- Email: [email protected]
Milky spots in the mouse omentum may play an important role in the origin of peritoneal macrophages
~) INSTITUTPASTEUR/ELSEVIER Paris 1992
Res. Immunol. 1992, 143, 401-409
Milky spots in the mouse omentum may play an important role in the origin of peritoneal macrophages J.F.A.M. Wijffels, R.J.B.M. Hendrickx, J.J.E. Steenbergen, I.L. Eestermans and R.H.J. Beelen (*) Division o f Electron Microscopy, Department o f Cell Biology, Medical Faculty, Free University, van der Boechorststraat 7, 1081 B T Amsterdam (The Netherlands)
SUMMARY
In the milky spots, precursors of cells belonging to the mononuclear phagocyte system (MPS), such as monoblasts, can be found on the basis of ultrastructural endogenous peroxidase cytochemistry. Therefore, in the present study, we investigated the milky spots using a panel of monoclonal antibodies, especially antibodies (ER-MP) that recognize macrophage precursor antigens. Early macrophage precursor antigens ER-MP12 and ER-MP58 were detected only on cells localized inside the milky spots. On the other hand, an antigen which disappears late in the course of macrophage differentiation, ER-MP20, was detected in high amounts on cells both inside and around the milky spots. This clearly indicates that macrophage precursors are centrally localized inside the milky spots, while more differentiated cells are found in peripheral areas. Moreover, long-term culture of milky spot tissue resulted in the forming of a monolayer of stromal cells which supported macrophage proliferation in vitro. In conclusion, both in situ and in vitro studies demonstrated that mouse milky spots have a microenvironment in which precursor cells of the MPS can home and proliferate, illustrating that milky spots play a role as a source of local macrophage generation, e.g. that of the free peritoneal macrophages. Key-words: Immunogenesis, Milky spot, Macrophage; Mononuclear phagocyte system, Precursor cells, In situ and in vitro studies.
INTRODUCTION Milky spots of the mouse appear as opaque patches in the transparent omentum, first described by Ranvier (1874). Our earlier studies showed that the milky spots in the omemtum must be regarded as strongly reactive structures,
originating in part from perivascular effusions in the omentum (Beelen et al., 1980a, b). We also concluded that at least some of the free peritoneal macrophages were derived from these milky spots, which was later confirmed by others (Daems and de Bakker, 1982; Dux, 1986). According to the concept of the mononuclear
Submitted January 31, 1992. This manuscript was originallyscheduledto appear in the colloquium on "Biology and Pathology of the macrophage" (Res. Immunol., 143, 1). (*) For correspondence.
402
J.F.A.M. WIJFFELS E T AL.
phagocyte system (MPS; van Furth et al., 1972), all macrophages present in serous cavities (e.g. peritoneal macrophages) and tissues are derived from blood monocytes, originating from a committed precursor in the bone marrow. Although this concept is generally accepted (The Lancet, 1980) there is major controversy as to the contribution of local proliferation to the macrophage pool in situ, as demonstrated by Wisse (1980), Daems et al. (1982) and Tarling et al. (1987). We suggested earlier that precursor cells may proliferate locally in the milky spots and so provide a fraction of the free peritoneal macrophages (Beelen et al., 1980b). Proliferative activity of mouse macrophages in the omentum has been suggested (Dux, 1986) and it was shown that milky spot stromal cells produce CSF-1 (colony-stimulating factor) (Ratajczak et al., 1987). Recently, long-term cultured (LTC) cell lines of murine peritoneal resident macrophages were established due to the presence of feeder layers of stromal ceils, which produce CSF-1 (Lombard et al., 1985 ; 1988). CSF-1 stimulates the growth of bone marrow-derived macrophage colonies in vitro (Yen et al., 1987 ; Wijffels et al., 1992) and recently we found that, in the continuous presence of CSF-1, cloned mouse macrophage
cell lines can be generated in vitro (Wijffels et al., 1991a). These data stress that, possibly due to local CSF-1 production by stromal cells, local proliferation in the milky spots can contribute in situ to the generation of peritoneal macrophages. In the present study, we therefore investigated the presence of both immature precursors and well-differentiated macrophages (all belonging to the MPS) in the milky spots in situ. Moreover, we investigated the haemopoietic activity of milky spot tissue during LTC in vitro and thus studied the contribution of local macrophage proliferation in milky spots to the generation of free peritoneal macrophages.
MATERIALS AND METHODS
Animals
Male BALB/c mice (20-30 g) were obtained from CPB-Harlan, Zeist (The Netherlands). The greater o m e n t u m
An inflammatory state was induced by intraperitoneal (i.p.) administration of 1 ml newborn calf serum (Gibco, Europe BV, Breda, The Netherlands) after 24 h or 1 ml Brewer's thioglycollate broth (TGB) (Difco, Detroit, MI) after 4 days according
Fig. 1A. Ulthrathin section of a mouse milky spot showing a framework of reticulum cells (R), lining mesothelial cells (Me) and capillary endothelial cells (En) populated with lymphocytes (L) and macrophages (Ma). Fig. lB. Milky spot stained for endogenous PA, containing promonocytes (Pro) surrounded by monocytes. Fig. 1C. Note the detailed P A staining in RER, nuclear envelope (NE), Golgi system (GS) and lys0somal granules (G), characteristic of promonocytes.
CFU-M CSF ER-MP LTC mAb MNP
= colony-forming unit-macrophage. = colony-stimulating factor. early macrophage precursor. = long-term culture. = monoclonal antibody. = mononuclear phagocyte. =
MOMA MPS PA PBS RER TGB
= = = = = =
macrophage marker. mononuclear phagocyte system. peroxidate activity. phosphate-buffered saline. rough endoplasmic reticulum. thioglycollate broth.
MILKY SPOTS AND ORIGIN OF PERITONEAL MACROPHAGES
.
M~ ~.~
•
°*
.,
403
"
~,'~
I
"b
".
t
A •
"'-
Pro
r,
"" '~:.,
,~
:" t '
."
