Journal

of Leukocyte Biology 50:252-262

(1991)

Determination of the Number and Distribution of

Macrophages, Lymphocytes, and Granulocytes in the
Mouse Uterus From Mating Through Implantation
Mamata
Department

De, Rajani
of Pathology,

Choudhuri,
University of Kansas

and

Gary

W: Wood

Medical Center,

Kansas

City

The concentration
and distribution
of F4-80 positive cells (macrophages)
and common
leukocyte antigen (CLA) positive (bone marrow derived) cells were assessed in mouse
uterus between days 1 and 8 of pregnancy.
High numbers of polymorphonuclear
leukocytes and lymphocytes
were present on days I and 2, but not thereafter. Granulocytes were found both in the endometrium
and within the luminal epithelium.
The
percentage of total cells contributed
by macrophages
was high on days 1 and 2. That
percentage decreased significantly on day 3, then increased again on day 5 and remained
high through day 8. Macrophages always were found in myometrial stroma. Macrophages
were found throughout
the endometrium
on days 2 through 8. High numbers of
macrophages
were observed near epithelia, particularly on days 1 2, 4, and 5. Few F4-80
or CLAP cells were observed within the developing primary and secondary decidua. The
results demonstrate
that an inflammation-like
cellular response occurs in the uterus
following mating and that macrophages
are a major cellular component
of the uterus
during early pregnancy.
,

Key words:

inflammation,

decidualization,

INTRODUCTION
Pregnancy-associated
intrauterine
events
may be divided into four distinct
periods:
preimplantation,
implantation,
placenta
development,
and parturition.
Uterine
cellular changes
which occur during each of these periods
are under the control
of gonadal
steroids
[1 ,6,7, 18,19].
Early morphologic
studies of the pregnant
uterus in mice
and rats during
the preimplantation
period described
an
inflammation-like
response
to mating
[16, 17]. Polymorphonuclear
leukocytes
(PMNs)
appear
in the uterus
within
hours.
PMNs
are recruited
from
the blood,
migrate
into the endometrium,
and pass through
the
luminal
epithelium.
As is the case with inflammatory
responses
which occur under pathologic
conditions,
this
early granulocytic
infiltrate
is closely
followed
by a
mononuclear
component,
which
peaks
on day 1 and
subsides
by day 3 . Beginning
on day 3 uterine
stromal
cells proliferate
and differentiate
in response
to estrogen
and progesterone
as part ofthe physiologic
preparation
of
the uterus for implantation
of the blastocyst
[7]. However,
there
is little evidence
from earlier
studies
for
inflammatory
cells
in the uterus
during
that period.
Blastocyst
implantation
initiates
another
series of cellular
changes
described
as decidualization.
Primary
and secondary decidua form around the implantation
site through
further proliferation
and differentiation
of uterine stromal
cells [35].
Significant
numbers
of macrophages
have
,

1991

Wiley-Liss,

Inc.

estrogen,

progesterone

been identified
in the deep endometrium
during
decidualization
[32].
Acute inflammatory
responses
are easily recognized
at
the light microscopic
level in tissue,
because
lymphocytes, granulocytes,
and some macrophages
are readily
distinguishable
from other tissue cells [ I 6, 17] . Except for
the studies
demonstrating
an acute
inflammatory
response immediately
following
mating
[ 15 16] morphologic, functional,
and immunohistochemical
studies have
detected
only low numbers
of inflammatory
cells
in
uterine
tissue
during
the early
stages
of pregnancy
[8,12-14,16,17,24,25,26,30,32,33,34,36].
In contrast,
cell suspension
analyses
have indicated
that the uterus,
like all organs which are partially
comprised
by connective tissue
[27], contains
significant
numbers
of macrophages
both
during
the estrous
cycle
and during
pregnancy
[5, 12,21 ,29,33].
The result
is that there is
considerable
controversy
regarding
the numbers
and
distribution
of uterine
macrophages
during
pregnancy.
Analyses
of macrophage
numbers
and distribution
in
tissue have two major difficulties.
First, strictly morphologic analysis
suffers
from the high degree
of morpho,

Received
Reprint
Center,

September

25,

1990;

accepted

requests:
Gary W. Wood,
39th and Rainbow
Boulevard,

January
University
Kansas

10, 1991.
of Kansas
Medical
City, KS 66103.

