| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) expression of the leukocyte integrins has been
implicated in their adhesion and migration in the adult. Little is
known, however, of the expression or function of these molecules during
development. This study defines the spatial and temporal sequences of
expression of the type 3 complement receptor (CR3) in the developing
mouse; establishes the functional efficacy of this molecule in
spreading, adhesion, and phagocytosis; and investigates its role in
inflammatory and constitutive migration. Expression of CR3 on monocytes
occurred early compared to M-restricted glycoprotein F4/80, but
expression on stellate tissue M appeared later than F4/80 and was
transient. Expression of CR3 on resident tissue M is more widespread
during development, being retained on only very specific M
populations in the adult. Neutrophil polymorphs expressed CR3 from day
17 of gestation onward. The anti-CR3 mAb 5C6 was used to
investigate the role of CR3 in adhesion, spreading, and phagocytosis by
neonatal M. Neonatal macrophages were found to adhere, spread, and
phagocytose by CR3-dependent mechanisms, and a CR3-independent system
was implicated in the spreading of neonatal M. The role of CR3 in
migration during development was then investigated. 5C6 had potent
effects on the early stages of the migration of myelomonocytic cells to
an inflammatory stimulus in vivo. Despite efficient transplacental
transfer of the Ab from pregnant mother to fetus, the process by which
monocytes generate populations of resident tissue M was undisrupted,
indicating the existence of CR3-independent mechanisms of monocyte
migration during development. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
)3 and their
environment during murine development (1). Cellular and soluble signals
coordinate the differentiation of stem cells into monocytes, which
enter the blood to form a pool of migration competent cells and
subsequently enter every organ in the body to form heterogeneous
populations of tissue M (2, 3, 4). The differentiation Ag F4/80 is
first expressed in the 10-day-old embryo on monocytes and some stellate
M in the yolk sac, and then on M populations in the liver (day
11), spleen (day 12), and bone marrow (day 17). Monocyte infiltration
and local proliferation also result in the appearance of abundant
F4/80+ M in mesenchymal organs during this period (5).
However, it is not clear whether this temporal progression of monocyte
appearance in various organs represents serial migration or whether
each organ is seeded from common sources of stem cells. Furthermore,
while previous studies have used mAb F4/80, which recognizes a
M-restricted plasma membrane glycoprotein (6), to describe the
distribution of M during development (5), the signals and receptors
involved in the entry, migration, and exit of monocytes within
developing tissues are not known.
In the adult mouse, constitutive migration of monocytes from the bone
marrow maintains most resident tissue M populations (7) and
continues alongside inflammatory and immunologically driven recruitment
processes in which a family of structurally and functionally related
surface glycoproteins, the leukocyte ß2 integrins (LFA-1,
CR3, and p150,95) (8), have been implicated (9, 10, 11). Each of these
molecules consists of an antigenically distinct 
-chain (150–190
kDa) noncovalently associated with a ß-chain (95 kDa) as an
1ß2 dimer (reviewed in 12 .
Congenital inability to express the ß2 chain results in
an impairment of surface expression of all three heterodimers, severe
defects in leukocyte accumulation in vivo and adherence in vitro, and
an increase in life-threatening infections (13). However, the role of
these molecules in the constitutive or inflammatory migration of
monocytes during development is unknown.
Similarly, while the distribution of leukocyte integrins in the adult
is well documented (14, 15), little is known of their expression during
ontogeny. In the adult, CR3 is expressed on monocytes, neutrophilic
polymorphonuclear leukocytes (PMN), and NK cells in blood and
hemopoietic tissues, but not on populations of stellate tissue M,
except microglia in the brain, marginal zone M in the spleen, and
M in the subcapsular sinus of the lymph node (16). This suggests
that CR3 expression is widely down-regulated as a consequence of
constitutive entry of monocytes into tissues and differentiation into
resident M such as Kupffer cells and implies a possible role for CR3
in development. In contrast, high levels of CR3 expression are retained
on M that have been recruited to tissues in response to inflammatory
signals.
The aim of this study, therefore, was to analyze the spatial and
temporal sequences of CR3 expression in fetal and neonatal tissues as a
first step to define its role in constitutive and inflammatory monocyte
recruitment during development. Subsequently, mAb 5C6, which blocks
CR3-dependent adhesion of M to bacteriologic plastic (BP) and has
potent effects on some forms of myelomonocytic migration in vivo
(17, 18, 19, 20), also provided a useful tool to determine the ability of CR3
on isolated neonatal M to mediate adhesion, spreading, and
phagocytosis. Having then established the efficacy of neonatal CR3 in
vitro, its role in the processes of constitutive and inflammatory
myelomonocytic migration during development was also investigated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Embryos and newborn mice (Pathology Oxford (PO)) were bred at the Sir William Dunn School of Pathology (Oxford, U.K.) and maintained under conventional laboratory conditions with free access to food and water. Females were inspected daily for the presence of a vaginal plug, which was designated day 0 of pregnancy. The day of birth (usually day 19 of pregnancy) was designated day 0 of neonatal life.