I
,
'~
j
•
•
%~.
0
.~
•
't.
"*
°
t °
Q
%,!
•
'o'
.%
L
,r
,
",.,:.
,',:/"'~
I
:
#, -
.~ . .,'-°%
:.o
°h,
..,:~.~-.
• .l(Slr'r--,-... ~ ~,~,
~ ' .
...,
~
'.~.
../°P
.. . ,." °m Q•
.
I I
~,,~.
•: =~-',c'~
"
,~ '.
.
.
•
'
:
1B
.
'~.~... t;o
1C
404
J.F.A.M.
WIJFFELS ET AL.
to Wijffels et al. (1991b). The greater omentum was dissected free from steady state as well as inflamed states according to Beelen et aL (1980a). Milky spot tissue was isolated and processed (a) for ultrastructural peroxidase cytochemistry as described earlier (Beelen et al., 1980b), (b) for immunochemistry as described below, or (c) for LTC. Milky spot tissue was directly put into culture or was completely dissociated with collagenase type III (1.0 mg/ml; 200 U/mg) (Worthington, Freehold, N J) as described elsewhere (Beelen, 1991).
lmmunochemistry Milky spots were frozen down in liver according to standard procedures, cryostat sections were made and fixed in acetone (Beelen, 1991). The slides were incubated for 1 h at room temperature with a panel of rat anti-mouse monoclonal antibodies (mAb) as outlined in table I. After washing the slides in phosphate-buffered saline (PBS), they were incubated with rabbit anti-rat IgG peroxidase (DAKO p162, pool Ig) in PBS with 1 07o normal mouse serum for 60 min at room temperature. After washing, the slides were stained for peroxidase using diaminobenzidine as trapping agent. Endogenous peroxidase was inhibited by 0.3 °7o H20 2 during acetone fixation. Cryostat sections were counterstained with haematoxylin. Control slides were incubated the same way, omitting the first step.
Culture of milky spots Milky spot tissue was cultured in RPMI-1640 (Flow), supplemented with 10 °70 foetal calf serum (Gibco), 100 U/ml penicillin, 100 ~.g/ml streptomycin, and 2 mM glutamine both without and with the addition of 20 o7o L929-cell-conditioned medium as a source of CSF-I (Wijffels et al., 1991a). Cultures were examined every other day and medium was
changed according to the method of Lombard et al. (1985). Alternatively, cell suspensions were made up of milky spots and processed further exactly as above.
RESULTS In the o m e n t u m of a normal steady-state mouse, numerous milky spots can be detected macroscopically as small opaque patches. After intraperitoneal injection of an inflammatory reagent, the blood flow throught the o m e n t u m increased strongly, followed by an increase in the number and size of the milky spots. The number of cells obtained after collagenase digestion o f milky spot tissue increased about 5 to 10-fold (normal steady-state 0.6 _+ 0.1 × 106 cells; n = 4; TG stimulated 4.7 + 2.5 × 106 cells; n = 4).
Ultrastructural cytochemistry Macrophages are the predominant cells in the mouse milky spots. The localization and kinetics o f different developmental stages - - mature resident macrophages outside and immature monocyte-derived macrophages inside - - suggested that at least a fraction of the free peritoneal macrophages are derived from the milky spots (fig. 1A). Most remarkably, precursor cells of the MPS were always found both after stimulation and in the normal steady state. These cells, found inside the milky spots, showed endogenous peroxidate activity (PA) in the nuclear envelope, the RER (rough endoplasmic
Table I. Panel of rat anti-mouse mAb used for characterization of MNP. mAb
Detected on :
Reference
MNP-precursor : CFU-M MNP-precursor: CFU-M MNP-lineage CFU-M till macrophages
Leenen et al. (1990) Leenen et al. (1990) Leenen et al. (1990)
all monocytes and macrophages mature macrophages
Wijffels et al. (1991) Austyn and Gordon (1981)
metallophilic macrophages stromal cells
Wijffels et al. (1991) van Vliet et al. (1986)
Immature:
ER-MP12 ER-MP58 ER-MP20 Mature:
MOMA-2 F4-80 Others:
MOMA-1 ER-TR7
M I L K Y SPOTS A N D ORIGIN OF P E R I T O N E A L M A C R O P H A G E S
,,,
,.
405
2A
o
-
2B
Fig. 2. Milky spot immunolabelled with MOMA-2 (A) and F4-80 (B). A) All macrophages are positive at the periphery of the milky spot and few central macrophages are positive. B) Mature macrophages are positive at the periphery of the milky spot and central macrophages are negative. Note the follicle-like area.
reticulum), the Golgi stacks and in the lysosomal granules (fig. 1B and 1C) and obviously represent promonocytes, which are normally found only in the bone marrow (van der Meer et al., 1979). This means that these promonocytes may divide locally and serve as a pool of monocytes and macrophages.
Immunochemistry Macrophages. With the panel of mAb (table I), which recognizes mouse macrophage subpopulations, we found that virtually all milky spot macrophages were positive for the general macrophage marker MOMA-2, i.e. all mature
406
J.F.A.M. WIJFFELS E T AL.
macrophages outside and the few immature macrophages inside (fig. 2A). F4-80, which recognizes only mature macrophages, was expressed only on the macrophages outside the milky spots (fig. 2B). These immunochemical data thus confirm the ultrastructural observations which indicate that mature macrophages (which enter the peritoneal cavity) may be derived from the immature milky spot macrophages. Most remarkably, as shown in figure 2, a follicle-like area is occasionally seen in the milky spots, containing mainly B lymphocytes. Also, a number of the milky spot macrophages were positive for MOMA-1 (not shown), which is normally seen only in lymphoid organs. From cell suspensions after collagenase digestion, about 50 °70 of the milky spot cells were macrophages, as detected by MOMA-2. Macrophage precursor cells. ER-MP 12 and ER-MP58 (both recognizing antigens on early macrophage precursors (CFU-M, see Leenen et al., 1990)), are found inside the milky spots in small numbers (fig. 3A). ER-MP20 (which recognizes a macrophage antigen which disappears late in the course of differentiation) is expressed at high levels on cells in the milky spot both inside and outside (fig. 3B). This clearly indicates that precursor ceils are localized in the milky spots of the mouse and proliferate locally.