Inflammatory

logic heterogeneity
exhibited
by macrophages.
Second,
other
localization
techniques,
including
immunohistochemistry,
are limited
by sensitivity
and specificity
problems.
The current
study was designed
to overcome
those difficulties
and determine
the numbers
and distribution of bone marrow
derived
cells, specifically
lymphocytes,
granulocytes,
and macrophages,
in the uterus

during

the preimplantation

(days

1-3) and periimplanta-

tion periods
(days 3-8). Our intent was to reconcile
cell
suspension
data with tissue distribution
data using immunohistochemistry
with antibodies
which
have been
successfully
applied
to immunohistochemical
analysis
of
leukocytes
in other organs
[2,8-11,27].
In mice, macrophages
may be accurately
quantitated
and localized
in tissue using the monoclonal
anti-macrophage
antibody,
F4-80,
because
F4-80 is specific
for
and expressed
on all mature
mouse
macrophages
[2,91 1 ,27]. Recently,
we used F4-80 and anti-mouse
cornmon leukocyte
antigen
(CLA)
[23,30]
to quantitate
and
determine
the distribution
of macrophages
and other bone
marrow derived
cells in the mouse
uterus
during
the
estrous
cycle [5] . Those studies demonstrated
that macrophages
account
for 7-10%
of total cells in the cycling
uterus and account
for nearly all bone marrow
derived
cells. The results further
demonstrated
that the presence
and distribution
of macrophages
in the uterus is dependent on cyclical
production
of estrogen
and progesterone
[5]. Those
findings
suggested
that, because
significant
changes
in sex hormone
levels occur during pregnancy,
precise localization
of macrophages
could provide
further
information
regarding
possible
relationships
between
estrogen,
progesterone
and macrophages
during
pregnancy. They also established
a technical
approach
which
could
be used to both quantitate
and determine
the

Cells

in Early

Pregnancy

253

experiment
were derived
from cell suspensions
prepared
from a pool of at least four uteri . All experiments
were
repeated
at least twice. Uterus was digested
for 60 mm at
an ambient temperature
with type I collagenase
(1 mg/rn!,
Worthington
Biochemical)
together
with type IX bacterial protease
(0.5 mg/rn!,
Sigma
Chemical
Co.).
The
dispersed
cells were washed
with RPMI
1 640 containing
5% fetal bovine
serum and cytocentrifuged
onto slides.
One set of slides was stained
with Wrights
stain for
morphologic
analysis.
A second
set of slides was prepared for immunocytochemistry.
The cells were vacuumdried, fixed in periodate-lysine-paraformaldehyde
(PLPbuffer)
[22] for 10 mm and immediately
stained
with
the selected
antibodies
by immunocytochernistry
using
Vectastain
ABC kits (Vector
Laboratories).
The procedure described
by the Vectastain
Kit insert was followed
precisely,
with the exception
that cytospins
were not
quenched
to remove
endogenous
peroxidase
and the

preblocking

step, which

employs

serum

supplied

by the

vendor,
was eliminated.
In order to obtain immunolocalization
in intact tissue,
frozen tissue sections
were prepared
and stained.
Uteri
were frozen in liquid nitrogen
and stored at -70#{176}C.Six
micron-thick
frozen sections
were cryostat-cut,
vacuumdried,
fixed for 10 mm in acetone,
and stained
on the
same day . Sections
were stained
as described
above for
cytocentrifuged
cells.

Antibodies

approach
to quantitate
and determine
the distribution
ofCLA
positive
cells was performed
in parallel,
because
of the high degree of controversy
regarding
the levels of
macrophages
and their relationship
to other uterine bone
marrow derived cells [8,12-14,16,17,25,26,30,32,33,

Three monoclonal
antibodies
were used in this study:
1) F4-80,
a rat IgG2
anti-mouse
macrophage
antibody
[2,9];
2) Ml/9.3.4.HL.2,
a rat IgG2
anti-mouse
CLA
[23,30,31];
and 3) 53-6.72,
a rat IgG2 anti-mouse
CD8,
a determinant
expressed
by cytotoxic
T cells [15], which
was employed
as the negative
control.
F4-80 is a widely
used mouse macrophage
specific
antibody
. The antigen
is
expressed
on all mature
macrophages.
The antibody
has
been used successfully
to determine
the distribution
of
macrophages
in various
tissues
and organs
using immunohistology
[2,9-1 1 ,27]. The anti-CLA
monoclonal
was

34,36].

employed

distribution

of macrophages

MATERIALS
Animals

in uterine

tissue.

AND METHODS

antibodies

were

determine

Females

and males were mixed

to obtain pregnant
animals.
Appearance
was defined
as day 1 of pregnancy.

were obtained
maintained
in
Animal
Care

euthanized
removed.

