: 030800 85 8 8:00:00 0:980 H0

N880o .o 8 0000.886 .8200: 8.8000800: H0 888on

o 3an

53

Figure 3

2
Percent labeled nuclei of each nuclear type in 0.09mm of corpus callosum
of the operated hemisphere.

Animals were injected with 100 )1c of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours prior to stab injury or tumor transplantation.
Included for comparison are the percent gold stained cells in 0. 0845mm2
of corpus callosum of these animals.

Indicates range

“MM... 7//////////////////////////////////////////////////////////
“MM... %./////////////////////////////////////////////////////fl
MV/////////////////////////////////////////////////////%

w gig/ii
..M........ V///////////////////////////////////////////.//./////////////

0
Gold stained cells

      

    

4...... //////

    

32“

7//////////////.
z//////////////1

      

           

JI.
3 0
Large, pole,

 

   

[III/l.

H ///////////////////////////
2////////.2
¢/////////%////////////////////////////////////////////%

    
 
    

   

SOIOSZ“

Hedium

       
 

 

       

”a... 2////////////. m m
gig/88%;??? o m
/////////////////////////////////////////////////////////////. s «

     
 

3204
d

 

M 2 O 8 6 4 2 O

oval-to - round

elongated

FIGURE 3

 

 

 

 

3. 3.3:...

55

.930 m :0 8:0: 208:5
0:: 8 0080000 88:: 0::0.:10:rH0>0 .08: .0wH0H :83 0:00 0: 0.83 0.8:: 00:0003 00:08:80 :0: 0.83 0030 >...

 

 

mm. A eowod Ho. V mew.\. Hoo. V Heom.HN we
moo . V NweH .0 cm . A ewmm .0 ON . A wwNe .H Nm
mo . A wNNH .m ow . A Como .o 3 . A oeHw .H eH
.. .. cm . A omwm .o no . A wemw .N m.
0:H0> : 0:H0> Wm 0:H0> : 0:H0> man 030.» a 030.» NWn
88:: 080:. 0: 88:: 88:: 00008030800
18.8 .08: .0mH0H 00:0w:0H0 8800—2 080.: 88002 .880 08:.

 

.:0H:0::0H00:0.:: 88:: .8 838 00:0 8::0 0H0>.8:8 08: 0080880 0:: ”H0

800 :0 00:8 0H0:0H>H08 0.3: 0:: :0080: 08:08:80 :0 00:H0> a 0:0 Nufi

080880 808:: 0.880800: .8: 08:00.: H00H:0H:0:m

OH 030p.

fililull l'l‘ III‘IIIIIII.‘ II.IIH‘II1‘.‘I.I.IVU

- pv-A~:._.

56

.0000 m :0 00:20: 08880
00: :: 00>.8000 88:: 0::0:- 0:-:0>0 .0:00 .0w:0: 0::3 0::00 0: 08.5 0.80: 00:0000 00:0::0:00 :0: 0.83 0030?.

 

 

moo . v we: .0: o: . A :000 .: mo . v 0000 .m 000 .0mm
mo . A. 0000 .o :oo . v 0500 .0: om . A 0000 .m 000 .0mm
00. A 3:06 00. A 003.0 om.A 00:05 0mm .00:
:oo . V 053 .0: 0o . v 030 .m. on . A 0000 .0 0mm .00:

cm . A :mmm .: :oo . V 03:. .0: :oo .. V +50 .0: 00: .w

... ... SKA mvwod :oo.v +3.0: 0 .00

0:4:0> 0 0::0> Nun 0::0> 0 0::0> Nun 0:8.» 0 0::0> Nun

88:: 0:50.? 0: 88:: 88:: 80008039800
..:0>0 .0:00 .0mn0u: 00:0w:0:0 8300:): 0::0: 8::002 8::0 08C.

 

.:0::0::0:00:0:: 88:: :0 0.8.3: 00:0 8::0 0:006:08: 08:: 00:0:w800
00: :0 0000 :0 00:3 00:08 00: 802500 000:08:::0 :0 00::0> 0 0:0 Nun

00:08:80 :00:0:: 0>::000:00: :0: 0:800: :00::0::0:m

:: 0:000.

57

Table 12 lists the frequency of labeled nuclei in 0. 04mm2 of cortical
gray matter along the wound in animals injected with tritiated thym idine 1 day
before operation. These data are presented as a bar chart in Figure 4. This
figure also includes data obtained from the corpus callosum which are listed in
Table 6. The data for the nuclear subgroups listed in Table 12 are represented
as a bar chart in Figure 5.

Table 13 lists data obtained from counts of radioautographs of 2
individuals injected with tritiated thymidine 60 days before tumor transplantation.
These data are included as a bar chart in Figure 4. In the corpus callosum and
the cortical gray matter along the wound, some of the cells of the medium -round
and medium-elongated nuclear subgroups were labeled. No large, pale, oval-
to-round radioactive nuclei were found in the corpus callosum. In the corpus
callosum of 1 of the animals, 3 labeled nuclei were found in the area subjacent
to the tumor. Two labeled medium-round nuclei and an unlabeled mitotic figure
were seen in the gray matter immediately above the corpus callosum and
within 200 )1. of the wound. The sections from the other animal did not demonstrate
labeled nuclei within 200 )1. of the wound either in the gray matter or the corpus
callosum. Most of the labeled cells counted in the corpus callosum were found

at distances greater than 250 )1. from the wound.

58

:0::0::0:00:0.:: .88: F...

 

0:V mm m m 0 .om mm 00m 0

m : m: 0 o m .0: mm :0: 0 00

0: v m: o 0 . : m :0 00: 0

w o: :: o m . : m 0 N 0m : 0 mm

:: m o: : N .0m mm mo: 0

: : n 0: v m .9: m0 0m: 0 0:

n 0: mm : 0 .mn vm N0 ... w

o m: m : o w .om mm .3 0

o N. w : m o . : m hm mm 0 m
080:- 0:- 8.6 00:00:08 080.: 0:80 00:000: 88 :: 8:0 :: :08::< :0K00w

.080 .0881: 8:: 00:): 8800:): 0080 800.80 0:800:00: :0 :0 :0::0.800
008.8080 .808 :7: 808:: :0:0.H 808:: :0:0.:. 8::0 08: E

 

.:0::0.800 .88 0:0?8::: 08:: 00:08:80 00: :0 00::::: 0.83 000:.
.:0::0::0:00:0H: 88:: .8 0.8.8: 00:0 0: 8:00 0.800 wm w:::::w00 0:0?88:
.800 v :0 008:: m 0:00:88: 00:0:::.:: :0 Bfioo: 0::3 00:00.8: 0.83 0888::

0.8000: 800 00:0800 00: :: 8::08 00.0w :00::.:00 :0 «8800 .o :: 88:: 03:80:00.: :0 88:00:”:

N: 0:009.