Milky spot culture In culture of TGB-stimulated milky spots, fibroblasfs and mesothelial cells (stromal cells) migrated from the tissue to the culture substrate and adhered. After 1-2 weeks, a confluent feeder layer of these stromal cells was formed. Precursor cells which had migrated from the milky spots showed contact with the stromal cells and then began to multiply actively. Centres of macrophage proliferation were formed within two weeks (see fig. 5). Addition of CSF-1 during LTC did not enhance macrophage proliferation. The loosely adherent cells could be subcultured again several times. In subculture, a feeder-layer of stromal cells first developed, followed by macrophage proliferation. All macrophages which developed in this culture system were positive for the general macrophage marker MOMA-2, but not for F4-80, indicating their relatively immature state.
DISCUSSION
Stromai cells. ER-TR7, which recognizes
Our data clearly show that precursor cells of the MPS are present and proliferate locally in the milky spots, based on ultrastructural c y t o c h e m i s t r y (fig. 1) as well as immunophenotyping (fig. 2 and 3) in situ and also after LTC in vitro (fig. 5).
stromal cells, was clearly visible in the milky spots (fig. 4).
This means that, notwithstanding the general importance of the MPS (van Furth et al.,
Fig. 3. Milky spot immunolabelled with ER-MPl2 (A) and ER-MP20 (B). A) Only a few MNP (CFU-M) are positive inside the milky spot. B) Interior and exterior MNP (precursor cells to macrophages) are positive. Fig. 4. Milky spot stained with ER-TR7. Note the characteristic staining of the stromal cells (mesothelial cells, reticulum cells). Fig. 5. LTC of milky spot tissue resulting in centres of macrophage proliferation on a feeder layer of stromal cells in vitro.
~,~..~'~,
. ~i
~ . ~
~,,,,9,..
~ ~.
"
'
"
•
'
.... "
~
~ i'~--~'," ~ " ~
. ~
i
~
408
J.F.A.M. WIJFFELS E T AL.
1972; van der Meer et al., 1985), local proliferation does contribute to the generation o f the free peritoneal macrophages, which coincides very well with findings o f others (Tarling et al., 1987; L o m b a r d et al., 1988). Our data imply that milky spots are haemopoietically active during adult life and give rise to mature macrophages. Thus not only in adult bone marrow (Fibbe et al., 1985) but also in the milky spots (Ratjazak et al., 1987), stromal cells (fig. 4) supply the production of CSF-1. Both this CSF-1 production and cell/cell contact between stromal ceils and precursor ceils p r o b a b l y play key roles in providing this haemopoietic microenvironment in the milky spots, as suggested for long-term bone marrow cultures by Dexter (1989) and L T C from several sources by van der Heuvel et al. (1988). Recently, it was shown that early in ontogeny the o m e n t u m provide L y l + B cells (Solvason et al., 1992). In this context, we cannot exclude that already at a very early stage o f embryonic development, stem cells migrate from the yolk sac to the omentum. Thus, the milky spots which are formed may become haemopoietically active for providing Lyl ÷ B cells, and also for cells belonging to the M P S . This hypothesis is underlined by ontogenetic data, which demonstrate precursor cells o f the M P S earlier in the yolk sac (Beelen et al., 1990) and in the liver (Deimann and Fahimi, 1980) than in the bone marrow. The finding by Liebermann (1983), o f simultaneous development o f the milky spots and the spleen, also supports this hypothesis. Stem cells o f the M P S lineage may migrate from the bone marrow to the microenvironment in the milky s p o t ~ where they can also home for local macrophage proliferation and differentiation. In conclusion, our data indicate that local proliferation o f mononuclear phagocytes in the milky spots may significantly contribute to the generation of free peritoneal macrophages, regulated by the local microenvironment in which stromaI cells play a crucial role.
Les taches laiteuses de I'dpiploon de la souris peuvent jouer un r61e important thins la gen~se des macrophages p~riton~aux Dans les taches laiteuses, des prdcurseurs de cellules appartenant au syst~me des phagocytes mononucldaires (MPS = mononuclear phagocyte system), tels les monoblastes, peuvent ~tre trouvds par l'dtude ultrastructurelle de la cytochimie de la peroxydase endog~ne. En consdquence, dans cette dtude, notre investigation sur les taches laiteuses a dtd faite avec des anticorps monoclonaux reconnaissant spdcialement les antig6nes prdcoces des prdcurseurs des macrophages (ER-MP). Les antig~nes ER-MP12 et ERMP58 ont 6td ddtectds dans le centre seulement des cellules des taches laiteuses. Mais un antigone, ER-MP20, disparaissant tardivement au cours de la diffdrenciation des macrophages, a dtd ddtectd en fortes quantitds dans le centre et h la pdriphdrie. Cela indique clairement que les prdcurseurs sont localisds centralement et que les cellules plus diffdrencides sont trouvdes h la pdriphdrie. De plus, la culture de longue durde du tissu des taches laiteuses permet de produire une couche monocellulaire du stroma, support de la prolifdration des lymphocytes in vitro. En conclusion, les dtudes in situ et in vitro montrent que les taches laiteuses de la souris ont un microenvironnement dans lequel les cellules prdcurseurs du MPS peuvent s'installer et prolifdrer. Cela illustre le fait que les taches laiteuses jouent un r61e de source dans la gen~se locale des macrophages, par exemple, en partie pour les macrophages pdritondaux libres. Mots-clds: Macrophage, Tache laiteuse, Immunogen6se; Syst6me des phagocytes mononucldaires, Cellules prdcurseurs, Etudes in situ et in vitro.