Antibodies
The following rat mAbs were used as hybridoma supernatants for immunohistochemistry. F4/80 (21) recognizes a 160-kDa Ag of unknown function. M1/70 (22) and 5C6 (18) were used in combination at a final dilution of 1/10 in PBS to recognize the CR3. Biotinylated rabbit anti-rat IgG were obtained from Vector Laboratories (Peterborough, U.K.). Polyclonal Abs recognizing F4/80 and sialoadhesin were prepared in our laboratory by Drs. P. Dri and P. Crocker and were routinely used at a dilution of 1/500 for immunohistochemistry.
Immunocytochemical techniques
Fixation and sectioning of murine organs. Embryos from day 8 to birth and neonatal animals up to 2 wk after birth were examined. Visceral yolk sac, liver, spleen, bone marrow, thymus, lung, kidney, and gut were dissected from conceptuses as they became readily identifiable. Organs were washed in PBS before freezing in OCT embedding medium (Miles, Elkhart, IN) cooled in isopentane (BDH-Merck, Poole, U.K.) over liquid nitrogen. Frozen sections were cut on a cryostat (Leica, Wetzlar, Germany) at 5 µm, air-dried for 1 h, and frozen at -20°C until further use. Fresh organs were fixed for 10 min in 2% paraformaldehyde in HEPES-buffered isotonic saline before staining; 2 mM calcium chloride was added to the fixative to maintain the structural conformation of the integrin.
Ag detection. Fixed sections were washed in PBS containing 0.1% (v/v) Triton X-100 (BDH-Merck) and treated with 2% normal rabbit serum for 30 min. Sections were incubated for 90 min in hybridoma supernatant, PBS, or isotype-matched control mAb. Endogenous peroxidase activity was inactivated by incubation of sections with 10-2 M glucose, 10-3 M NaN3, and 40 U glucose oxidase in 100 ml 0.1 M phosphate buffer for 15 min at 37°C (23). Affinity-purified, mouse-adsorbed, biotinylated second Ab (Vector) was used at 1% for 45 min, and avidin-biotin-peroxidase complex (ABC elite, Vector) (24) was used according to the supplier’s recommendation. The presence of Ag was revealed by incubation with 0.5 mg/ml 3,3'-diaminobenzidine HCl (Polysciences, Northampton, U.K.) and 0.024% H2O2 in 10 mM imidazole in PBS, pH 7.4.
Double labeling. To detect a second Ag, polyclonal Abs were incubated on sections for 90 min followed by 1% alkaline phosphatase-conjugated goat anti-rabbit second Ab for 60 min (Vector). The Ag was revealed by incubation with a 5-bromo-4-chloro-3-indolylphosphate/nitro blue tetrazolium alkaline phosphatase substrate kit (Vector), and sections were counterstained in cresyl fast violet acetate (BDH-Merck) and mounted in DPX (BDH-Merck). Sections in which primary Ab, secondary Ab, or avidin-biotin-peroxidase complex reagent were omitted and sections treated with irrelevant control Abs showed no staining.
In vivo detection of Ags.
Pregnant mice were injected i.v. with 0.5 mg of 5C6 mAb or
isotype-matched control mAb, 1C5 (which does not react with the surface
of murine leukocytes) (25), on days 14 and 17 of pregnancy. Newborn PO
mice were injected i.p. with 0.05 mg of purified rat mAb. Offspring of
the pregnant mice or the newborn mice injected directly were
subsequently fixed by perfusion with 2%
periodate-lysine-paraformaldehyde solution (PLP) (26). After postfixing
in the same solution for 4 h, tissues were impregnated with 20%
sucrose in 0.1 M phosphate overnight before freezing and sectioning, as
described. No further fixation of these sections was required.
Affinity-purified, mouse-adsorbed, biotinylated anti-rat Ab
(Vector) was used to locate injected rat Abs, and polyclonal rabbit
anti F4/80 antiserum was used to define the distribution of
F4/80+ monocytes or M in these sections.
Photography. Representative black and white photographs were taken using a blue filter (Wratten Gelatin Filter no. 47, Kodak, Rochester, NY) that intensifies the brown precipitate. Color photographs were taken using a 15-cc magenta filter.