59

 

 

m ..m 0: 00:: m .: N 00: 0
m .0 0m mom 0 .m m 00: 0
00800: 88:: 88:: 00800: 88:: 88::
800.80 03:08:00.: :0 800.80 03:08:00.: :0 :0
:0 808:: 8:00. 808:: 8:00. 808:: 8:00. 808:: 8:00.
8:00:80 0:0.80 00:0 80::80 8880.

 

.:0::0::0:00:0.:: :88 0000 w 00:::: 0.83 0000. .:0::
-0::0:00:0.:: 88:: 0: .8:.:0 0000 00 0:0?8::: .800 v :0 008:: 0
080:800: 00:0:::.:: :0 0:000 0::3 00:00.5: 0.83 08880 0:808:

08:00:80: 8:088 00: :: 8:00:80 0:080 00: :0 0:0 00:03 00: 0:08

.8::08 00:0 808.80 :0 N88:5 .o :: 88:: 0>::000:00.: :0 00:03:05.:

m: 0:80.

60

Figure 4

2
Percent radioactive nuclei in 0. 04mm of the cortical gray matter along the
wound and in 0. 09mm2 of corpus callosum in the operated hemisphere.

Animals were injected with 100 Jim of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours before stab injury or tumor transplantation.
Each pair of bars represents a different mouse. Animals designated 8*
received 30 )1C of tritiated thymidine 6 time s at 4 hour intervals 60 days
before tumor transplantation.

801

70: l_

a:

221%,;

manor

V///////////////////////////////////////////////////////A

gig/lg?

mgggggggflg

-Corpus callosum
W Grey

igggggggggg

 

wgggggggg

gig

d
0 m w 2 m 0

1r—
.4
4

4
O
.202. 02000. 20200

 

   

 

d J 4 d

Tune interval (days) of?" operation

FIGURE 4

62

Figure 5

Percent radioactive nuclei of each nuclear type in 0. 04mm2 of cortical gray
matter along the wound in the operated hemisphere.

Animals were injected with 100 )1C of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours prior to stab injury or tumor transplantation.

 

00 0//////////////////////////////////////////,
00 ///////////////////////////////////////////////////////////0

00000000/////////////////0
0 0////////////////////////////0

Indicates ronge

 

D L

7//////////////////////////////////////////////0
00000000888888
% ig/Zgggggggggg
. 4. _ :

 

22..

2H2... 05.00063 2.3030 000.34

:0 m. 6 4 2 o a 6 4 2 o
2 l

3 O I63264

Large, pale,

84 3 8 IS 32 64

30|632

Medium tound

Medium

oval-to-rauna

elongated

FIGURE 5

64
Mitotic figures

 

A search was made for mitotic figures at each time interval. In the
animals injected with tritiated thymidine 1 day before injury, mitotic figures
were seen only in the sections from animals killed 3 days after stab injury.
Thirteen mitotic figures were counted in 0. 270mm2 of corpus callosum
(0.045mm2 in each of 6 sections). There were 3 prophases, 8 metaphases
and 2 anaphases. All of the figures were considered radioactive except 1
prophase which had 3 silver grains overlying it. Eight mitotic figures were
seen in 0. 110mm2 of corpus callosum (0. 0275mm2 in each of 4 sections).

There were 2 prophases, 4 metaphases and 2 anaphases. Of these, 1 prophase,
l metaphase and l anaphase were not radioactive. One mitotic figure in
metaphase was seen in 0. 04mm2 of cortical gray matter of 1 animal injected

60 days before tumor transplantation. It was not radioactive.

Radioautographs of 20 )1. thick sections showed various degrees of labeling.
Some sections appeared devoid of labeled cells in the tissue surrounding stab
injury or tumor transplantation. Other sections showed labeled nuclei, but
none of the sections showed lesioned areas with heavily labeled cells (>20 grains
per nucleus). No quantitation of labeled nuclei was done on the slides from
any of the animals because of the variability of the technique. These sections
were used, rather, to count the total number of nuclei per 0. 0845mm2 of the

corpus callosum (Table 3).

Formalin-ammonium bromide perfused and fixed material

 

Five slides which were prestained with gold chloride sublimate were viewed
under the microscope before radioautographic processing. Large gold stained
cells were seen in the corpus callosum of the injured hemisphere. After the
emulsion was developed and fixed, the emulsion over the entire surface of the
slides appeared dis colored, the sections had turned a dark brown color and the
cells could not be distinguished from the background. The slides that were
processed for radioautography before staining with gold discolored when they
were placed in the gold bath, and neither silver grains nor cells could be

demonstrated.

65
Large, well-stained cells were demonstrated in gold stained sections
collected in the alternating serial section method. These cells were found
in the corpus callosum of the injured hemisphere. The alternating sections
which were processed for radioautography demonstrated no silver grains over

the tissue up to and including 14 weeks of exposure at 50C.

66

Figure 6

Three small, lightly-stained astrocytic cell bodies are seen in the center
of the photomicrograph. Small pieces of fibrous material and cell bodies
which did not appear to have processes were not counted. Unoperated
animals and the uninjured hemisphere of operated animals showed a pre-
dominance of these cells. Cajal's gold chloride sublimate stain for

astrocytes. X 640.

Figure 7

Large, well-stained astrocytic cell bodies are demonstrated in the upper
left quadrant of the photomicrograph. The processes of these cells appear
thicker and more heavily stained than those of the small astrocytic cell
bodies. Fragments of fibrous material and cell bodies which did not

appear to have processes were not counted. A predominance of large, well-
stained cells was found in the corpus callosum of operated hemispheres.
Cajal's gold chloride sublimate stain for astrocytes. X 640.

 

Figure 7. Large gold stained cells in operated hemisphere

68

Figure 8

Large, well-stained cells are located in the corpus callosum and
hippocampus, in the lower one-third of the photomicrograph. The
lesion is located at the upper left and cortical gray matter occupies
the remaining upper two-thirds of the photomicrograph. Cajal's
gold chloride sublimate stain for astrocytes. X 280.