References Austyn, J.M. & Gordon, S. (1981), F4/80, a monoclonal antibody directed specifically against the mouse macrophage. Europ. J. Immunol., 11, 805-815. Beelen, R.H.J., Fluitsma, D.M. & Foefsmit, E.C.M. (1980a), Cellular composition of milky spots and fine structure of milky spot macrophages and reticulum cells. J. Reticulendothel. Soc., 28, 585-600. Beelen, R.H.J., Fluitsma, D.M. & Hoefsmit, E.C.M. (1980b), Peroxidatic activity o f mononuclear phagocytes developing in milky spots. J. Reticulendothel. Soc., 28, 601-609. Beelen, R.H.J. (1991), The greater omentum: physiology and immunological concepts. Neth. J. Surg., 43, 5-11.
MILKY SPOTS AND ORIGIN OF PERITONEAL Beelen, R.H.J., van Rees, E.P., Bos, H.J., Kamperdijk, E.W.A. & Dijkstra, C.D. (1990), Ontogeny of antigen-presenting cells, in "Immunology of the fetus" (G. Chaouat) (pp. 33-41). CRC Press, West Palm Beach. Daems, W.T. & de Bakker, J.M. (1982), Do resident macrophages proliferate? Irnmunobiol., 161,204-212. Deimann, W. & Fahimi, H.D. (1980), Hepatic granulomas induced by glucan. An ultrastructural and peroxidasecytochemical study. Lab. Invest., 43, 172-181. Dexter, T.M. (1989), Biology of haematopoiesis. Oncology Rev., 4, 7-9. Dux, K. (1986), Proliferative activity of macrophages in the greater omentum of the mouse in relation to the early postnatal development of the vascular structures. J. Leuk. Biol., 40, 445-458. Fibbe, W.E., van Damme, J., Billiau, A., Goselink, H.M., Voogt, P.J., Ralph, P., Altrock, B.W. & Falkenburg, J.H.F. (1985), Interleukin 1 induces human marrow stromal cells in long-term cultures to produce granulocyte colony-stimulating factor and macrophage colony-stimulating factor. Blood, 71, 993-1004. van Furth, R., Cohn, Z.A., Hirsch, J.G., Humprey, J.L., Spector, J.H. & Langevoort, H.L. (1972), The mononuclear phagocyte system: a new classification of macrophages, monocytes and their precursor cells. Bull. Org. mond. Santd, 46, 845-850. van de Heuvel, R.L., Schroeters, G.E.R. & Vanderborght O.L.J. (1988), Haemopoiesis in long-term cultures of liver, spleen, and bone marrow of pre- and postnatal mice: CFU-GM production. Brit. J. Haematol., 70, 273-277. The Lancet, Editorial. (1980), 8161, 130-132. Leenen, P.J., Melis, M., Slieker, W.A.T. & van Ewijk, W. (1990), Murine macrophage precursor characterization. - - II. Monoclonal antibodies against macrophage precursor antigens. Europ. J. immunoL, 20, 27-36. Liebermann-Meffert, D. (1983), The greater omentum. Anatomy, pathology, surgery (White). SpringerVerlag, Heidelberg, Berlin. Lombard, Y., Ulrich, B. & Poindron, P. (1985), In vitro multiplication and apparently indefinite subcultures of normal mouse resident macrophages. Biol. Cell., 53, 219-230. Lombard, Y., Bartholeyns, J., Chokri, M, Illinger, D., Hartmann, D., Dumont, S., Kaufmann, S.H.E.,
MACROPHAGES
409
Landmann, R., Loor, F. & Poindron, P. (1988), Establishment and characterization of long-term cultured cell-lines of murine resident macrophages. J. Leuk. Biol., 44, 391-401. van der Meer, J.W.M., Beelen, R.H.J., Fluitsma, D.M. & van Furth, R. (1979), Ultrastructure of mononuclear phagocytes developing in liquid bone marrow cultures - - a study on peroxidatic activity. J. exp. Med., 149, 17-26. van der Meer, J.W.M., van de Gevel, J., de Water, R., Ginsel, L., Daems, W.T. & van Furth, R. (1985), Proliferation and differentiation of mononuclear phagocytes in vitro, in "Mononuclear phagocytes" (R. van Furth) (pp. 243-254). Martinus Nijhoff, The Hague. Ranvier, L. (1874), Du d~veloppement et de I'accroissement de vaisseaux sanguins. Arch. Physiol., 1, 429-450. Ratjaczak, M.Z., Jaskulski, D. & Pojda, Z. (1987), Omental lymphoid organ as a source of macrophage colonystimulating activity in peritoneal cavity. Clin. exp. Immunol., 69, 198-203. Solvoson, N., Lehuen, A. & Kearney, J.F. (1991), An embryonic source of Lyl but not conventional B cells. Int. lmmunol. (in press). Tarling, J., Lin, H. & Hsu, F. (1987), Self renewal of pulmonary macrophages: evidence from radiation chimera studies. J. Leuk. Biol., 42, 443-450. van Vliet, E.P., Melis, M., Foidart, J.M. & van Ewijk, W. (1986), Reticular fibroblasts in peripheral lymphoid organs identified by a monoclonal antibody. J. Histochem. Cytochem., 34, 883-890. Wijffels, J.F.A.M., de Rover, Z., van Rooijen, N., Kraal, G. & Beelen, R.H.J. (1991a), Clonal expansion of mouse macrophage cell lines, which can be repopulated in vitro, in "Lymphatic tissues and in vivo immune responses" (pp. 845-850). Marcel Dekker, New York. Wijffels, J.F.A.M., de Rover, Z., van Rooyen, N., Kraal, G. & Beelen, R.H.J. (1991b), Chronic inflammation induces the expression of dendritic cell markers not related to functional antigen presentation on exudate macrophages. Immunobiol., 184, 83-92. Wisse, E. (1980), On the kupffer cell origin of Kupffer cells. Blood Cells, 6, 91-92. Yen, S.E., Beelen, R.H.J. & Walker, W.S. (1987), Expression of Ia antigen of colonies of bone marrow derived macrophages. J. Leuk. Biol., 42, 697-703.
