It is well established that macrophages have a function in scavenging apoptotic bodies from cells undergoing programmed cell death. Here we show that macrophages can also induce apoptosis of normal cells. Using injected toxic liposomes to eliminate macrophages in the anterior chamber of the rat eye, we provide direct evidence that, in vivo, macrophages induce apoptosis in normal vascular endothelial cells during programmed capillary regression. Macrophage elimination resulted in the survival of endothelial cells that normally would die and the persistence of functional capillaries. Furthermore, replacement of eliminated macrophages with bone-marrow-derived macrophages ‘rescued’ lack of capillary regression. Viability of the persistent target cells was demonstrated through their lack of apoptotic morphology, expression of intracellular esterases and synthesis of DNA. These results uncover a new function for macrophages in the remodeling of tissues through the induction of programmed cell death and provide direct evidence of a key role for macrophages in capillary regression.

Despite decades of interest in the function of macrophages, the in vivo significance of their ability to kill non-immune target cells in culture has remained obscure. Two disparate develop- mental systems have suggested that phagocytes might have activity in non-immune induction of target cell death. Laser ablation studies in C. elegans show that death of the gonadal linker cell (a component of the male sexual apparatus) requires the action of adjacent phagocytic epidermal cells (Sulston and Thompson, 1980), while cellular ablation transgenic mice (Lang and Bishop, 1993) have led to the proposal that, in mammals too, programmed cell death might sometimes be dependent on phagocytes.

In mammalian systems, mediators of macrophage cytotoxicity in vitro include TNFα (van de Loosdrecht et al., 1993) and nitric oxide (Cui et al., 1994). It has been presumed that these cytotoxic activities might be consistent with a role for macrophages in tissue maintenance (Vlassara et al., 1988) or in the case of a tumor cell target, with anti-neoplastic surveillance (Urban et al., 1986). However, in all cases characterized so far (Bernasconi et al., 1991), macrophage killing has required stimulation with inflammatory mediators and the relevance of this activity to normal physiology is unclear. In most cases, macrophage cytotoxicity results in the death of target cells through apoptosis (Aliprantis et al., 1996).

The pupillary membrane is a temporary capillary network in the anterior chamber of the eye (Matsuo and Smelser, 1971). It adheres to the anterior surface of the lens and iris diaphragm and regresses between weeks 1 and 3 after birth in the rat (Lang et al., 1994). Previous studies on the pattern of cell death in the pupillary membrane have suggested that capillary regression occurs in two stages. The first stage was proposed to be a macrophage-dependent induction of endothelial cell apoptosis (Lang and Bishop, 1993; Lang et al., 1994) and the second a coordinated apoptosis of capillary endothelial cells dependent on cessation of plasma flow (Meeson et al., 1996). The accessibility and relative simplicity of the pupillary membrane recommend it as a system in which to study vascular regression and, in particular, macrophage-endothelial cell interactions in an in vivo setting.

In the current study, we have used the liposome-mediated macrophage elimination technique (Van Rooijen, 1989) combined with macrophage-reconstitution to show that programmed capillary regression is dependent upon macrophage- induced apoptosis. Elimination of macrophages in the anterior chamber results in prevention of endothelial cell apoptosis and a dramatic persistence of the pupillary membrane in functional form. Macrophage reconstitution following macrophage elimination results in a ‘rescue’ of the persistent membrane. These experiments demonstrate that macrophages can induce apoptosis in normal cells in vivo and that they play a key role in programmed capillary regression.

Animal breeding

Timed pregnant Sprague-Dawley and inbred Fisher rats were obtained from Taconic Farms (German Town, NY). Ages of rats are noted as days after birth (e.g. A14). Animal care was in accordance with institutional guidelines.

Liposomes

Control and toxic liposomes were generated according to established procedures (Van Rooijen, 1989). The liposome-dependent drug dichloromethylene bisphosphonate (clodronate) was kindly provided by Boehringer Mannheim GMBH. For macrophage elimination in the anterior chamber, liposome suspensions were injected trans-corneally in volumes ranging from 50 to 650 nl.

Trans-corneal injection

Trans-corneal injection was performed using a modifi- cation of previously described vital observation technique (Fig. 1; Meeson et al., 1996). Beveled glass injection needles were used to deliver defined volumes of liposomes to the anterior chamber using a Nanoject Variable injector (Drummond, PA). Injection was observed with the 5× objective of a Zeiss Axiophot microscope (Zeiss, Thornwood, NY). A custom- designed stage plate allowed mounting of micromanipulators for injector and the optic fibre channel that provided illumination. Anesthesia of rat pups was performed as previously described except that Ketamine and Xylazine were delivered intraperitoneally and the local anesthetic tetracaine hydrochloride (0.5%) applied prior to parting of the eyelids. Injection was performed into the right anterior chamber with the rat pups positioned in a silastic body mold.

Fig. 1.

The diagram shows the arrangement of elements for transcorneal injection. A 5× air objective is positioned above the cornea of an anaesthetized rat. An optic fibre light guide that provides illumination also functions to limit eye movement. A beveled glass needle mounted on a micromanipulator is axially apposed to the eye holder and in this position can penetrate the cornea for delivery of material directly onto the pupillary membrane (pm).

Fig. 1.