 

-‘ o o ‘ t f c' . ,
é. '5’ ‘ ‘ Y \‘r’af” .‘. ”r
. . , ‘ ‘ s t. a. ’2'). .‘ . .
' 2 ’ W " V”,
‘ fl . e ...- ‘0' ,0 : 8
4:. rm - 3:0»... _ .
fig):- (‘8‘. ‘ . . ‘fi“ '. Jig. . d it; .-, .
.. . ' . ‘ \
”if” ‘ . j . “‘.. “' ..flS‘ .
f". . " 3‘“! ’ A, O .‘9 ‘ . ...‘4
' " ‘ 1‘ ’5'.“ '0'qu '9’. .a
» {b . a_ ,k' .. ’3 . l
A ‘ t . 1.57."? ' 7‘ U .
, l v s . ‘ .0 . ‘ . l‘%\‘ ‘ 1". i‘
L ' . . . ‘ V
.049. °.‘, ‘.‘N”'~g~t frvi
y- " ‘ .

‘ ‘-¢';f
“Iv . , ' ‘ .u.
. ‘2: a}? . 3":Q{. ‘4 0“ s‘ 0" .‘1 . b?
“ filv e'
. . 0’4 ,

3
1".“ 9.19”" ‘3‘

 

'0

Figure 8. Large gold stained cells in the corpus callosum of the operated
hemisphere 3 days after stab injury

60

Figure 4

2
Percent radioactive nuclei in 0. 04mm of the cortical gray matter along the
wound and in 0. 09mm2 of corpus callosum in the operated hemisphere.

Animals were injected with 100 )LC of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours before stab injury or tumor transplantation.
Each pair of bars represents a different mouse. Animals designated 8*
received 3041c of tritiated thymidine 6 time s at 4 hour intervals 60 days
before tumor transplantation.

BOT

70: L_

gig/gin

V/////////////////////////////////////////M///////,

,//////////////////////////////////////////////////////A

7//////////////////////////////////////////////////////0

Corpus cal Ioeurn
Gray matter

gig/g?

gig/i

-

“gig/ZZZ/Zggg

 

gig

igggggn

0 w 0

q d

o

2 '
.202. 00.32 23.00

 

   

 

J
O 0

Tune interval (days) atter operation

FIGURE

4

62

Figure 5

Percent radioactive nuclei of each nuclear type in 0. 04mm2 of cortical gray
matter along the wound in the operated hemisphere.

Animals were injected with 100 )1C of tritiated thymidine 3 times at 4 hour
intervals beginning 24 hours prior to stab injury or tumor transplantation.

 

00%,, 7/////////////,////////////////////////////%
MMM wig/gig

Indicates range

.rl ll 00000000//////////0/////0
%%Z%/%%§/%§// ////////////////////////////////////////////////////////////0
giggiA

7////////////0
////////////////////////////////////////////////////////,

MMMMHMMMHM 7/////////////////////////////////////////////////////

% gig/80888888888888???

a 4:: 0 .o m. m M m m a s 4 2
3 2 2 2 8.0:... 2.200202 .522. 009.34

 

J

O

3 O '632C4

Large, pale,

30l63204 38l63264
Medium

Medium round
FIGURE 5

oval-to-round

elongated

64
Mitotic figures

 

A search was made for mitotic figures at each time interval. In the
animals injected with tritiated thymidine 1 day before injury, mitotic figures
were seen only in the sections from animals killed 3 days after stab injury.
Thirteen mitotic figures were counted in 0. 270mm2 of corpus callosum
(0.045mm2 in each of 6 sections). There were 3 prophases, 8 metaphases
and 2 anaphases. All of the figures were considered radioactive except 1
prophase which had 3 silver grains overlying it. Eight mitotic figures were
seen in 0. 110mm2 of corpus callosum (0.0275mm2 in each of 4 sections).

There were 2 prophases, 4 metaphases and 2 anaphases. Of these, 1 prophase,
l metaphase and l anaphase were not radioactive. One mitotic figure in
metaphase was seen in 0. 04mm2 of cortical gray matter of 1 animal injected

60 days before tumor transplantation. It was not radioactive.

Radioautographs of 20 )1. thick sections showed various degrees of labeling.
Some sections appeared devoid of labeled cells in the tissue surrounding stab
injury or tumor transplantation. Other sections showed labeled nuclei, but
none of the sections showed lesioned areas with heavily labeled cells (>20 grains
per nucleus). No quantitation of labeled nuclei was done on the slides from
any of the animals because of the variability of the technique. These sections
were used, rather, to count the total number of nuclei per 0. 0845mm2 of the

corpus callosum (Table 3).

Formalin-ammonium bromide perfused and fixed material

 

Five slides which were prestained with gold chloride sublimate were viewed
under the microscope before radioautographic processing. Large gold stained
cells were seen in the corpus callosum of the injured hemisphere. After the
emulsion was developed and fixed, the emulsion over the entire surface of the
slides appeared dis colored, the sections had turned a dark brown color and the
cells could not be distinguished from the background. The slides that were
processed for radioautography before staining with gold discolored when they
were placed in the gold bath, and neither silver grains nor cells could be

demonstrated.

65
Large, well-stained cells were demonstrated in gold stained sections
collected in the alternating serial section method. These cells were found
in the corpus callosum of the injured hemisphere. The alternating sections
which were processed for radioautography demonstrated no silver grains over

the tissue up to and including 14 weeks of exposure at 5°C.

66

Figure 6

Three small, lightly-stained astrocytic cell bodies are seen in the center
of the photomicrograph. Small pieces of fibrous material and cell bodies
which did not appear to have processes were not counted. Unoperated
animals and the uninjured hemisphere of operated animals showed a pre-
dominance of these cells. Cajal's gold chloride sublimate stain for
astrocytes. X 640.

Figure 7

Large, well-stained astrocytic cell bodies are demonstrated in the upper
left quadrant of the photomicrograph. The processes of these cells appear
thicker and more heavily stained than those of the small astrocytic cell
bodies. Fragments of fibrous material and cell bodies which did not

appear to have processes were not counted. A predominance of large, well-
stained cells was found in the corpus callosum of operated hemispheres.
Cajal's gold chloride sublimate stain for astrocytes. X 640.

 

Figure 7. Large gold stained cells in operated hemisphere

68

Figure 8

Large, well-stained cells are located in the corpus callosum and
hippocampus, in the lower one -third of the photomicrograph. The
lesion is located at the upper left and cortical gray matter occupies
the remaining upper two-thirds of the photomicrograph. Cajal's
gold chloride sublimate stain for astrocytes. X 280.

 

Figure 8. Large gold stained cells in the corpus callosum of the operated
hemisphere 3 days after stab injury

70

Figure 9

Radioactive nuclei located in the cortical gray matter 50 )1 - 100 )1 from

the edge of tumor 8 days after allografting. The animal was injected with
100 )1c of tritiated thymidine 3 times at 4 hour intervals beginning 24 hours
prior to tumor transplantation. The photomicrograph demonstrates labeled
glial —like nuclei which are derived from the mononuclear leukocytes of the
blood. Hematoxylin and eosin. X 720.