Res. Immunol. 1992, 143, 401-409
Milky spots in the mouse omentum may play an important role in the origin of peritoneal macrophages J.F.A.M. Wijffels, R.J.B.M. Hendrickx, J.J.E. Steenbergen, I.L. Eestermans and R.H.J. Beelen (*) Division o f Electron Microscopy, Department o f Cell Biology, Medical Faculty, Free University, van der Boechorststraat 7, 1081 B T Amsterdam (The Netherlands)
SUMMARY
In the milky spots, precursors of cells belonging to the mononuclear phagocyte system (MPS), such as monoblasts, can be found on the basis of ultrastructural endogenous peroxidase cytochemistry. Therefore, in the present study, we investigated the milky spots using a panel of monoclonal antibodies, especially antibodies (ER-MP) that recognize macrophage precursor antigens. Early macrophage precursor antigens ER-MP12 and ER-MP58 were detected only on cells localized inside the milky spots. On the other hand, an antigen which disappears late in the course of macrophage differentiation, ER-MP20, was detected in high amounts on cells both inside and around the milky spots. This clearly indicates that macrophage precursors are centrally localized inside the milky spots, while more differentiated cells are found in peripheral areas. Moreover, long-term culture of milky spot tissue resulted in the forming of a monolayer of stromal cells which supported macrophage proliferation in vitro. In conclusion, both in situ and in vitro studies demonstrated that mouse milky spots have a microenvironment in which precursor cells of the MPS can home and proliferate, illustrating that milky spots play a role as a source of local macrophage generation, e.g. that of the free peritoneal macrophages. Key-words: Immunogenesis, Milky spot, Macrophage; Mononuclear phagocyte system, Precursor cells, In situ and in vitro studies.
INTRODUCTION Milky spots of the mouse appear as opaque patches in the transparent omentum, first described by Ranvier (1874). Our earlier studies showed that the milky spots in the omemtum must be regarded as strongly reactive structures,
originating in part from perivascular effusions in the omentum (Beelen et al., 1980a, b). We also concluded that at least some of the free peritoneal macrophages were derived from these milky spots, which was later confirmed by others (Daems and de Bakker, 1982; Dux, 1986). According to the concept of the mononuclear
Submitted January 31, 1992. This manuscript was originallyscheduledto appear in the colloquium on "Biology and Pathology of the macrophage" (Res. Immunol., 143, 1). (*) For correspondence.
402
J.F.A.M. WIJFFELS E T AL.
phagocyte system (MPS; van Furth et al., 1972), all macrophages present in serous cavities (e.g. peritoneal macrophages) and tissues are derived from blood monocytes, originating from a committed precursor in the bone marrow. Although this concept is generally accepted (The Lancet, 1980) there is major controversy as to the contribution of local proliferation to the macrophage pool in situ, as demonstrated by Wisse (1980), Daems et al. (1982) and Tarling et al. (1987). We suggested earlier that precursor cells may proliferate locally in the milky spots and so provide a fraction of the free peritoneal macrophages (Beelen et al., 1980b). Proliferative activity of mouse macrophages in the omentum has been suggested (Dux, 1986) and it was shown that milky spot stromal cells produce CSF-1 (colony-stimulating factor) (Ratajczak et al., 1987). Recently, long-term cultured (LTC) cell lines of murine peritoneal resident macrophages were established due to the presence of feeder layers of stromal ceils, which produce CSF-1 (Lombard et al., 1985 ; 1988). CSF-1 stimulates the growth of bone marrow-derived macrophage colonies in vitro (Yen et al., 1987 ; Wijffels et al., 1992) and recently we found that, in the continuous presence of CSF-1, cloned mouse macrophage
cell lines can be generated in vitro (Wijffels et al., 1991a). These data stress that, possibly due to local CSF-1 production by stromal cells, local proliferation in the milky spots can contribute in situ to the generation of peritoneal macrophages. In the present study, we therefore investigated the presence of both immature precursors and well-differentiated macrophages (all belonging to the MPS) in the milky spots in situ. Moreover, we investigated the haemopoietic activity of milky spot tissue during LTC in vitro and thus studied the contribution of local macrophage proliferation in milky spots to the generation of free peritoneal macrophages.
MATERIALS AND METHODS
Animals
Male BALB/c mice (20-30 g) were obtained from CPB-Harlan, Zeist (The Netherlands). The greater o m e n t u m
An inflammatory state was induced by intraperitoneal (i.p.) administration of 1 ml newborn calf serum (Gibco, Europe BV, Breda, The Netherlands) after 24 h or 1 ml Brewer's thioglycollate broth (TGB) (Difco, Detroit, MI) after 4 days according
Fig. 1A. Ulthrathin section of a mouse milky spot showing a framework of reticulum cells (R), lining mesothelial cells (Me) and capillary endothelial cells (En) populated with lymphocytes (L) and macrophages (Ma). Fig. lB. Milky spot stained for endogenous PA, containing promonocytes (Pro) surrounded by monocytes. Fig. 1C. Note the detailed P A staining in RER, nuclear envelope (NE), Golgi system (GS) and lys0somal granules (G), characteristic of promonocytes.
CFU-M CSF ER-MP LTC mAb MNP
= colony-forming unit-macrophage. = colony-stimulating factor. early macrophage precursor. = long-term culture. = monoclonal antibody. = mononuclear phagocyte. =
MOMA MPS PA PBS RER TGB
= = = = = =
macrophage marker. mononuclear phagocyte system. peroxidate activity. phosphate-buffered saline. rough endoplasmic reticulum. thioglycollate broth.
MILKY SPOTS AND ORIGIN OF PERITONEAL MACROPHAGES
.
M~ ~.~
•
°*
.,
403
"
~,'~
I
"b
".
t
A •
"'-
Pro
r,
"" '~:.,
,~
:" t '
."
I
,
'~
j
•
•
%~.
0
.~
•
't.