Neonatal peritoneal M.
Resident peritoneal M (RPM) were obtained from the cavities of
3-day-old PO mice by lavage with 1 ml of PBS. Recruited cells were
similarly harvested from 3-day-old PO mice that had been injected i.p.
with 25 µl of Brewer’s complete thioglycolate broth on the day of
birth.
Adhesion to artificial substrata (18).
Cells to be assayed for adhesion were resuspended in RPMI 1640 with
10% FBS and plated at a density of 3 x 105
M/well in flat-bottom BP or tissue culture plastic (TCP) 96-well
plates (Flow Laboratories (Rickmansworth, U.K.) and Nunc/Life
Technologies (Paisley, U.K.)). mAb (5 µg/ml) alone or in combination
with divalent cation chelator (5 mM EDTA) were added to test wells and
incubated for 30 min at 4°C. After incubation for 90 min at 37°C,
plates were washed three times in PBS, and adherent cells were fixed in
methanol and stained with 40% Giemsa solution for 1 h. Plates
were washed in tap water, and the retained dye was solubilized in
methanol and quantified by measuring absorbance at 450 nm in an
automatic plate reader (Anthos III, Denley Instruments, Billinghurst,
U.K.). In some assays cell viability was assessed at each stage using
trypan blue dye exclusion (Sigma, Poole, U.K.).
Spreading assays
Cells were resuspended in RPMI 1640 plus 10% FBS and plated at
a density of 105 M/chamber in multichamber slides.
mAb (2 µg/ml 5C6 or control 1C5) or chelator (5 mM EDTA) was added to
test chambers and retained throughout the experiment. Slides were
washed in PBS, and adherent cells were fixed in 0.25% glutaraldehyde
at 30 min or 1, 3, 6, 10.5, or 24 h. The number of cells with a
spread morphology was assessed using phase contrast microscopy and
expressed as a percentage of the total adherent cells.
Phagocytosis of opsonized red cells
Sheep erythrocytes (Becton Dickinson, Oxford, U.K.; stored in
Alsever’s solution at 4°C until used) were incubated with 30% mouse
IgM anti-sheep erythrocyte in PBS (Nordic Immunologic Laboratories,
Tilburg, The Netherlands) for 45 min at 4°C, followed by fresh mouse
serum for 30 min at 37°C to fix iC3b. Opsonized sheep erythrocytes
(EAiC3b) were used at a 5% (v/v) suspension for 60 min to assay for
rosetting at 4°C or phagocytosis at 37°C (27, 28). Binding was
quantified by counting the number of erythrocytes attached to 100 M.
Noninternalized erythrocytes were lysed in 0.147 M
NH4Cl before counting the number of erythrocytes
ingested per 100 M.
LPS-induced myelomonocytic recruitment
Newborn PO mice were injected intradermally with 10 µl of PBS or LPS (0.2 µg) containing 1% Monestral blue to mark the site of the lesion and at the same time i.p. with 0.05 mg of purified 5C6 or isotype-matched control mAb (1C5). Animals were perfusion fixed with PLP at 1, 4, 8, 10. 5, 24, and 56 h after injection and processed as described above to reveal injected mAb and F4/80 on recruited cells.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To establish a basis for manipulation of myelomonocytic cell
migration during the ontogeny of the mononuclear phagocyte system, the
expression of CR3 on monocytes, M, and PMN during murine development
was investigated. Tissue samples were taken throughout fetal and
neonatal life and stained using a combination of two mAbs, M1/70 and
5C6, which show differential sensitivity to digestion of the cell
surface with pronase and are therefore believed to recognize
different epitopes of CR3 (18). No staining was observed with the
irrelevant isotype-matched mAbs CAMPATH-1G or 1C5. Organs
from at least two mice from each of three independent litters were
examined at each developmental stage, and in each case a comparison was
made with consecutive sections stained with the F4/80 polyclonal
Ab.
Lymphohemopoietic organs
Liver (Fig. 1
).