Figure 10

The section demonstrates radioactive medium round and medium elongated
nuclei located in the cortical gray matter 100 )1 - 150 )1 from the edge of the
tumor. The animal was injected with 100 1.00 of tritiated thymidine 3 times
at 4 hour intervals beginning 24 hours prior to tumor allograft and was
killed 8 days later. Hematoxylin and eosin. X 640.

 

ne fa
a. .4.

., C 4"
4e " ‘
1 . . ‘S .

fl 9" F": 4’.
o ‘ 4...; V. ' j. .
y 2.»: ‘0 " ’5
LP ._ i" '2?
,5. L
o . 3 e .’° ‘- ‘3

 

Figure 10. Radioactive nuclei in the cortical gray matter

72

Figure 11

Large, well-stained cell bodies are located in the corpus callosum and in the
cortical gray matter lying just above it. The unstained tumor mass occupies
the upper left quadrant of the photomicrograph; the corpus callosum is located
in the lower right corner. Cajal's gold chloride sublimate stain for astrocytes.

X 240.

 

Figure 11. Large gold stained cells in the corpus callosum of the operated
hemisphere 8 days after tumor allografting

74

Figure 12

The unstained tumor mass lies in the left one -third of the photomicrograph.
The large, well-stained cells in the corpus callosum and the cortical gray
matter lying just above it are located in the lower one -half of the photomi-
crograph. The animal from which this section was taken was injected with
tritiated thymidine as a juvenile animal. The animal from which the section
shown in Figure 11 was taken was injected with radioisotope 1 day prior to
transplantation. Cajal's gold chloride sublimate stain for astrocytes. X 280.

‘\\
. 4. 1‘

 

.14 V v.4;

Figure 12. Large gold stained cells in the corpus callosum of the operated
hemisphere 8 days after tumor allografting

76

Figure 13

Large, well -stained cell bodies located in the corpus callosum around the
lesion 16 days after stab injury. The lower part of the lesion appears filled
with debris. The corpus callosum can be seen at the lower right corner;
cortical gray matter occupies the remainder of the photomicrograph. Cajal's
gold chloride sublimate stain for astrocytes. X 280.

 

Figure 13. Large gold stained cells in the corpus callosum of the operated
hemisphere 16 days after stab injury

78

Figure 14

Large, well-stained cell bodies can be seen in the corpus callosum at the
bottom of the photomicrograph. The lower portion of the wound track is
almost entirely filled with necrotic debris. A few astrocytes can be seen

in the cortical gray matter just above the corpus callosum on both sides

of the lesion. This wound extended deep into the hippocampal region. Cajal's
gold chloride sublimate stain for astrocytes. X 280.

"\i

4'.-
a
\' 8

.‘. 1‘” g

g,

  

0.! \\
K \

’ 3‘
1‘
\

it
‘gyx ..
._ s
.\
\ ~ .

F
\e

\
t
\

\

Figure 14. Large gold stained cells in the corpus callosum of the operated
hemisphere 32 days after stab injury

80

Figure 15

Large, well-stained cell bodies located in the corpus callosum in the lower
left corner of the photomicrograph. Cajal's gold chloride sublimate stain
for astrocytes. X 280.

 

Large gold stained cells in the corpus callosum of the operated
hemisphere 64 days after stab injury

Figure 15.

DISCUSSION

The experiments described above were carried out to determine the
answer to the question: Do the mononuclear leukocytes of the blood that
respond to injuries of central nervous tissue give rise to cells that can be
stained with gold chloride sublimate and which might, as a result, be
mistaken for glial cells taking part in scar formation?

Mononuclear leukocytes are known to infiltrate central nervous tissue
in response to injury (Adrian and Walker, 1962). Monocytes are reported to
respond to stab injury of central nervous tissue (Konigsmark and Sidman, 1963;
Huntington and Terry, 1967). Lymphocytes have been suggested as the agent
responsible for rejection of allografted material in central nervous tissue
(Murphy and Sturm, 1923; Smith and Walker, 1967) as well as in non-nervous
tissues. Since lymphoid cells have been shown to enter an area of allografted
tissue in non -nervous tissue with hypertrophy and formation of multiple, long
thin cellular processes occurring in the area underlying the graft (Walker and
Goldman, 1963; Walker, _e_t__a_l_. , 1964; 1965), it would not be unreasonable to
expect this to happen in nervous tissue. The matrix of central nervous tissue
is the dense neuropil with its highly interwoven glial and axonal network.
Hypertrophy of the cytoplasm of lymphoid cells with attenuation of the cytoplasmic
expansions due to the density of the neuropil could be expected. The attenuated
cytoplasmic expansions might have an increased affinity for heavy metal stains.
If so, the morphology could suggest a likeness to the astrocytic glia, cells
indigenous to the central nervous system. In non -nervous tissue, hypertrophied
lymphoid cells have been reported to form a layer underlying an allograft. This
description is not unlike the "felting" of astrocytes around wounds of central
nervous tissue reported by Linell (1929). Thus, if lymphoid cells demonstrated
an affinity for heavy metal staining, they could be the cells which have been
reported to take part in glial scar formation. Hypertrophy and the formation of

82

83
cytoplasmic processes could so change cellular morphology that the only means
of identifying the origin and fate of these cells would be radioautographic
techniques and the specific heavy metal stain for astrocytes.

Attempts to study injured central nervous tissue with the combined
techniques of radioautography and specific heavy metal stains have given in-
consistent results (Konigsmark and Sidman, 1963; 1964; Hommes and Leblond,
1967). Konigsmark and Sidman (1963) reported a combined gold stain-
radioautographic technique in which the brain of an animal injected with
thymidine -C14 was fixed in formalin-ammonium bromide. Sections, 20 )1.
in thickness, were processed for radioautography after being prestained by
Cajal's gold chloride sublimate stain. They reported labeled cells with large
nuclei that did not take the astrocyte stain and no cells which were stained
were labeled. At a later date (1964), they reported that, whereas radioautographs
stained with toluidine blue showed 100 labeled cells per area, those stained
with gold showed only 10-20 labeled cells in the same area. Hommes and Leblond
(1967) reported that none of the well stained astrocytes appeared to be labeled
in their study utilizing combined radioautography and Cajal's gold chloride sublimate
stain. They too used 20 J1. thick sections, but their material had been fixed in
10% formalin.