"*
°
t °
Q
%,!
•
'o'
.%
L
,r
,
",.,:.
,',:/"'~
I
:
#, -
.~ . .,'-°%
:.o
°h,
..,:~.~-.
• .l(Slr'r--,-... ~ ~,~,
~ ' .
...,
~
'.~.
../°P
.. . ,." °m Q•
.
I I
~,,~.
•: =~-',c'~
"
,~ '.
.
.
•
'
:
1B
.
'~.~... t;o
1C
404
J.F.A.M.
WIJFFELS ET AL.
to Wijffels et al. (1991b). The greater omentum was dissected free from steady state as well as inflamed states according to Beelen et aL (1980a). Milky spot tissue was isolated and processed (a) for ultrastructural peroxidase cytochemistry as described earlier (Beelen et al., 1980b), (b) for immunochemistry as described below, or (c) for LTC. Milky spot tissue was directly put into culture or was completely dissociated with collagenase type III (1.0 mg/ml; 200 U/mg) (Worthington, Freehold, N J) as described elsewhere (Beelen, 1991).
lmmunochemistry Milky spots were frozen down in liver according to standard procedures, cryostat sections were made and fixed in acetone (Beelen, 1991). The slides were incubated for 1 h at room temperature with a panel of rat anti-mouse monoclonal antibodies (mAb) as outlined in table I. After washing the slides in phosphate-buffered saline (PBS), they were incubated with rabbit anti-rat IgG peroxidase (DAKO p162, pool Ig) in PBS with 1 07o normal mouse serum for 60 min at room temperature. After washing, the slides were stained for peroxidase using diaminobenzidine as trapping agent. Endogenous peroxidase was inhibited by 0.3 °7o H20 2 during acetone fixation. Cryostat sections were counterstained with haematoxylin. Control slides were incubated the same way, omitting the first step.
Culture of milky spots Milky spot tissue was cultured in RPMI-1640 (Flow), supplemented with 10 °70 foetal calf serum (Gibco), 100 U/ml penicillin, 100 ~.g/ml streptomycin, and 2 mM glutamine both without and with the addition of 20 o7o L929-cell-conditioned medium as a source of CSF-I (Wijffels et al., 1991a). Cultures were examined every other day and medium was
changed according to the method of Lombard et al. (1985). Alternatively, cell suspensions were made up of milky spots and processed further exactly as above.
RESULTS In the o m e n t u m of a normal steady-state mouse, numerous milky spots can be detected macroscopically as small opaque patches. After intraperitoneal injection of an inflammatory reagent, the blood flow throught the o m e n t u m increased strongly, followed by an increase in the number and size of the milky spots. The number of cells obtained after collagenase digestion o f milky spot tissue increased about 5 to 10-fold (normal steady-state 0.6 _+ 0.1 × 106 cells; n = 4; TG stimulated 4.7 + 2.5 × 106 cells; n = 4).
Ultrastructural cytochemistry Macrophages are the predominant cells in the mouse milky spots. The localization and kinetics o f different developmental stages - - mature resident macrophages outside and immature monocyte-derived macrophages inside - - suggested that at least a fraction of the free peritoneal macrophages are derived from the milky spots (fig. 1A). Most remarkably, precursor cells of the MPS were always found both after stimulation and in the normal steady state. These cells, found inside the milky spots, showed endogenous peroxidate activity (PA) in the nuclear envelope, the RER (rough endoplasmic
Table I. Panel of rat anti-mouse mAb used for characterization of MNP. mAb
Detected on :
Reference
MNP-precursor : CFU-M MNP-precursor: CFU-M MNP-lineage CFU-M till macrophages
Leenen et al. (1990) Leenen et al. (1990) Leenen et al. (1990)
all monocytes and macrophages mature macrophages
Wijffels et al. (1991) Austyn and Gordon (1981)
metallophilic macrophages stromal cells
Wijffels et al. (1991) van Vliet et al. (1986)
Immature:
ER-MP12 ER-MP58 ER-MP20 Mature:
MOMA-2 F4-80 Others:
MOMA-1 ER-TR7
M I L K Y SPOTS A N D ORIGIN OF P E R I T O N E A L M A C R O P H A G E S
,,,
,.
405
2A
o
-
2B
Fig. 2. Milky spot immunolabelled with MOMA-2 (A) and F4-80 (B). A) All macrophages are positive at the periphery of the milky spot and few central macrophages are positive. B) Mature macrophages are positive at the periphery of the milky spot and central macrophages are negative. Note the follicle-like area.
reticulum), the Golgi stacks and in the lysosomal granules (fig. 1B and 1C) and obviously represent promonocytes, which are normally found only in the bone marrow (van der Meer et al., 1979). This means that these promonocytes may divide locally and serve as a pool of monocytes and macrophages.
Immunochemistry Macrophages. With the panel of mAb (table I), which recognizes mouse macrophage subpopulations, we found that virtually all milky spot macrophages were positive for the general macrophage marker MOMA-2, i.e. all mature
406
J.F.A.M. WIJFFELS E T AL.
macrophages outside and the few immature macrophages inside (fig. 2A). F4-80, which recognizes only mature macrophages, was expressed only on the macrophages outside the milky spots (fig. 2B). These immunochemical data thus confirm the ultrastructural observations which indicate that mature macrophages (which enter the peritoneal cavity) may be derived from the immature milky spot macrophages. Most remarkably, as shown in figure 2, a follicle-like area is occasionally seen in the milky spots, containing mainly B lymphocytes. Also, a number of the milky spot macrophages were positive for MOMA-1 (not shown), which is normally seen only in lymphoid organs. From cell suspensions after collagenase digestion, about 50 °70 of the milky spot cells were macrophages, as detected by MOMA-2. Macrophage precursor cells. ER-MP 12 and ER-MP58 (both recognizing antigens on early macrophage precursors (CFU-M, see Leenen et al., 1990)), are found inside the milky spots in small numbers (fig. 3A). ER-MP20 (which recognizes a macrophage antigen which disappears late in the course of differentiation) is expressed at high levels on cells in the milky spot both inside and outside (fig. 3B). This clearly indicates that precursor ceils are localized in the milky spots of the mouse and proliferate locally.