CLA is an accepted

by CO2 poisoning,
and
Data for each day for

in parallel

contribution

bone marrow

derived

antigenic

marker

to identify
cells.
The

principally

of macrophages

cell population.

and
two

to
to the

The

plug

uteri
each

culture
tation,

in a ratio of 2:1
of a vaginal

employed

the relative

total uterine

it can be used
marrow
derived

anti-CD8,
which contains
a high concentration
of antibodies
reactive
with CD8 positive
T cells,
and which,
like the other two antibodies,
is very effective
in immunohistochemistry
and immunocytochemistry
using
the
anti-rat
IgG Vectastain
ABC developing
system,
was
employed
as a control
for systematic
non-specific
staining. All monoclonal
antibodies
were
prepared
from

Tissue
Mice were
were surgically

because

for leukocytes.
Thus,
quantitate
uterine
bone

Six to eight week-old

Swiss/CD1
mice
from Harlan/Sprague
Dawley,
bred, and
the University
of Kansas
Medical
Center

Facility.

A similar

supernatants
by 50% ammonium
sulfate precipifollowed
by dialysis against Dulbeccos
phos-

254

Deetal.

TABLE 1. Quantitation
of Lymphocytes,
PMNs, and
Macrophages
In Uterine Cell Suspensions
From
Pregnant
Mice8
D

Lymphocytes

I
2
3
4
5
6
7
8
cell

F4-80

1.0
36
l7C
3.0c
4.4
3.6
2.2
2.4
0.5

0.4
2.6
0.7
1.4
1.2
1.7
1.4
1.3
0.5

0h

aUterine

PMNS

suspensions

from

timed

phate buffered
saline (PBS).
were determined
by titrating
sections.
CLAP

9
2lC
27C
l7C
19
35
40
NCd
35
pregnant

mice

Description
of the General Approach to
Distributional
Analysis of Cells Expressing
F4-80 and/or CLA

10
73C

60
1SC
19

Uterus was obtained

from normal mice during proestrus and from each day of pregnancy.
Analysis
was
performed
on at least eight different
mice at each time
point. Tissue was mounted
upright
in order that cross

39C

36
NC
34
were

sections
would be cut. At least five sections
were placed
on each slide,
and three slides were stained
for each
antibody.
Tissue
was examined
for general
distribution
and approximate
number
of positive
cells. Tissue
was
examined
to determine
whether
or not positive
cells were
evenly
distributed
around
the uterine
lumen,
e.g.
were
cells concentrated
on one side of the uterus.
Tissue also
was examined
to determine
whether
or not positive
cells
were concentrated
in myometrium,
deep endometrium,
or superficial
endometrium.
Tissue
was examined
to
determine
whether
or not positive
cells were associated
with particular
structures,
e.g.
blood vessels or glandular epithelium.
The immunostaining
was controlled
in the following

stained

with Wrights
or by immunocytochemistry
with F4-80 or anti-CLA.
Lymphocytes
and granulocytes
were quantitated
by morphologic
examination.
For each data point, a total of 1 ,000 cells was counted
on three slides. Data are expressed
as mean of the percentage
of
positive
cells in cell suspensions.
Standard
errors were calculated
and
always
were less than 20% of the mean. Controls
(PBS and antiCD8)
routinely
were negative.
Data were pooled from at least four
mice at each time point.
hDay zero data are from uteri removed
from mice during proestrus.
cThe difference
between
this value and the value obtained
on the
previous
day is significant
(p
0.05),
ANOVA
and Newman-Keuls
multiple
dNC
not counted.
Technical
failure.

as determined
comparison

by one
test.

way

,-

-f-

4.
..

:#{149}-

:-

Optimum
working
dilutions
each antibody
on uterine

..

-7?;4-.;

.-

Fig. 1 . Section of uterus from mouse In proestrus. Section stained by immunoperoxidase

Immunohistochemistry
with F4-80. Control sections stained with anti-CD8 were completely
negative. LE, luminal epithelium;
GE, glandular epithellum; EN, endometrium.
x560.

Inflammatory

I_

in Early Pregnancy

Cells

.7?

!=

255

-4,

#{149}r
I-.)

.;.