In the present study, the radioautographs of material perfused and
fixed in Cajal's formalin-ammonium bromide (FAB) solution demonstrated
complete inhibition of the emulsion reaction to the B-particles from the tissue.

No silver grains were seen over the radioactive tissue. Sections prestained or
poststained with Cajal's gold chloride sublimate stain resulted in the loss of
definition of gold stained cells. Because there was no consistent method found

to combine the gold stain with the radioautographic technique, and since there was
variability noted in radioautographs of 29;].thick frozen sections of material

fixed in 10% formalin, a radioautographic technique that had given reproducible
results in the past was used for analysis in this study. Evaluation of the role

that lymphoid cells play in the gold stained cell reaction to central nervous

injury had to be inferred from studies of the two populations.

84

The experimental approach was based on the hypothesis that if the
population of gold stained cells responding to central nervous tissue injury
were derived from the population of mononuclear leukocytes, then, within a
defined area, the two populations would be expected to react similarly. If,
on the other hand, the two populations showed dissimilar responses to injury,
stab wound and the immunological stimulus of tumor allograft, then the
hypothesis would be considered false.

Large numbers of labeled cells, over 50% of the total number of cells,
were seen in the cortical gray matter and the corpus callosum surrounding the
tumor allograft. Over 20% of the cells surrounding a stab womd 3 days after
injury were labeled. These labeled cells were known to have infiltrated
the lesioned area from the blood because the radioisotope was injected 1 day
prior to injury. Thus, none of the label was available to cells synthesizing
DNA in response to the injury. Injection of radioisotope into unoperated animals
would have been expected to label extremely few cells in the nervous system
(Table 1; Smart and Leblond, 1961; Adrian and Walker, 1962 ). This method of
tracing labeled leukocytes has been used in the past to study injured muscle
(Bintliff and Walker, 1960), spinal cord (Adrian and Walker, 1962; Walker,

1963; Adrian, 1968) and brain (Konigsmark and Sidman, 1963; Huntington and
Terry, 1967; Smith and Walker, 1967).

Fluctuations in the labeled cell population were noted at different survival
intervals after stab injury, and a large influx of labeled cells occurred in response
to tumor allograft (Tables 6 and 12, Figure 4). These fluctuations in the labeled
cell population are consistent with the theory that mononuclear leukocytes migrate
in and out of a lesioned area. Macrophages from the blood infiltrate the area,
phagocytize necrotic debris and many leave the area. Lymphoid cells would be
expected to migrate into an area in response to antigenic stimulation. This
population of cells would be expected to remain in the area until the stimulation
subsided. Over 2 months after stab injury, many labeled cells were still seen to
surround the lesion (Tables 6 and 12). Significant differences were noted in the
labeled cell populations at different time intervals after stab injury and after
tumor allograft (Table 8).

85

The large gold stained cell population of the corpus callosum, on the other
hand, reacted differently than the labeled cell population. Large fluctuations
in the size of this population were not seen, either in response to stab injury
or to tumor allograft (Table 3, Figure 2). No significant differences were
demonstrated between individuals killed at the same interval or between animals
killed at different time intervals after stab injury (Tables 4 and 5, Figure 2).
Significant differences were not seen between the two individuals that had
received tumor allograft (Table 4). Nor were there significant differences
noted in the population of gold stained cells between animals which received stab
injury and those which received tumor allograft to the brain (Table 5). Large gold
stained cells were not seen in the cortical gray matter surrounding the lesion.
Comparison of the labeled cell population with the large gold population in the
corpus callosum showed that, within approximately the same area (0. 0900mm2
and O. 0845mm2), the two populations do not respond similarly to either stab
injury or tumor allograft (Tables 3 and 6, Figure 2).

To determine whether one of the subgroups of labeled nuclei rather than
the total labeled population influenced the gold cell population, the latter was
compared with the subgroup populations. Each of the nuclear subgroups
fluctuated with the type of stimulus and the survival interval after stab injury,
just as had been noted for the total labeled cell population (Tables 9 and 12,
Figures 3 and 5). Significant differences were noted for each nuclear type between
the individuals with the same survival time and between animals with different
survival intervals after stab injury (Tables 10 and 11). The results of this
comparison demonstrated that these subgroup populations respond differently
than does the gold cell population. Therefore, the gold cell population is not
influenced either by the total labeled cell population or a subgroup thereof.

Furthermore, when the glial population was labeled in the juvenile animals,
there was no evidence of migration of these cells to an area of tumor allograft.
This finding was consistent with previous work (Smith and Walker, 1967). There
was a decrease in the frequency of labeled nuclei 8 days after allografting, but this
could be explained by the number of infiltrating mononuclear luekocytes, which

would not have been radioactive due to dilution of label below detectable levels.

86

Lymphoid cells do infiltrate central nervous tissue in response to stab
wounds and tumor allografts, but these cells do not take on the gold stain.
Considering the behavior of the gold stained cells of the corpus callosum, it
is reasonable to accept the classical definition that these are astrocytes and
that they had hypertrophied in response to the stab wound or tumor allograft.
Lymphoid cells may hypertrophy after infiltrating central nervous tissue, but
they don't contribute to the gold staining cell (astrocyte) population of the
corpus callosum.

Nuclear hypertrophy was noted in response to stab injury of the brain
(Tables 9 and 12, Figures 3 and 5). This finding is consistent with previously
reported work (Walker, 1963) which suggested that lymphocytes infiltrated
an area surrounding stab injury in spinal cord and their nuclei hypertrophied
into glial-like nuclei. In the present study, the labeled medium -round and
medium -elongated nuclei decreased in frequency between 3 days and 16 days
after stab injury. The population of large, pale, oval-to -round nuclei in-
creased in frequency over the length of the experiment, as had been reported
previously (Walker, 1963).

Labeled mitotic figures noted at 3 days after stab injury indicate that the
mononuclear leukocytes which infiltrate an area of central nervous tissue injury
continue to divide as reported by Konigsmark and Sidman (1963) and Smith and

Walker (1967).

SUMMARY AND CONCLUSIONS

Tritiated thymidine radioautography and Cajal's gold chloride sublimate
stain for astrocytes were used to determine if the mononuclear leukocytes
which invade central nervous tissue in response to injury might be stained
with gold, and thus, might be mistaken for the gold stained cells which are
reported to take part in glial scar formation. Unoperated adult animals had a
low frequency of labeled cells in the brain. Unoperated hemispheres of brains
of injured animals also showed a low frequency of labeled cells in the brain.