The liver primordium develops relatively earlier in the mouse than in
man and is visible at only 9 days of gestation, when it is composed of
broad hepatic cords separated by large sinusoids containing nucleated
erythroblasts (Eb). CR3+ monocytes were found in the
fetal liver of whole embryos from this time onward, whereas
F4/80+ was only detected on monocytes from day 10 of
gestation (day 10e; not shown). The density of CR3+
monocytes in liver increased as hemopoiesis progressed from day 11e
(Fig. 1a) and by day 13e, when the liver is well
developed, labeled monocytes could be seen in clusters around liver
sinusoids (Fig. 1b). Hemopoietic foci were found
intermingled within the hepatic cords, but unlike the stellate M
detected by F4/80, CR3+ monocytes were not associated with
developing Eb. The F4/80+ stellate M first exhibited CR3
expression on day 15 (Fig. 1c). By day 17e,
CR3+ PMN were apparent and located in clusters of erythroid
and myeloid cells around a CR3+ stromal M (Fig. 1d). CR3+ PMN and monocytes increased in
number until shortly after birth (Fig. 1, e and
f) and continued to be distributed unevenly
throughout the liver, with high numbers of cells adjacent to the
sinusoids. CR3 staining of stellate M was maximal around 8 days
after birth (Fig. 1g) even though at this stage the
M were no longer associated with remaining erythroid cells.
Hemopoietic clusters containing CR3+ monocytes were rare by
this stage, and mature M accounted for most of the CR3 staining.
After birth, declining hemopoietic activity in the liver was associated
with a decreased number of CR3-staining monocytes and PMN, while CR3
staining of mature M progressively decreased in intensity. In the
adult, CR3 expression on spindle-shaped sinus-lining M (Kupffer
cells) was much reduced or absent (Fig. 1h), in
contrast to F4/80, which is highly expressed on mature Kupffer cells in
the liver (not shown).
|
).
By 11 days of gestation, the epithelial lining of the dorsal mesentery
of the stomach is thickened and represents the anlage of the spleen.
CR3+ monocytes first appeared among the cells of this
rudimentary spleen on day 12e (Fig. 2, a and
b), at the same time as F4/80+ expression
was first detected on splenic monocytes. Ramified F4/80+
M appear on day 15e and are surrounded by developing red cells, but
at this stage these cells do not express CR3 (Fig. 2c). From day 16e (Fig. 2d) the
splenic architecture becomes more complex, with distinct arteries and
veins and regions destined to become lymphoid areas, containing none of
the monocytes or numerous CR3+ granulocytic cells that are
then apparent in the developing red pulp (Fig. 2e).
CR3+ monocytes and PMN continue to be detected in the red
pulp immediately after birth and during adult life (Fig. 2, f and g). In contrast, CR3+ stellate
M were most evident in the expanding marginal zone of the white pulp
1 wk after birth (Fig. 2h), and in the adult could be
detected both in the marginal zone and occasionally around the central
arteriole in the white pulp (Fig. 2i). The network of
stromal M expressing F4/80 in the red pulp never expressed
CR3.
|
and monocytes within and along
vessels of the vitelline yolk sac from the ninth day of gestation (day
9e; Fig. 3a), which is
1 day earlier than F4/80 expression has been observed on cells seen
mainly in the yolk sac mesoderm; CR3 is also expressed on more stellate
M within vessels than F4/80. The density of CR3 staining peaked on
day 10e, when nucleated Eb were also seen within the blood spaces. At
no stage were M associated with clusters of Eb.
|
invade the
cartilaginous rudiment of bone on day 16e, resulting in the formation
of the marrow cavity. CR3+ monocytes and PMN were evident
within the stroma of the bone marrow from day 17e, when hemopoietic
activity is initiated (Fig. 3
b). Clusters of
erythroid and CR3+ myeloid cells surround central
F4/80+ M during development and in the adult mouse, but
CR3 staining was only observed on the stromal M of the bone marrow
from day 17e until shortly after birth and, unlike F4/80, was absent
from these M in the adult as previously reported (29). Nonhemopoietic tissues
F4/80+ and CR3+ monocytes were seen
throughout the undifferentiated tissue within which organs develop. As
organogenesis proceeds, F4/80+ M persist to become
components of loose connective tissue of differentiated organs, whereas
CR3 expression is variable depending on the organ.
Gut.
CR3 was first detected on spindle-shaped M around the capsule of the
gut on day 11e. One day later, when the lumen became visible,
CR3+ M were still confined to the capsule, whereas
F4/80+ M appeared in the lamina propria at this stage.
From day 13e CR3+ monocytes appeared in the lamina propria
circumferential to the developing villi, and on day 15 mature M were
visible within the villi. The pattern of moderate CR3 expression on the
stellate M within the villi and strong expression on monocytes of
the lamina propria continued throughout development and in the adult
(Fig. 3, c and d).
Lung.