9,
1:

#{149}1

4.
,
Fig. 2. Cytocentrifuge
preparations
of uterine cells from I day pregnant mice. A: Cells stained
with Wrights stain. B: Cells stained with antl-CD8. PBS control sections were identical. C: Cells
stained with anti-CLA. D: Cells stained with F4-80. All magnifications
x I ,l20.

ways. First, only monoclonal

antibodies
with well characterized
specificity
were employed.
Second,
PBS and
rat IgG monoclonal
antibody
to an unrelated
antigen
always were included
to control for non-specific
staining.
The anti-CD8
antibody
employed
has been
used
in
separate
immunohistologic
studies to localize
CD8
cells
in thymus,
spleen,
and lymph
nodes
of Swiss
mice.
Third, the approximate
number
of positive
cells detected
with F4-80 and anti-CLA
in this study was similar to the
number
of macrophages
identified
in uterine cell suspensions during previous
studies using independent
methodology [12,21,33].

Statistics
Statistical
positive

cells

analysis

of differences

on different

days

between
of pregnancy

numbers
was

of
done

using the Newman-Keuls

multiple comparison
test following the one way analysis of variance (ANOVA
for
repeated
mark),

measures)
test using BMDP
1988 version.
As the data

(registered
tradewere obtained
as

percentages,

arcsine

transformation

RESULTS
Tissue Distribution

of Antigen

was performed

prior

to ANOVA.

Positive

Cells

On day zero,
immediately
before
mating,
the proestrous
uterus
contains
significant
numbers
of macrophages
and few other bone marrow
derived
cells (Table
1 Fig. 1). Evidence
for this derives
from the fact that
CLA and F4-80
staining
patterns
and the numbers
of
CLAP
and F4-80
cells in suspensions
are virtually
identical
in the cycling
uterus
(Table
1 ). Immediately
prior to mating,
CLAP
and F4-80
cells contribute
approximately
10% of total uterine
cells
(Table
1).
Twenty-four
hours later, on day 1 of pregnancy,
numbers
of both CLAP
and F4-80
cells are substantially
increased
(Table
1 Figs.
2, 3). However,
on day 1,
macrophages
contribute
only
a fraction
of the total
number
of bone marrow
derived
cells.
This is clearly
,

256

Deetal.
evidenced
by the fact that immunostaining
with antibody
to F4-80 outlined
a lower number
of cells in the tissue
than did anti-CLA
(Figs.
2, 3), and cell suspensions
contained
lymphocytes
and PMNs
in addition
to macrophages
(Table 1 Fig. 2) . The distributional
patterns
for
CLA and F4-80 expression
on day 1 are similar.
Cells
expressing
both antigens
are concentrated
in superficial
endometrium.
However,
CLAP
but not F4-80
cells
are detected
within the luminal
epithelium
(Fig. 3).
The uterine
cellular
picture
undergoes
qualitative
as
well as quantitative
changes
between
days 1 and 2. The
number of CLAP cells is reduced,
while the number
of
F4-80
cells remains
approximately
the same (Table
1).
Distribution
of CLAP
and F4-80
cells on day 2 is
similar
to day 1 except
that CLAP
cells are no longer
detectable
in the epithelium.
Both observations
reflect the
waning
inflammatory
response,
where lymphocytes
and
PMNs usually
predominate
in early phases,
but rapidly
,

decrease

in number.

Distribution

F4-80
cells is changed
the first 2 days, there are
many more CLAP
than F4-80
cells. However,
on day
3, CLAP
and F4-80
cell distribution
cannot
be distinguished
(Fig. 4), and the numbers
of CLAP
and F4-80
cells in suspensions
are similar
(Table
1). Few granulocytes or lymphocytes
are detectable
on day 3 (Table
1)
and this is a trend which continues
for the remainder
of
pregnancy.
CLAP
and F4-80
positive
cells also no
longer are focused in the superficial
endometrium
on day
3. They are relatively
evenly
distributed
throughout
the
endometrium
and myometrium
(Fig. 4). The distributional pattern
for CLAP
and F4-80
cells on day 3 is

considerably

similar
during

Fig. 3. Sections of uterus from 1 day pregnant mice. Sections

stained by immunoperoxidase
immunohistochemistry.
A: Saction stained with anti-CD8. PBS control sections were identical.
Note endogenous
peroxidase
positive cells, which were more
evident on day 1 than on later days of pregnancy.
B: Section
stained with anti-CLA. Note positive cells in LE, but not GE. C:
Section stained with monoclonal antibody to mouse macrophages F4-80. EN, endometrium;
GE, glandular epithelium;
LE,
luminal epithelium;
M, myometrium.
All magnifications
x560.