In animals injected with tritiated thymidine 1 day before operation, large
numbers of labeled cells were seen in the corpus callosum and the cortical

gray matter along the wound 3 days after stab injury and 8 days after tumor
transplantation. Because of the time of labeling, these cells were known to be
mononuclear leukocytes. The labeled cell population fluctuated in response

to the type of stimulation, stab injury or the immunological stimulation of tumor
allograft, and with survival time after stab injury. Furthermore, when the
cells were classified as to nuclear type, the subgroups of the labeled cell
population also fluctuated with the type of stimulation and with survival time
after stab injury.

The role of glial cells in the response to tumor transplantation was
evaluated by injecting juvenile mice with tritiated thymidine (6 gram body weight)
60 days before operation. Injured hemispheres of these animals showed no
evidence for migration of labeled cells to the area of the tumor allograft.

The gold cell population of the corpus callosum remained stable with both
types of stimulation and with time of survival after stab injury. These cells were
not seen in the cortical gray matter along stab wounds or surrounding tumor
allografts.

Because the mononuclear leukocyte population fluctuated in response to
injury or transplantation and the gold cell population remained stable under the

87

88
same conditions, it was concluded that the two populations were independent
and that the hypertrophied lymphoid cells were not stained by Cajal's gold
chloride sublimate stain for astrocytes. The gold stained cell population
of the corpus callosum (astrocytes) hypertrophied in response to stab injury

or tumor transplantation.

BIBLIOGRAPHY

Adams, R. D. 1958. Implications of the biology of the neuroglia and
microglia cells for clinical neuropathology , p. 245-287. In: W. F.
Windle [ed] Biology of Neuroglia. Charles C. Thomas, Springfield,
Illinois.

Adrian, E. K. . Jr. , and B. E. Walker. 1962. Incorporation of thymidine-
H3 by cells in normal and injured mouse spinal cord. J. Neuropath.
Exp. Neurol. 21:597-609.

Adrian, E. K. , Jr. 1968. Cell division in injured spinal cord. Am. J. Anat.

Adrian, E. K. , Jr. , and R. D. Smothermon. 1970. Leukocytic infiltration
into hypoglossal nucleus following injury to hypoglossal nerve. Anat.
Rec. 166:99-115.

Altman, J. 1963. Autoradiographic investigation of cell proliferation in the
brains of rats and cats. Anat. Rec. 145:573 -S92.

Baggenstoss, A.H. , J. W. Kernohan, and J. F. Drapiewski. 1943. The
healing process in wounds of the brain. Amer. J. Clin. Path. 13:
333-348.

Bender, L. 1925a. Experimental production of gliosis. I. Effects on the
nervous system of the rabbit of intravenous and intraspinal injections
of cholesterol emulsions. Amer. J. Path. 1:657-666.

Bender, L. 1925b. Experimental production of gliosis. 11. Reaction of
brain tissue to the lipoid fractions and to the residue of brain extract.
Amer. J. Path. 1:667-675.

Bintliff, S. and B. E. Walker. 1960. Radioautographic study of skeletal
muscle regeneration. Amer. J. Anat. 106:233-246.

Cammermeyer, J. 1967. Microglia cells in diffuse and granulomatous
encephalitis in the rabbit. Acta Neuropath. 7:261—274.

Carmichael, F.A., J. W. Kernohan, and A. W. Adson. 1939. Histopatho-
genesis of cerebral abscess. Arch. Neurol. Psychiat. 42:1001-1029.

Cavanagh, J. B. 1970. The proliferation of astrocytes around a needle wound
in the rat brain. J. Anat. 106:471-487.

89

9O

Collias, J. C. , and E. E. Manuelidis. 1957. Histopathological changes
produced by implanted electrodes in cat brains. Comparisons with
histopathological changes in human and experimental puncture wounds.
J. Neurosurg. 14:302-328.

Cone, W. 1928. Acute pathological changes in neuroglia and in microglia.
Arch. Neurol. Psychiat. 20:34-68.

Cronkite, E. P., V. P. Bond, T. M. Fliedner, and J. R. Rubini. 1959. The
use of tritiated thymidine in the study of DNA synthesis and cell
turnover in hemopoietic tissues. Lab. Invest. 8:263-277.

Croxton, F. E. 1959. Elementary Statistics with Applications in Medicine
and the Biological Sciences. Dover Publications, Inc. , New York.
376 p.

Dalton, M. M. , O. R. Hommes, and C. P. Leblond. 1969. Correlation of
glial proliferation with age in the mouse brain. J. Comp. Neur. 134:
397-4000

Del Rio Hortega, P. 1919. E1 "tercer elemento" de los centros nerviosos.
1. La microglia en estado normal. 2. Intervencion de la micrOglia en
los procesos patologicos. 3. Naturaleza probable de la microglia. Bol.
Soc. Espan. Biol. 9:69-120. Cited by Del Rio Hortega, 1932.

Del Rio Hortega, P. , and W. Penfield. 1927. Cerebral Cicatrix; the reaction
of neuroglia and microglia to brain wounds. Bull. Johns Hopkins Hosp.
41:278 -303.

Del Rio Hortega, P. 1932. Microglia, p. 482-534. In: w. Penfield [edJ
Cytology and Cellular Pathology of the Nervous System, V0. 11. Paul
B. Hoeber, Inc., New York, N. Y.

Dougherty, T. F. 1944. Studies on the cytogenesis of microglia and their
relation to cells of the reticulo -endothelial system. Am. J. Anat.
74:61-89.

Dublin, W. B. 1945. Histogenesis of compound granular corpuscles in response
to cerebral trauma. Amer. J. Clin. Path. 15:228-229.

Dublin, W. B. 1967. Fundamentals of Neuropathblogy. Second ed. Charles
C. Thomas, Springfield, Illinois. 467 p.

Duckett, S. , and A. G. E. Pearse. 1969. Periventricular astrocytosis and
acute necrotizing encephalitis. Acta Neuropath. 12:90 -93.

Dunning, H. S. , and J. Furth. 1935. Studies on the relation between microglia,
histiocytes and monocytes. Amer. J. Path. 11:895-913.

91

Essick, C. R. 1919. Pathology of experimental traumatic abscess of the
brain. Arch. Neurol. Psychiat. 1:673 -704.

Ferraro, A. , and L. Davidoff. 1928. Reaction of oligodendroglia to injury
of the brain. Arch. Path. 6:1030-1053.

Friede, R. L. 1967. Incorporation patterns of tritiated thymidine following
cutting of the sciatic nerve in rats. J. Neuropath. Exp. Neurol. 26:
134-135e

Friede, R. L. , and M.A. Johnstone. 1967. Responses of thymidine labeling
of nuclei in gray matter and nerve following sciatic transection. Acta
Neuropath. 7:218-231.