The lung anlage develops on day 9e as rudiments of
larynx-trachea-bronchi. CR3+ M, with extensive
processes, were evident around the outer margins of the mesenchyme of
the developing lung from day 11e, at the same stage that
F4/80+ cells were observed randomly distributed within the
mesenchyme. By day 13 the lungs were clearly subdivided into lobes, and
by day 14 the disappearance of mesenchyme from the lung was accompanied
by the formation of airways and vessels with which some
CR3+ monocytes and more spindle-shaped M were associated
(Fig. 3e). CR3+ cells, often in pairs,
and F4/80+ M were distributed throughout the
intraalveolar septa of the lung from just before birth (Fig. 3f), whereas the alveolar M populations of the
adult failed to express CR3 or F4/80.
M of the kidney failed to express CR3 at any point during
development, whereas F4/80 was expressed on M between the tubules
from day 12e.
Thymus.
The thymus becomes visible as epithelial lobes surrounded by a thin
outer capsule of mesenchyme at 13 days of gestation. Flattened
F4/80+ cells can be seen in the thymus from this time.
As the organ grows, the F4/80+ M increase in number and
become more stellate, the majority of M were found in the cortical
regions among developing thymocytes. By contrast, CR3+
monocytes could only be detected in the vascularized corticomedullary
region of the thymus from day 16e onward when the thymus increases
strikingly in volume and the flattened F4/80+ M do not
express detectable levels of CR3 at any point during development or in
the adult (Fig. 3, g and h).
Peritoneal cells
Resident and elicited neonatal peritoneal M were immunoreactive
for CR3 (not shown).
The immunohistochemistry outlined above established that CR3 was
expressed on monocytes and tissue M during development, but was
gradually lost after birth from mature tissue M in many regions. The
pattern of CR3 expression differed from that of F4/80 both spatially
and temporally (summarized in Table I).
CR3 was only retained on the marginal zone M of the spleen and the
subcapsular M of the lymph nodes (not shown), which may be involved
in trafficking of leukocytes into these organs in the adult. This was
suggestive of a specific role for CR3 during the developmental period
and was consistent with a possible role for CR3 in the adhesion or
migration of monocytes into and within organs before they differentiate
into resident tissue M. It was impossible to obtain neonatal
peripheral blood monocytes in sufficient numbers to assess their use of
CR3 in vitro; however, in the adult RPM have been successfully used as
surrogate monocytes in adoptive transfer assays of migration (25).
Isolated neonatal peritoneal M were therefore used to establish the
functional competence of the CR3 expressed during murine development.
The spreading, adhesion, and phagocytosis of neonatal peritoneal M
were assessed at a stage when there was still significant expression of
CR3 on the stromal M of spleen and liver.
|
mAb 5C6 was initially selected for its ability to inhibit the
adhesion of adult M to BP by blocking the interaction between CR3
and an undefined component of FBS that coats the plastic surface (18).
This interaction is also inhibited by EDTA, since the receptor function
of integrins depends on the presence of divalent cations (30, 31). In
this study, short term adhesion assays to BP were performed with
neonatal and adult peritoneal cells lavaged from the resting peritoneal
cavity or after its stimulation 3 days previously with thioglycolate
broth. The comparative adhesion qualities of the cells and the effects
of chelation of divalent cations or preincubation of the cells with mAb
5C6 are shown in Figure 4, A
and B. The adhesion of neonatal peritoneal M to BP, like
that of adult M, was divalent cation dependent and almost completely
inhibited by 5C6. The CR3 of neonatal cells is therefore functional in
adhesion to BP. However, in contrast to the adult where
thioglycolate-elicited peritoneal M (TPM) were more adherent than
RPM, neonatal RPM adhered more readily to BP than neonatal
TPM.
|
The interaction of M with more complex substrata such as serum
coated glass or TCP can be described in the two phases: adhesion and
spreading. 5C6 and EDTA, which inhibit M adhesion to BP, reduce the
spreading of adult cells on glass and TCP, but fail to inhibit their
adhesion. The kinetics of spreading of adult and neonatal peritoneal
M on glass and its dependence on CR3 were assessed using a method
previously described by Haynes et al. (32). Neonatal RPM spread more
rapidly than adult RPM, such that 100% adherent neonatal cells had an
extensively arborized morphology after only 3 h of incubation
compared with 24 h for the adult RPM (Fig. 4C).
Chelation of divalent cations inhibited the spreading of neonatal RPM
on TCP (p < 0.01), whereas blocking CR3 with
5C6 mAb up to the 6 h point did not significantly inhibit their
spreading on this surface. This indicates that the spreading of
neonatal M involves divalent cation-dependent receptors other than
CR3 or a conformation of CR3 not recognized by 5C6 (Fig. 4D). Since the proportion of adult M fully spread
at the 6 h point was small (only 18.7%), the inhibitory effect of
5C6 and EDTA (20 and 58%) did not reach statistical significance. This
time point was chosen to fall on the up-slope of spreading for both
cell populations, thus potentially allowing both positive and negative
effects of the Ab to be observed. The inhibitory effect of 5C6 and EDTA
on adult cells at time points that allow full spreading has been well
documented (18).