of CLAP

and

on day 3 . During

to the pattern observed

in the cycling uterus
proestrus
and estrus (Fig. 1). The number of

CLAP and F4-80

cells is significantly
lower on day 3
than on day 2, but still is higher than in the cycling
uterus
(Table
1; p
0.01 ). The same pattern
also is evident
with the granulocyte
population.
The percentage
of
granulocytes
found in the uterus on day 3 is significantly
less than on day 2 and significantly
greater
than in the
cycling
uterus (p
0.05).
The number
of CLAP
and F4-80
cells remains
unchanged
through
day 4 (Table
1), but the cells appear
to undergo
additional
distributional
changes.
CLAP
and
F4-80
cells reaccumulate
in the endometrial
stroma
underlying
all glandular
and luminal
epithelia,
regardless
of their position
in the uterus.
This pattern
continues
through
day 5 (Fig. 5). On day 5, there is a significant
increase
in the number
of CLAP
and F4-80
cells in
suspensions
(Table 1), which persists through day 8 . The
subepithelial
localization
of antigen
positive
cells
is
observed
throughout
the uterus and, in fact, CLAP
and
F4-80
cells are observed
even in the superficial
endo-

metrium

adjacent

Major changes
cells again are

to the implanting

blastocyst

(Fig.

6).

in the distribution
of CLAP and F4-80
evident
in the anti-mesometrial
uterus

Inflammatory Cells in Early Pregnancy

.4

.a

r
I,

.
-

4-

:;)
,.

.t;!b4

4A,

j!

:;.;

#{149}#{149}

..

I
,t

?#
lr

;a:

$44

..

.,1.

-:

,bP_A.

.:e::.

#{149}

r-

p.
,

*
#{149}
...

Fig. 4. Sections of uterus from 3 day pregnant mice. Sections stained by immunoperoxidase
immunohistochemistry.
A: Section stained with anti-CLA. B: Section stained with F4-80. LE,
luminal epithelium;
M, myometrium;
EN, endometrium.
All magnifications
x560.

257

258

Deetal.

ifr
:

%!
ii.

-.

a
f

;bmr

-.4

vt
-

.#{149}

-.

,.

/
%

.1

:(

Fig. 5. Sections of uterus from 5 day pregnant mice. Sections stained by immunoperoxidase
immunohistochemistry.
A: Section stained with anti-CIA. B: Section stained with F4-80. Control
sections stained with anti-CD8 or buffer contained isolated peroxidase positive cells. LE, luminal
epithellum; GE, glandular epithelium; EN, endometrium.
All magnifications
x560.

ufr

Inflammatory

4,

Pregnancy

259

%.

LE

-1

.,

I,,.

M
Iy.

in Early

: .,.#
%---t

.
,

Cells

#{149}

UL

#{149}.:#{248}#{163}#{149}

a..,

#{149}

\,

#{149}..

#{149}, r.

.,CCb

41;c..ft

dI:U

#{149}.im.

(L;S.%

.-?..-.;:q
!

#{149}j::

:h

#\

.4.

t8

r11i

Fig. 6. Cross section through uterus of 5 day pregnant mice.

Sections stained by immunoperoxidase
immunohistochemistry
with anti-CLA. Section is through the implantation
site. Note
positive cells (arrows) lining the uterine lumen (UL) at the
implantation
site. Some cells in the inner cell mass (double
arrows) also were positive. B, blastocyst;
LE, luminal epithehum; M, myometrium;
EN, endometrium.
x560.

;,

,#{149}:

at,.

:i

48

,,,_

4ld1tI

I,

#{149}

,..

beginning

on day

6.

The

formation

of the primary

decidua
results
in a shift in the distribution
of antigen
positive
cells in the endometrium.
As decidual
cells
proliferate,
CLAP
and F4-80
cells primarily
are detected in portions
of the uterus peripheral
to the decidua.
CLAP
and F4-80
cells no longer
are seen near the
implantation
site but are evident
in the deep endometrium
and within
the myometrium
(Fig. 7). Rare CLAP
and
F4-80
cells are observed
mixed
among
the decidual
cells. The trend detected
in the uterus on day 6 is more

obvious

on day 7. Nearly

all anti-mesometrial

CLAP

and

F4-80
cells are concentrated
in the deep endometrium,
and the number
of cells is substantially
decreased
from
day 6. A few isolated
positive
cells still are present within
the primary
and secondary
decidua.
On days 8 through
12, during which time the primary
and secondary
decidua
are regressing,
there are very few CLAP or F4-80
cells
seen in the anti-mesometrial
uterus.
During
the second
half of pregnancy,
epithelium
returns
to line the uterine
lumen,
and epithelial
cells again are a prominent
feature
of the anti-mesometrial
uterus.
In those areas, the uterine
lumen and luminal
epithelial
cells remain
intact. At that
time,
the anti-mesometrial
endometrium
and myometrium
contain
high numbers
of CLAP
and F4-80
cells (Fig. 8).