Fuentes, C. , and R. Marty. 1970. Oligodendrocytose réactionnelle au cours
de la de’gene’rescence expe’rimentale du cortex ce’rébral. J. Neurol. Sci.

Globus, J. H. 1928. Glia response in chronic vascular disease of the brain.
Arch. Neurol. Psychiat. 20:14-33.

Good, R.A. 1950. Experimental allergic brain inflammation, a morphological
study. J. Neuropath. Exp. Neurol. 9:78-92.

Greenfield, J. G. 1932. Inflammatory cells in the central nervous system.
p. 1220-2138. In: w. Penfield [ed] Cytology and Cellular Pathology
of the Nervous System, Vol. HI. Paul B. Hoeber, Inc., New York, N. Y.

Hain, R. F. , W. O. Rieke, and N. B. Everett. 1960. Evidence of mitosis
in neuroglia as revealed by radioautography employing tritiated thymidine.
J. Neuropath. Exp. Neurol. 19:147-148.

Hassin, G. B. 1936. Reacting cells in the brain in the presence of a foreign
body. Arch. Neurol. Psychiat. 36:231-247.

Hicks, S. P. 1947. Brain repair. 11. Role of brain lipids in genesis of gliosis.
Arch. Path. 43:15-27.

Hommes, O. R. , and C. P. Leblond. 1967. Mitotic division of neuroglia in
the normal adult rat. J. Comp. Neur. 129:269-278.

Hughes, W. L., V. P. Bond, G. Brecher, E. P. Cronkite, R. P. Painter, R.
Quastler, and R. G. Sherman. 1958. Cellular proliferation in the mouse
as revealed by autoradiography with tritiated thymidine. Proc. Nat.
Acad. Sci. 44:476-483.

92

Huntington, H. W. , and R. D. Terry. 1966. The origin of the reactive cells
in the cerebral stab wound. J. Neuropath. Exp. Neurol. 25:646-653.

Klatzo, I. , A. Piraux, and E.J. Laskowski. 1958. The relationship between
edema, blood-brain barrier, and tissue elements in a local brain injury.
J. Neuropath. Exp. Neurol. 17:548-564.

Konigsmark, B. W. , and R. L. Sidman. 1963. Origin of brain macrophages
in the mouse. J. Neuropath. Exp. Neurol. 22:643-676.

Konigsmark, B. W. and R. L. Sidman. 1964. Response of astrocytes to brain
injury. J. Neuropath. Exp. Neurol. 24:142-144.

Kosunen, T. U. , B.H. Waksman, and I.K. Samuelsson. 1963. Radioautographic
study of cellular mechanisms in delayed hypersensitivity. II. Experimen-
tal allergic encephalomyelitis in the rat. J. Neuropath. Exp. Neurol.
22:367-380.

Kovak, W. and H. Denk. 1968. Der Hirnstamm der Maus. Springer-Verlag.
New York. 150 pp.

Kreutzberg, G. 1966. Autoradiographische Untersuchung iiber die Beteiligung
von Gliazellen an der axonalen Reaktion in Facialiskern der Ratte. Acta
Neuropath. 7:149-161.

Lampert, P. 1967. Electron microscopic studies on ordinary and hyperacute
experimental allergic encepalomyelitis. Acta Neuropath. 9:99-126.

Leblond, C. P., B. Messier, and B. Kopriwa. 1959. Thymidine-H3 as a tool
for investigation of the renewal of cell populations. Lab. Invest. 8:296 -308.

Levine, S., A. Hirano, and H. M. Zimmerman. 1965. Hyperacute allergic
encephalomyelitis. Amer. J. Path. 47:209-221.

Lewis, R. C. , and R. L. Swank. 1953. Effect of cerebral microembolism on
the perivascular neuroglia. J. Neuropath. Exp. Neurol. 12:57-63.

Linell, E.A. 1929. The histology of neuroglial changes following cerebral
trauma. An experimental investigation. Arch. Neurol. Psychiat.
22:926-958.

Macklin, C. C. , and M. J. Macklin. 1920. A study of brain repair in the rat
by the use of trypan blue; with special reference to the vital staining of
the macrophages. Arch. Neurol. Psychiat. 3:353 -394.

Marsh, R. F., and R. P. Hanson. 1969. Transmissible mink encephalopathy:
neuroglial response. Am. J. Vet. Res. 30:1643-1653.

Maxwell, D. S. , and L. Kruger. 1965. Small blood vessels and the origin of
phagocytes in the rat cerebral cortex following heavy particle irradiation.
Exp. Neurol. 12:33-54.

93

Maxwell, D. S. , and L. Kruger. 1966. The reactive oligodendrocytes. An
electron microscopic study of cerebral cortex following Ct-particle
irradiation. Am. J. Anat. 118:437-460.

Medawar, P. B. 1948. Immunity to homologous grafted skin. III. The fate
of skin homografts transplanted to the brain, to subcutaneous tissue
and to the anterior chamber of the eye. Brit. J. Exp. Path. 29:58 -69.

Messier, B., C. P. Leblond, and 1. Smart. 1958. Presence of DNA synthesis
and mitosis in the brain of young adult mice. Exp. Cell Res. 14:224-226.

Messier, B. , and C. P. Leblond. 1960. Cell proliferation and migration as
revealed by radioautography after injection of thymidine -H3 into male rats
and mice. Am. J. Anat. 106:247-285.

Mori, S. ,. and C. P. Leblond. 1969a. Identification of microglia in light and
electron microscopy. J. Comp. Neur. 135:57-80.

Mori, S. , and C. P. Leblond. 1969b. Electron microscopic features and
proliferation of astrocytes in the corpus callosum of the rat. J. Comp.
Neur. 137:197-226.

Mori, S. , and C. P. Leblond. 1970. Electron microscopic identification of
three classes of oligodendrocytes and a preliminary study of their
proliferative activity in the corpus callosum of young rats. J. Comp.
Neur. 139:1-29.

Murphy, J. B. , and E. Sturm. 1923. Conditions determining the transplantability
of tissues in the brain. J. Exp. Med. 38:183-197.

Noetzel, H. 1959. Autoradiographische Untersuchungen iiber die physiologische
Regeneration der Gliazellen. Verh. Deutsch. Ges. Path. 46:341-342.
Cited by Hommes and Leblond, 1967.

Noetzel, H. , and J. Rox. 1964. Autoradiographische Untersuchungen iiber
Zellteilung und Zellentwicklung im Gehirn der erwachsenen Maus und
des erwachsenen Rhesus-Affen nach Injektion von radioaktivem Thymidin.
Acta Neuropath. 3:326-342.