Role of CR3 in phagocytosis by neonatal M
In addition to adhesion and spreading, CR3 is known to mediate
binding and ingestion of complement-opsonized particles by adult
peritoneal M (33, 34). Figure 5
summarizes the binding and phagocytosis of iC3b EAiC3b by neonatal and
adult resident and recruited M. As previously documented (34), adult
TPM were able to both bind and ingest EAiC3b, whereas RPM bound
opsonized particles well, but ingested relatively few. Similarly, in
the neonate both RPM and TPM bound EAiC3b, but ingestion of EAiC3b was
greater by TPM than by RPM. In each case EAiC3b rosetting was inhibited
by 5C6, indicating that the CR3 on both the resident and elicited
neonatal peritoneal M is capable of binding iC3b and to some extent
mediates ingestion of EAiC3b.
|
Since the epitope of CR3 recognized by 5C6 and implicated in adult myelomonocytic cell recruitment is present and active in adhesion and phagocytosis in the neonate, the ability of 5C6 to block inflammatory recruitment of cells in the neonate was examined. It is well documented that thioglycolate broth elicits a 10- to 20-fold influx of myelomonocytic cells into the adult peritoneal cavity 2 to 4 days after injection. Recruitment of myelomonocytic cells to the neonatal peritoneal cavity was examined by i.p. injection of sterile thioglycolate broth. However, in contrast to the adult, there was no detectable increase in the total number of cells within the peritoneal cavity over the 7 days after thioglycolate injection. Furthermore, the developmental, weight-related increase in the number of cells in the cavity was suppressed by thioglycolate (not shown). Since injection of control Ab, PBS, or 5C6 also prevented any developmental increase in cell number in the cavity, it is likely that any injection results in a sufficient inflammatory stimulus to induce the resident cells to leave or die within the cavity. For this reason an alternative model of inflammatory recruitment in neonatal mice was employed. It has been established that an acute inflammatory response in adult murine skin may be elicited by LPS. After intradermal injection, myelomonocytic recruitment into adult skin occurs within 2 h (35). This model was therefore adapted for the neonate.
One-day-old mice were injected i.p. with control Ab or 5C6 and intradermally with PBS or LPS, including monestral blue to mark the site of injection. Animals were fixed by PLP perfusion at various times, and the lesions were analyzed immunohistochemically.
One hour after injection, many CR3-labeled cells were found at the
lesion, but few could be identified as monocytes using polyclonal
F4/80. Cell recruitment was increased at 4 h; it was maximal
8 h after injection and consisted mainly of PMN (Fig. 6, a and b).
By 24 h the total numbers of recruited cells had started to
decline, and the infiltrate became predominantly monocytic in character
(Fig. 6, c and d). The early, largely
neutrophil, response was inhibited by the i.p. injection of 5C6 mAb.
Figure 6, E and F, shows the absence of
F4/80+ monocytes/M or CR3+ monocytes and PMN
at the 24 h lesion. The later phase of monocytic recruitment
appeared to be at least partially 5C6 resistant (Fig. 6, e–h), with both CR3- and F4/80-staining cells
detectable in the vicinity of the LPS injection after 56 h. This
result confirms the activity of CR3 in the inflammatory recruitment of
neonatal PMN and monocytes and also suggests that, as in some models in
the adult, other adhesion molecules are also involved.
|
We next investigated whether the migration of cells
expressing CR3 during prenatal development could be affected by 5C6. It
was essential to know that the Ab could reach sites of expression if
its effects on monocyte migration were to be assessed. Animals were
injected with mAb i.v. on days 14 and 17 of pregnancy, and their
offspring were fixed by perfusion 1 day after birth. 5C6 could be
detected on M in the neonatal spleen (Fig. 7a), lamina propria of
the gut (Fig. 7b), and Kupffer cells in the liver
(Fig. 7c) and had therefore crossed the placenta.