Distribution
of Antigen
Interimplantation
Sites

Positive

Cells in

Primary
and secondary
decidua
rapidly
the anti-mesometrial
uterus immediately

develop
following

within
blas-

1#{149}.

%.,

tp

.*

,,

Fig. 7. Section of uterus from 6 day pregnant mice. Section

stained by immunoperoxidase
immunohistochemistry.
Antimesometrial
uterus stained with F4-80. Note isolated positive
cells in secondary decidua (SD) and higher concentration
of
positive cells in deep endometrium
(EN) and myometrium
(M).
Control sections stained with anti-CD8 were negative. x560.
#{149}:
#{149},

,-

j:#{149}:

...--

..

rr

.t,_
q

.-#{149}

4,

..
.

,,

.;

..-

:.

.
-

-..

-.

_f

-.

#{149}

4,,,

-.-

#{149}1
-.

,1

#{243}fr:.

-;

.-

:.:

Fig. 8. Section of anti-mesometrial

uterus from 15 day pregnant mice. Section stained by immunoperoxidase
immunohistology with F4-80. Control sections stained with anti-CD8 were
negative. EN, endometrium;
LE, luminal epithelium;
M, myometrium; P, placenta. x560.

260

De et al.

Fig. 9. Section of uterus from 8 day pregnant mice. Section of interimplantation

site stained by
immunoperoxidase
immunohistochemistry
with F4-80. EN, endometrium;
LE, iuminal epithehum; M, myometrium.
x560.

tocyst implantation.

Decidua
development
is focal, leaving a portion of the uterus between
each implantation
site
free of decidual
cells. Interimplantation
sites contain high
concentrations
of CLAP
and F4-80
cells (Fig. 9).

DISCUSSION
Two

major
cellular
events
involving
macrophages
in the uterus
during
early pregnancy.
The first
appears
to be a classical-type
acute inflammatory
response
[16, 17]. The cells which contribute
to this early
inflammatory
response
are recruited
from
the blood.
Evidence
for this comes
from the fact that very little
proliferation
by endometrial
cells occurs
before
day 3
[1,4,6,7,19,20].
The response
is short-lived.
By day 3,
the only evidence
that an inflammatory
response
occurred
is that the number
of macrophages
is higher
than in the
non-pregnant
uterus.
At present,
it is not clear how this
response
is induced.
Its most obvious
effect is to eliminate semen from the uterine lumen through
phagocytosis
and degradation
by PMNs.
An additional
possible
role
may be to eliminate
microorganisms.
Since mating
is not
a sterile interaction,
the response
could have evolved
as
systematic
protection
against
microbial
invasion.

occur

Prior to implantation,
uterine
cells undergo
estrogen
and progesterone-dependent
proliferation
and differentiation,
which
are essential
for successful
implantation.
The data from the present
study demonstrate
that, at the
time
of implantation,
macrophages
are very
clearly
increased
in number
and appear
to be associated
with
glandular
and luminal
epithelium.
Their
subepithelial
localization
and numerical
increase
may be explained
by
the fact that estrogen
and progesterone
stimulate
uterine
epithelium
to make macrophage
colony
stimulating
factor (CSF-1)
[3,28],
which is specifically
chemotactic
for
macrophages.

Thus,

CSF- 1 released

from uterine

epithe-

hum could recruit macrophage

precursors
from the blood
and cause their migration
from blood vessels through
the
endometnum
into subepithelial
locations.
Also, the CSF1 could attract tissue macrophages
from one portion
of
the tissue to another.
These studies
do not exclude
the
possibility
that CSF- I is involved
in placental
development and trophoblast
proliferation
as originally
speculated [3 ,28]
but they do suggest
that uterine
CSF- 1
recruits
macrophages
to the uterus.
Implantation,
in combination
with appropriate
hormona! preparation
of the uterus,
induced
rapid endometrial cell proliferation
which is described
as decidualiza,

Inflammatory

tion.
Prior
to development
of the decidua,
uterine
macrophages
are very prominent
in the endometrium,
but
they are replaced
by the developing
primary
decidua.
Few macrophages
are found among the decidual
cells. It
is likely
that the increasingly
peripheral
locations
in
which macrophages
are found result from the fact that
destruction
of the uterine
epithelium
in the area immediately
surrounding
the developing
embryo
eliminates
a
major
source
for CSF- 1 . Macrophages
continue
to be
prominent
in interimplantation
sites, where the uterine
epithelium
remains
intact.
This study is consistent
with the numerous
studies
in
the literature
demonstrating
large numbers
of leukocytes

in the pregnant uterus [2,12,14,16,17,25,26,30,32,33,

34,36].
The function
(s) of inflammatory
cells in pregnancy remains
to be determined,
but their high numbers
suggest
that they play a significant,
if not essential
role.