Milky spot culture In culture of TGB-stimulated milky spots, fibroblasfs and mesothelial cells (stromal cells) migrated from the tissue to the culture substrate and adhered. After 1-2 weeks, a confluent feeder layer of these stromal cells was formed. Precursor cells which had migrated from the milky spots showed contact with the stromal cells and then began to multiply actively. Centres of macrophage proliferation were formed within two weeks (see fig. 5). Addition of CSF-1 during LTC did not enhance macrophage proliferation. The loosely adherent cells could be subcultured again several times. In subculture, a feeder-layer of stromal cells first developed, followed by macrophage proliferation. All macrophages which developed in this culture system were positive for the general macrophage marker MOMA-2, but not for F4-80, indicating their relatively immature state.
DISCUSSION
Stromai cells. ER-TR7, which recognizes
Our data clearly show that precursor cells of the MPS are present and proliferate locally in the milky spots, based on ultrastructural c y t o c h e m i s t r y (fig. 1) as well as immunophenotyping (fig. 2 and 3) in situ and also after LTC in vitro (fig. 5).
stromal cells, was clearly visible in the milky spots (fig. 4).
This means that, notwithstanding the general importance of the MPS (van Furth et al.,
Fig. 3. Milky spot immunolabelled with ER-MPl2 (A) and ER-MP20 (B). A) Only a few MNP (CFU-M) are positive inside the milky spot. B) Interior and exterior MNP (precursor cells to macrophages) are positive. Fig. 4. Milky spot stained with ER-TR7. Note the characteristic staining of the stromal cells (mesothelial cells, reticulum cells). Fig. 5. LTC of milky spot tissue resulting in centres of macrophage proliferation on a feeder layer of stromal cells in vitro.
~,~..~'~,
. ~i
~ . ~
~,,,,9,..
~ ~.
"
'
"
•
'
.... "
~
~ i'~--~'," ~ " ~
. ~
i
~
408
J.F.A.M. WIJFFELS E T AL.
1972; van der Meer et al., 1985), local proliferation does contribute to the generation o f the free peritoneal macrophages, which coincides very well with findings o f others (Tarling et al., 1987; L o m b a r d et al., 1988). Our data imply that milky spots are haemopoietically active during adult life and give rise to mature macrophages. Thus not only in adult bone marrow (Fibbe et al., 1985) but also in the milky spots (Ratjazak et al., 1987), stromal cells (fig. 4) supply the production of CSF-1. Both this CSF-1 production and cell/cell contact between stromal ceils and precursor ceils p r o b a b l y play key roles in providing this haemopoietic microenvironment in the milky spots, as suggested for long-term bone marrow cultures by Dexter (1989) and L T C from several sources by van der Heuvel et al. (1988). Recently, it was shown that early in ontogeny the o m e n t u m provide L y l + B cells (Solvason et al., 1992). In this context, we cannot exclude that already at a very early stage o f embryonic development, stem cells migrate from the yolk sac to the omentum. Thus, the milky spots which are formed may become haemopoietically active for providing Lyl ÷ B cells, and also for cells belonging to the M P S . This hypothesis is underlined by ontogenetic data, which demonstrate precursor cells o f the M P S earlier in the yolk sac (Beelen et al., 1990) and in the liver (Deimann and Fahimi, 1980) than in the bone marrow. The finding by Liebermann (1983), o f simultaneous development o f the milky spots and the spleen, also supports this hypothesis. Stem cells o f the M P S lineage may migrate from the bone marrow to the microenvironment in the milky s p o t ~ where they can also home for local macrophage proliferation and differentiation. In conclusion, our data indicate that local proliferation o f mononuclear phagocytes in the milky spots may significantly contribute to the generation of free peritoneal macrophages, regulated by the local microenvironment in which stromaI cells play a crucial role.
Les taches laiteuses de I'dpiploon de la souris peuvent jouer un r61e important thins la gen~se des macrophages p~riton~aux Dans les taches laiteuses, des prdcurseurs de cellules appartenant au syst~me des phagocytes mononucldaires (MPS = mononuclear phagocyte system), tels les monoblastes, peuvent ~tre trouvds par l'dtude ultrastructurelle de la cytochimie de la peroxydase endog~ne. En consdquence, dans cette dtude, notre investigation sur les taches laiteuses a dtd faite avec des anticorps monoclonaux reconnaissant spdcialement les antig6nes prdcoces des prdcurseurs des macrophages (ER-MP). Les antig~nes ER-MP12 et ERMP58 ont 6td ddtectds dans le centre seulement des cellules des taches laiteuses. Mais un antigone, ER-MP20, disparaissant tardivement au cours de la diffdrenciation des macrophages, a dtd ddtectd en fortes quantitds dans le centre et h la pdriphdrie. Cela indique clairement que les prdcurseurs sont localisds centralement et que les cellules plus diffdrencides sont trouvdes h la pdriphdrie. De plus, la culture de longue durde du tissu des taches laiteuses permet de produire une couche monocellulaire du stroma, support de la prolifdration des lymphocytes in vitro. En conclusion, les dtudes in situ et in vitro montrent que les taches laiteuses de la souris ont un microenvironnement dans lequel les cellules prdcurseurs du MPS peuvent s'installer et prolifdrer. Cela illustre le fait que les taches laiteuses jouent un r61e de source dans la gen~se locale des macrophages, par exemple, en partie pour les macrophages pdritondaux libres. Mots-clds: Macrophage, Tache laiteuse, Immunogen6se; Syst6me des phagocytes mononucldaires, Cellules prdcurseurs, Etudes in situ et in vitro.