When Ab was injected i.p. into newborn mice, followed by PLP
perfusion-fixation of the whole animal, 5C6 could be detected in the
liver, thymus, and spleen 1, 4, 6, and 9 days after injection using a
rat monoclonal detection system. In control animals that received PBS
or isotype-matched control Ab, no in vivo labeling of cells could be
detected (not shown). In animals injected with 5C6, PMN, monocytes, and
some tissue M were strongly labeled in the developing hematolymphoid
organs. Monocytes and marginal zone M were labeled in the spleen
(Fig. 7d), monocytes were labeled in the
corticomedullary region of the thymus (Fig. 7e), and
Kupffer cells and monocytes were labeled in the liver (Fig. 7f). In the nonhematolymphoid organs of gut, brain,
and skin, CR3+ cells were detected in architectural
arrangements similar to those of conventionally stained specimens (not
shown).
|
populations. (The detection
system employed to locate the rabbit polyclonal Abs did not bind to the
injected rat mAb.) No morphologic disruption or interference with the
expression of F4/80 or sialoadhesin was observed over the extended time
course over which the injected Ab could still be detected. In a number
of Ab-treated animals there appeared to be an increased frequency of
mitoses in the neonatal liver and spleen and an enhanced sinusoidal
location of monocytes in the liver. Figure 7g shows pF4/80
staining in the spleen, and Figure 7h shows Kupffer cells in
the liver 9 days after 5C6 administration, illustrating the morphologic
integrity of the organs after Ab injection. This pattern of staining
was identical with that detected in animals preinjected with control Ab
and with that detected in native animals by Morris et al. (5).
Therefore, while 5C6 exhibits the capacity to cross the placenta and be
retained on the surface of myelomonocytic cells for several days after
birth, such intervention does not interfere with the ability of the
cells to find their organ-specific locations. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
were distinguished on account of their round or stellate
morphology and frequent presence of phagocytic material in M.
Neutrophils, which only became evident on day 16 or 17 of gestation,
were small with a multilobed nucleus. Flow cytometric analysis has
previously demonstrated large populations of CR3-positive myeloid cells
in the fetal liver from day 11 of gestation, increasing up to birth and
decreasing from the first to third weeks after birth (37). However, the
study in question does not detect heterogeneity of expression within
individual organs or distinguish between monocytes produced as a result
of hemopoietic activity at the height of the CR3 expression and mature
resident tissue M. Here we have compared the immunohistochemical
distribution of CR3 during development with that of F4/80 (Table I) to
analyze the expression of a known adhesion molecule on a population of
developing cells identified by an independent marker. The expression of
CR3 on monocytes within most developing organs preceded the appearance
of F4/80+ monocytes or M, but the expression of CR3 on
stellate tissue M appeared later than F4/80 and was transient, often
being lost shortly after birth. The expression of CR3 on resident
tissue M is therefore more widespread in the fetus and neonate than
in the adult, except on very specific populations of adult M (Table II). LFA-1 was detected on monocytes and
M at the same stage as CR3 in most organs, but was retained during
development and in the adult (not shown). This raises interesting
questions about the regulation of CR3 expression during development,
its function on monocytes and M in the embryo and neonate, and the
significance of its down-regulation in the adult. It may be that its
function is specific to an early window of monocyte/M life cycle and
is lost as a consequence of monocyte differentiation within tissues.
For example, CR3 expression on monocytes may be related to their
capacity for migration and may no longer be required once a M is
fixed within a tissue. CR3 expression on mature M in the lymph node
and spleen occurs in areas of leukocyte entry into these organs (38);
elimination of marginal zone M in the spleen reduces the
accumulation of lymphocytes in this area (39), suggesting that these
M have an active role in the trafficking of other cells.
Furthermore, adoptively transferred RPM are first evident in the
marginal zone of the spleen and subcapsular sinus of the lymph node
(25), so that CR3 expressed in these regions might be related to newly
extravasated cells that have not yet assumed a tissue phenotype.
|
isolated from the peritoneal cavities of neonatal mice
was functional in in vitro assays of adhesion to BP and could be
completely inhibited by mAb 5C6 or EDTA. However, the spreading of
neonatal RPM, which was more rapid and extensive than that of adult
RPM, could not be completely inhibited by 5C6. Integrins on leukocytes
are normally inactive, in that they bind with only very low avidity to
their ligands but can be triggered remotely by other membrane receptors
or their own ligands to an activated conformation (40, 41). Further
studies are required to establish whether the spreading of neonatal RPM
is due to an activation-specific epitope of CR3 or to a different
molecule. Enhanced spreading may be a reflection of recent
extravasation from the blood to form the resident population or may be
due to the presence of mediators within the newborn cavity, such as
IFN-
(42) and TNF (43) or macrophage CSF (44), which are known to
enhance the spreading of murine adult RPM in vitro.