ACKNOWLEDGMENTS
We would
like to thank
Dr. Robert
Klein and
Frank Holladay
for their help and encouragement.

gratefully

acknowledge

the technical

contributions

Dr.
We

of

Jean Selanders
and Grisel
Lopez.
We also appreciate
the excellent
photographic
help and advice
of Dennis
Friesen.
This research
was supported
by NIH grant
HD17678.

REFERENCES
I . Austin,
CR. , and Short, R.V. Hormonal
Control
of Reproduction. New York: Cambridge
University
Press,
1984.
2. Austyn,
J.M. , and Gordon,
S. F4/80;
a monoclonal
antibody
directed
specifically
against mouse macrophages.
Eur. J. Immunol. 10,805, 1981.
3. Bartocci,
A. , Pollard,
J.W. , and Stanley,
ER.
Regulation
of
colony-stimulating
factor-l
during
pregnancy.
J. Exp.
Med.
164,956,
1986.
4. Das, R.M. The time course
of the mitotic response
to estrogen
in
the epithelium
and stroma
of mouse
uterus.
J. Endocrinol.
55,203, 1984.
5. De, M. , and Wood,
G.W. Influence
of oestrogen
and progesterone on macrophage
distribution
in the mouse uterus.
J. Endocrinol.
126,417,
1990.
6. Finn, CA. , and Martin,
L. Endrocrine
control of gland proliferation in the mouse uterus.
Biol. Reprod.
8,585, 1973.
7. Finn,
CA. , and Martin,
L. The control
of implantation.
J.
Reprod.
Fertil. 39,195,
1974.
8. Head,
JR.
Lymphoid
components
in the rodent
uterus.
In
Immunoregulation
and Fetal Survival,
(Gill, Ti. , and Wegmann,
T.G. , his.). New York: Oxford
University
Press, p. 46, 1987.
9. Hume,
D. , Robinson,
A.P. , Macpherson,
G.S. , and Gordon,
S.
The mononuclear
phagocytic
system
of the mouse defined
by
immunohistochemical
localization
of antigen F4/80. J. Exp. Med.
158,1522,
1983.
10. Hume,
D.A. , Loutit,
iF. , and Gordon,
S. The mononuclear
phagocytic
system of the mouse defined by immunohistochemical
localization
of antigen
F4/80. Macrophages
associated
with bone
marrow
and connective
tissue. J. Cell. Sci. 66, 189, 1984.