References Austyn, J.M. & Gordon, S. (1981), F4/80, a monoclonal antibody directed specifically against the mouse macrophage. Europ. J. Immunol., 11, 805-815. Beelen, R.H.J., Fluitsma, D.M. & Foefsmit, E.C.M. (1980a), Cellular composition of milky spots and fine structure of milky spot macrophages and reticulum cells. J. Reticulendothel. Soc., 28, 585-600. Beelen, R.H.J., Fluitsma, D.M. & Hoefsmit, E.C.M. (1980b), Peroxidatic activity o f mononuclear phagocytes developing in milky spots. J. Reticulendothel. Soc., 28, 601-609. Beelen, R.H.J. (1991), The greater omentum: physiology and immunological concepts. Neth. J. Surg., 43, 5-11.
MILKY SPOTS AND ORIGIN OF PERITONEAL Beelen, R.H.J., van Rees, E.P., Bos, H.J., Kamperdijk, E.W.A. & Dijkstra, C.D. (1990), Ontogeny of antigen-presenting cells, in "Immunology of the fetus" (G. Chaouat) (pp. 33-41). CRC Press, West Palm Beach. Daems, W.T. & de Bakker, J.M. (1982), Do resident macrophages proliferate? Irnmunobiol., 161,204-212. Deimann, W. & Fahimi, H.D. (1980), Hepatic granulomas induced by glucan. An ultrastructural and peroxidasecytochemical study. Lab. Invest., 43, 172-181. Dexter, T.M. (1989), Biology of haematopoiesis. Oncology Rev., 4, 7-9. Dux, K. (1986), Proliferative activity of macrophages in the greater omentum of the mouse in relation to the early postnatal development of the vascular structures. J. Leuk. Biol., 40, 445-458. Fibbe, W.E., van Damme, J., Billiau, A., Goselink, H.M., Voogt, P.J., Ralph, P., Altrock, B.W. & Falkenburg, J.H.F. (1985), Interleukin 1 induces human marrow stromal cells in long-term cultures to produce granulocyte colony-stimulating factor and macrophage colony-stimulating factor. Blood, 71, 993-1004. van Furth, R., Cohn, Z.A., Hirsch, J.G., Humprey, J.L., Spector, J.H. & Langevoort, H.L. (1972), The mononuclear phagocyte system: a new classification of macrophages, monocytes and their precursor cells. Bull. Org. mond. Santd, 46, 845-850. van de Heuvel, R.L., Schroeters, G.E.R. & Vanderborght O.L.J. (1988), Haemopoiesis in long-term cultures of liver, spleen, and bone marrow of pre- and postnatal mice: CFU-GM production. Brit. J. Haematol., 70, 273-277. The Lancet, Editorial. (1980), 8161, 130-132. Leenen, P.J., Melis, M., Slieker, W.A.T. & van Ewijk, W. (1990), Murine macrophage precursor characterization. - - II. Monoclonal antibodies against macrophage precursor antigens. Europ. J. immunoL, 20, 27-36. Liebermann-Meffert, D. (1983), The greater omentum. Anatomy, pathology, surgery (White). SpringerVerlag, Heidelberg, Berlin. Lombard, Y., Ulrich, B. & Poindron, P. (1985), In vitro multiplication and apparently indefinite subcultures of normal mouse resident macrophages. Biol. Cell., 53, 219-230. Lombard, Y., Bartholeyns, J., Chokri, M, Illinger, D., Hartmann, D., Dumont, S., Kaufmann, S.H.E.,
MACROPHAGES
409
Landmann, R., Loor, F. & Poindron, P. (1988), Establishment and characterization of long-term cultured cell-lines of murine resident macrophages. J. Leuk. Biol., 44, 391-401. van der Meer, J.W.M., Beelen, R.H.J., Fluitsma, D.M. & van Furth, R. (1979), Ultrastructure of mononuclear phagocytes developing in liquid bone marrow cultures - - a study on peroxidatic activity. J. exp. Med., 149, 17-26. van der Meer, J.W.M., van de Gevel, J., de Water, R., Ginsel, L., Daems, W.T. & van Furth, R. (1985), Proliferation and differentiation of mononuclear phagocytes in vitro, in "Mononuclear phagocytes" (R. van Furth) (pp. 243-254). Martinus Nijhoff, The Hague. Ranvier, L. (1874), Du d~veloppement et de I'accroissement de vaisseaux sanguins. Arch. Physiol., 1, 429-450. Ratjaczak, M.Z., Jaskulski, D. & Pojda, Z. (1987), Omental lymphoid organ as a source of macrophage colonystimulating activity in peritoneal cavity. Clin. exp. Immunol., 69, 198-203. Solvoson, N., Lehuen, A. & Kearney, J.F. (1991), An embryonic source of Lyl but not conventional B cells. Int. lmmunol. (in press). Tarling, J., Lin, H. & Hsu, F. (1987), Self renewal of pulmonary macrophages: evidence from radiation chimera studies. J. Leuk. Biol., 42, 443-450. van Vliet, E.P., Melis, M., Foidart, J.M. & van Ewijk, W. (1986), Reticular fibroblasts in peripheral lymphoid organs identified by a monoclonal antibody. J. Histochem. Cytochem., 34, 883-890. Wijffels, J.F.A.M., de Rover, Z., van Rooijen, N., Kraal, G. & Beelen, R.H.J. (1991a), Clonal expansion of mouse macrophage cell lines, which can be repopulated in vitro, in "Lymphatic tissues and in vivo immune responses" (pp. 845-850). Marcel Dekker, New York. Wijffels, J.F.A.M., de Rover, Z., van Rooyen, N., Kraal, G. & Beelen, R.H.J. (1991b), Chronic inflammation induces the expression of dendritic cell markers not related to functional antigen presentation on exudate macrophages. Immunobiol., 184, 83-92. Wisse, E. (1980), On the kupffer cell origin of Kupffer cells. Blood Cells, 6, 91-92. Yen, S.E., Beelen, R.H.J. & Walker, W.S. (1987), Expression of Ia antigen of colonies of bone marrow derived macrophages. J. Leuk. Biol., 42, 697-703.