To investigate the role of CR3 in inflammatory migration in neonates,
LPS was introduced into the skin of neonatal mice. The initial phase of
LPS-induced myelomonocytic recruitment in the neonate was blocked by
mAb 5C6. However, despite the continued presence of 5C6-labeled M at
the lesion, a later phase of myelomonocytic recruitment was still
evident and therefore was independent of CR3. Analogous studies in the
adult indicate that T cell-dependent inflammation induced by specific
antigenic challenge in sensitized mice is also biphasic and is
inhibited by 5C6 only in its initial phase even though the Ab is
present and functional for 4 or more days (20) and recruitment of cells
to pulmonary granulomata following Calmette-Guérin bacillus
infection or to hepatic sinusoids in murine malaria are mostly
independent of CR3-mediated mechanisms (9). The present study shows
that induced monocyte migration in the neonate also has CR3-dependent
and -independent components.
Administration of 5C6 to pregnant or newborn mice resulted in labeling
of the CR3 Ag in vivo and demonstrated that mAb 5C6 was able to cross
the placenta. This is of interest since 5C6 is also able to cross the
blood-brain barrier (45), whereupon it induces mitoses and apoptosis of
microglia. Low levels of mitosis were also observed in neonatal spleens
and thymi after injection of this Ab, indicating that ligation of CR3
in the periphery may also stimulate cell division, but further
quantitative studies are needed to establish whether this was enhanced
relative to the background in the present model. 5C6 failed to disrupt
constitutive monocyte migration to nonhematolymphoid organs or to
affect myelomonocytic cell distribution in hematolymphoid organs. The
expression of other differentiation markers, including F4/80 and
sialoadhesin, was unaffected by the administration of 5C6 at or before
birth. A variety of Abs did not detect any significant disruption of
M distribution in tissues by 5C6 up to 9 days after its original
administration. This is in keeping with studies indicating that
migration of fluorescently labeled M from the blood to normal tissue
in the adult is CR3 independent (25) and that normal M populations
are found in CR3-deficient human patients to the extent examined (13).
The development of mice deficient in CR3 is also grossly normal;
however, details of M distribution during the embryonic period of
these animals have not been reported (46).
This study has demonstrated high levels of CR3 expression in
hematolymphoid and nonhematolymphoid organs during development that
coincide with high levels of hemopoiesis and decline after birth.
Prominent expression during the developmental period is compatible with
a role for CR3 in hemopoietic cell migration. CR3 expressed on isolated
neonatal peritoneal M is able to mediate adhesion and phagocytosis,
and it contributes to the initial phase of myeloid cell recruitment to
an inflammatory stimulus in vivo. However, mAb 5C6 was unable to
inhibit the largely monocytic, delayed phase of recruitment to LPS or
the constitutive migration of monocytes during development. M
migration in the normal developing animal is therefore largely
independent of CR3, which may perform additional, undiscovered
functions in tissue modelling and myelomonocytic cell ontogeny.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Siamon Gordon, Sir William Dunn School of Pathology, University of Oxford, South Parks Rd., Oxford, U.K. OX1 3RE.
3 Abbreviations used in this paper: M, macrophage; PMN, polymorphonuclear leukocyte; BP, bacteriologic plastic; PO, pathology Oxford; PLP, periodate-lysine-paraformaldehyde solution; RPM, resident peritoneal macrophages; TCP, tissue culture plastic; EAiC3b, opsonized sheep red cells; Eb, erythroblasts; TPM, thioglycolate-elicited peritoneal macrophages.
Received for publication April 9, 1997. Accepted for publication January 5, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
elimination. Immunology 68:227.[Medline]
-interferon on cachectin expression by mononuclear phagocytes: reversal of lps(d) (endotoxin resistance) phenotype. J. Exp. Med. 164:1791.This article has been cited by other articles:
|
|
 |
|
  J. Y. Bertrand, A. Jalil, M. Klaine, S. Jung, A. Cumano, and I. Godin Three pathways to mature macrophages in the early mouse yolk sac Blood, November 1, 2005; 106(9): 3004 - 3011. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Geutskens, T. Otonkoski, M-A. Pulkkinen, H. A. Drexhage, and P. J. M. Leenen Macrophages in the murine pancreas and their involvement in fetal endocrine development in vitro J. Leukoc. Biol., October 1, 2005; 78(4): 845 - 852. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. H. Horwitz, J. P. Mizgerd, M. L. Scott, and C. M. Doerschuk Mechanisms of granulocytosis in the absence of CD18 Blood, March 15, 2001; 97(6): 1578 - 1583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Lichanska, C. M. Browne, G. W. Henkel, K. M. Murphy, M. C. Ostrowski, S. R. McKercher, R. A. Maki, and D. A. Hume Differentiation of the Mononuclear Phagocyte System During Mouse Embryogenesis: The Role of Transcription Factor PU.1 Blood, July 1, 1999; 94(1): 127 - 138. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