Olsson, Y. , and J. Sj'cistrand. 1969. Origin of macrophages in Wallerian
degeneration of peripheral nerves demonstrated autoradiographically.
Exp. Neurol. 23:102-112.

Paterson, P. Y. 1960. Transfer of allergic encephalomyelitis in rats by means
of lymph node cells. J. Exp. Med. 111:119-136.

Penfield, W. 1925. Microglia and the process of phagocytosis in gliomas.
Amer. J. Path. 1:77-90.

94

Penfield, W. , and R. C. Buckley. 1928. Punctures of the brain. Arch.
Neurol. Psychiat. 20:1-13.

Penfield, w. 1932. Neuroglia, p. 423-479. In: w. Penfield [0.] Cytology
and Cellular Pathology of the Nervous System, Vol. IL Paul B. Hoeber,
Inc. , New York, N. Y.

Penfield, W. , and W. V. Cone. 1961. Neuroglia and microglia (the metallic
methods), p. 399-431. In: R. M. Jones [edJ McClung's Handbook of
Microscopical Technique. Third ed. Hafner Publishing Co. , Inc. , New
York, N. Y.

Ram6n y Cajal, S. 1913. Contribucional concimiento de la neuroglia del
cerebro humano. Trab. Lab. Invest. Biol. Univ. Madr. 11:225-315.
Cited by Mori and Leblond, 1969b.

Rand, C. W. , and C. R. Courville. l932a. Histologic changes in the brain in
cases of fatal injury to the head. III. Reaction of microglia and
oligodendroglia. Arch. Neurol. Psychiat. 27 :605 -644.

Rand, C. W. , and C. R. Courville. 1932b. Histologic studies of the brain
in cases of fatal injury to the head. IV. Reaction of the classic neuroglia.
Arch. Neurol. Psychiat. 27:1342-1379.

Reznikov, K. Yu. 1968. Effect of brain injury and intracerebral injection of
RNA on thymidine -H3 incorporation into adult mouse brain cells. Dokl.
Akad. Nauk. SSSR. 181:467-469.

Roessmann, U. , and R. L. Friede. 1967. Entry of labeled donor cells from the
bloodstream into the CNS. J. Neuropath. Exp. Neurol. 26:144-145.

Roessmann, U. , and R. L. Friede. 1968. Entry of labeled monocytic cells into
the central nervous system. Acta Neuropath. 10:359-362.

Russell, D. S. 1929. Intravital staining of microglia with trypan blue. Amer.
J. Path. 5:451-458.

Schultz, R. L. , and D. C. Pease. 1959. Cicatrix formation in rat cerebral cortex
as revealed by electron microscopy. Amer. J. Path. 35:1017-1041.

Shirai, Y. 1921. On the transplantation of the rat sarcoma in adult heterogenous
animals. Jap. Med. World. 1:14-15. Cited by Murphy and Sturm, 1923.

Silver, M. L. , and A. E. Walker. 1947. Histopathology of thermocoagulation
of the cerebral cortex. J. Neuropath. Exp. Neurol. 6:311-322.

Sj'ostrand, J. 1965a. DNA synthesis in glia cells during nerve regeneration.
Experientia 21:142-143.

95

Sjc'istrand, J. 1965b. Proliferative changes in glial cells during nerve
regeneration. Z. Zellforsch. 68:481-493.

Sj6strand, J. 1966. Studies on glial cells in the hypoglossal nucleus of the
rabbit during nerve regeneration. Acta Physiol. Scand. 67(Supp1. 270):
1 -17.

Smart, 1. 1960. DNA synthesis in the brain of juvenile and adult white mice.
Anat. Rec. 136:279 -280.

Smart, 1. 1961. The subependymal layer of the mouse brain and its cell
production as shown by radioautography after thymidine -H3 injection.
J. Comp. Neur. 116:325-347.

Smart, 1. , and C. P. Leblond. 1961. Evidence for division and transformations
of neuroglia cells in the mouse brain as derived from radioautography
after injection of thymidine-H3. J. Comp. Neur. 116:349-367.

Smith, C. W. , and B. E. Walker. 1967. Glial and lymphoid cell response to
tumor implantation in mouse brain. Texas Rep. Biol. Med. 25:585 -600.

Swank, R. L. , and R. J. Hain. 1952. The effect of different sized emboli on
the vascular system and parenchyma of the brain. J. Neuropath. Exp.
Neurol. 11:280-299.

Tschistowitsch, T. 1898. Ueber die Heilung aseptischer traumatischer
Gehirnverletzungen. Ziegler's Beitréige 23:321. Cited by Essick, 1919.

Vaughn, J. E. 1965. Electron microscopic study of the vascular response to
axonal degeneration in rat optic nerve. Anat. Rec. 1512428. (Abstr.)

Waksman, B. H. , and R. D. Adams. 1962. A histologic study of the early
lesion in experimental allergic encephalomyelitis in the guinea pig and
rabbit. Amer. J. Path. 41:135-162.

Walker, B. E. 1959. Radioautographic observations on regeneration of
transitional epithelium. Texas Rep. Biol. Med. 17:375-384.

Walker, B. E. 1963. Infiltration and transformation of lymphoid cells in areas
of spinal cord injury. Texas Rep. Biol. Med. 21:615-630.

Walker, B. E. , and A. S. Goldman. 1963. Thymidine-H3 radioautography of
skin grafts in mice. Texas Rep. Biol. Med. 21:425-441.

Walker, B. E. , R. D. Yates, and D. Duncan. 1964. Cell morphology in tissue
underlying skin grafts. Anat. Rec. 149:651-670.

96

Walker, B. E., R. D. Yates, and D. Duncan. 1965. Morphologic analysis of
the tumor homograft reaction. Texas Rep. Biol. Med. 23:814-827.

Watson, W. E. 1965. An autoradiographic study of the incorporation of nucleic
acid precursors by neurones and glia during nerve regeneration. J. Physiol.
180:741-753.

Wells, A.Q. , and E.A. Carmichael. 1930. Microglia: An experimental study
by means of tissue culture and vital staining brain. Brain 53:1-10.

Wiener, J., D. Spiro, and P. S. Russell. 1964. An electron microscopic study
of the homograft reaction. Amer. J. Path. 44:319-347.

Wilson, R. B. 1926. Brain repair. Arch. Neurol. Psychiat. 15:75-84.
Wolman, M. 1958. Studies on the impregnation of nervous tissue elements.

IV. Mechanism of impregnation of oligodendroglia and microglia.
Lab. Invest. 7:52 -57.

MICHIGAN STATE UNIVERSITY LIBRARIES

l 1:le llllllllllllll

3143 0196

31293