Cells in Early Pregnancy

261

1 1 . Hume,
D.A. , Perry,
V.H. , and Gordon,
S. The mononuclear
phagocytic
system of the mouse defined by immunohistochemical
localization
of the antigen
F4/80.
Macrophages
associated
with
epithelia.
Anat. Res. 210,503,
1985.
12. Hunt, J.S., Manning,
L.S. , Selanders,
iS.,
and Wood,
G.W.
Localization
and characterization
of macrophages
in the murine
uterus.
J. Leukoc.
Biol. 38,255,
1985.
13. Kearns,
M. , and Lala, P.K. Bone marrow
origin of decidual
cell
precursors
in the pseudopregnant
mouse
uterus.
J . Exp. Med.
155,1537, 1982.
14. Keams,
M. , and Lala, P.K. Autoradiographic
analysis
of surface
markers
on decidual
cells shared
by cells of lymphomyeloid
tissues.
Am. J. Reprod.
Immunol.
Microbiol.
9,39,
1985.
15. Ledbetter,
J.A. , and Herzenberg,
L.A. Xenogeneic
monoclonal
antibodies
to mouse lymphoid
differentiation
antigens.
Immunol.
Rev. 47,63,
1979.
16. Lobel,
B.L. , Levy,
E. , and Shelesnyak,
MC.
Studies
on the
mechanism
of nidation XXXIV.
Dynamics
of cellular interactions
during progestation
and implantation
in the rat. Part I, progestation. Acta Endocrinol.
55,7,
1967.
17. Lobel,
B.L. , Levy,
E. , and Shelesnyak,
MC.
Studies
on the
mechanism
of nidation XXXIV.
Dynamics
of cellular
interactions
during
progestation
and implantation
in the rat. Part I, nidus
formation.
Acta Endocrinol.
55,47, 1967.
18. Martin,
L. , and Finn, CA. Hormonal
regulation
of cell division
in epithelial
and connective
tissues
of the mouse
uterus.
J.
Endocrinol.
41,363,
1968.
19. Martin,
L. , and Finn, CA.
Duration
of progesterone
treatment
required
for a stromal response
to estradiol
17b in the uterus of the
mouse.
J. Endocrinol.
44,279,
1969.
20. Martin,
L. , Finn,
CA. , and Trinader,
0. Hypertrophy
and
hyperplasia
in the mouse
uterus after oestrogen
treatment:
An
autoradiographic
study. J. Endocrinol.
56, 133, 1973.
21. Mathews,
C.J. , Adams,
AM. , and Searle,
R.F. Detection
of
macrophages
and the characterization
of Fc receptor
bearing
cells
in the mouse decidua,
placenta
and yolk sac using the macrophage-specific
antibody
F4/80. J. Reprod.
Immunol.
7,3 15, 1985.
22. McLean,
lW. , and Nakane,
P.K. Periodate-lysine-paraformaldehyde fixative.
A new fixative for immune
electron
microscopy.
J. Histochem.
Cytochem.
22,1077,
1974.
23. Milstein,
C. , Galfre,
G. , Secher,
D.S. , and Springer,
T. Monclonal antibodies
and cell surface antigens.
Cell. Biol. mt. Rep.
3,1, 1979.
24. Noun,
A. , Acker,
G.M. , Chaouat,
G. , Antoine,
iC. , and
Garabedian,
M. Cells bearing the granulocyte-macrophage
and T
lymphocyte
antigens
in the rat uterus before
and during
ovum
implantation.
Clin. Exp. Immunol.
78,494,
1989.
25. Papanicolaou,
G.N.
Observations
on the origin
and specific
function
of the histiocytes
of the female genital tract. Fert. Steril.
4,472,
1953.
26. Peel, S. , Stewart,
Ii. , and Bulmer,
D. Experimental
evidence
of
the bone marrow
origin of granulated
metrial gland cells of the
mouse uterus. Cell. Tissue Res. 233,647,
1983.
27. Perry, V.H. , Hume, D.A. , and Gordon,
S. Immunohistochemical
localization
of macrophages
and microglia
in the adult and
developing
mouse brain. Neuroscience
15,313,
1985.
28. Pollard,
J.W. , Bartocci,
A. , Arceci,
R. , Orlofsky,
A. , Ladner,
MB.
, and Stanley, ER. Apparent role of the macrophage
growth
factor, CSF-l , in placental
development.
Nature Lond. 330,484,
1988.
29. Redline,
R.W. , and Lu, C.Y. Specific
defects in the anti-listerial
immune
response
in discrete
regions
of the murine
placenta
account
for susceptibility
to infection.
J. Immunol.
140,3947,
1988.
30. Redline,
R.W. , and Lu, C.Y. Localization
of fetal major histo-

262

31

32.
33.

Deetal.
compatibility
complex
antigens
and maternal
leukocytes
in muring
placenta:
Implications
for maternal
fetal immunological
relationship. Lab. Invest.
61 ,27, 1989.
Springer,
T. , Galfre,
G. , Secher,
D.S., and Milstein,
C. Monoclonal
xenogeneic
antibodies
to murine
cell-surface
antigens:
identification
of novel leukocyte
differentiation
antigens.
Eur. J.
Immunol.
8,539,
1978.
Tachi,
C. , and Tachi,
S. Macrophages
and implantation.
Ann.
N.Y. Acad. Sci. 476,158, 1986.
Tawfik,
OW. , Hunt, iS. , and Wood,
G.W.
Partial characterization of the uterine cells responsible
for suppression
of murine

34.

35.

36.

maternal
anti-fetal
immune
responses
.
J . Reprod . Immunol.
9,213, 1986.
Wegmann,
T.G.
Placental
immunotrophism:
maternal
T cells
enhance placental
growth and function . Am . J . Reprod . Immunol.
15,67,
1987.
Weitlauf,
H.M. Biology
of implantation.
In The Physiology
of
Reproduction.
Vol. 1 . , (Knobl,
E. , and Neil, K. , Eds.).
New
York: Raven Press, p. 231 , 1988.
Wood, G.W. Immunohistologic
identification
of macrophages
in
murine
placenta,
yolk sac and pregnant
uteri. Placenta
1,309,
1980.