The diagram shows the arrangement of elements for transcorneal injection. A 5× air objective is positioned above the cornea of an anaesthetized rat. An optic fibre light guide that provides illumination also functions to limit eye movement. A beveled glass needle mounted on a micromanipulator is axially apposed to the eye holder and in this position can penetrate the cornea for delivery of material directly onto the pupillary membrane (pm).

Culture of bone-marrow-derived macrophages

Bone-marrow-derived macrophages were prepared according to the following protocol (I. Riviere, personal communication). Bone marrow cells were harvested from the femurs of 1- to 2-week-old Fisher rats. 107 cells were plated in 10 ml of RPMI medium containing 10% fetal calf serum, 5×10−5 M β-mercaptoethanol and incubated for 2-4 hours at 37°C. The non-adherent cells were removed, replated on Teflon culture dishes and cultured overnight. Non-adherent cells were discarded and the remaining adherent cells grown in RPMI, 10% fetal calf serum, 5×10−5 M β-mercaptoethanol, 20% L929 supernatant, 10% horse serum (Gibco BRL). Immature macrophages were grown for 5-6 days and during this period fed with fresh medium every other day. Prior to trans-corneal injection, macrophages were removed from the culture dish with a cell dissociation solution, washed with serum-free medium and resuspended at a concentration of 2×106/ml. 1 μl of this cell suspension was injected for macrophage reconstitution experiments.

Histological procedures

The pupillary membrane/iris diaphragm complex was dissected from the rat eye as previously described (Lang et al., 1994) and nuclei stained with Hoechst 33258 (Sigma, St Louis, MO) or with hema-toxylin. Apoptosis was recognized as the characteristic condensation of apoptotic bodies (Wyllie et al., 1980). Macrophages were identified using indirect immunofluorescence with the primary antibody OX42 (Robinson et al., 1986)(Cedarlane Laboratories Ltd., Hornby, Ontario) and rhodamine-conjugated goat anti-mouse IgG as a secondary antibody. An assessment of cellular proliferation through S-phase labeling of cells with 5-bromo-2′-deoxyuridine was performed according to standard procedures (Takahashi et al., 1993). Calcein AM (Papadopoulos et al., 1994; Molecular Probes, Eugene, OR) was used to stain dissected pupillary membranes at 50 μM in PBS for 30 minutes at 37°C prior to fluorescence visualization.

Timing of trans-corneal injections

The following diagram (Fig. 2) summarizes the times, relative to normal pupillary membrane regression, when trans-corneal injection of liposomes and macrophages was performed for the different experiments described.

Fig. 2.

The timing of normal pupillary membrane regression in rats relative to experimental manipulations. Morphological signs of regression occur between A7 and A14 as indicated in the three representations of the eye at the top of the figure. Experiments shown in Figs 3 and 4 were performed with a liposome injection at A7 and analysis at A11. Experiments shown in Figs 5 and 7 were performed with two injections of liposomes, the first at A3 and the second at A7. Analysis after dissection was performed at A14 at a time when a normal pupillary membrane has all but completely regressed. Macrophage reconstitution experiments (Fig. 6) were performed with a liposome injection at A7, macrophage injection at A12 and dissection for analysis at A14.

Fig. 2.

The timing of normal pupillary membrane regression in rats relative to experimental manipulations. Morphological signs of regression occur between A7 and A14 as indicated in the three representations of the eye at the top of the figure. Experiments shown in Figs 3 and 4 were performed with a liposome injection at A7 and analysis at A11. Experiments shown in Figs 5 and 7 were performed with two injections of liposomes, the first at A3 and the second at A7. Analysis after dissection was performed at A14 at a time when a normal pupillary membrane has all but completely regressed. Macrophage reconstitution experiments (Fig. 6) were performed with a liposome injection at A7, macrophage injection at A12 and dissection for analysis at A14.

Toxic liposomes containing PBS-clodronate and control liposomes containing only PBS were injected into the anterior chamber of young rat eyes at A7 and the pupillary membranes assessed histologically after removal by dissection at A11 (Lang et al., 1994). Control liposomes encapsulating PBS and tagged with the lipophilic fluorescent dye DiO effectively labeled macrophages associated with the pupillary membrane (Fig. 3A,C). In contrast, liposomes encapsulating the cytotoxic agent clodronate (Van Rooijen, 1989) and carrying DiO did not label macrophages implying that these cells were rapidly eliminated; this was apparent from the absence of labeled nuclei lying between capillaries (Fig. 3B) and the lack of DiO-positive cells on pupillary membranes (Fig. 3D). Macrophage elimination was confirmed through the labeling of pupillary membrane preparations with the macrophage-specific anti-CR3 antibody (Fig. 3E-H).

Fig. 3.

Toxic liposomes clear macrophages from the anterior chamber. (A-D) Pupillary membranes at A11 after liposome injection at A7. (A,B) chromatin labelling with Hoechst 33258. (C,D) The same pupillary membrane preparations as in A and B, respectively, illuminated for DiO tagged control (C) and toxic (D) liposomes. (E-H) Micrographs of A8 pupillary membranes treated with liposomes at A7. (E,F) DIC illumination; (G,H) fluorescence illumination for CR3 receptor staining. (E,G) Control liposomes; (F,H) toxic liposomes.

Fig. 3.

Toxic liposomes clear macrophages from the anterior chamber. (A-D) Pupillary membranes at A11 after liposome injection at A7. (A,B) chromatin labelling with Hoechst 33258. (C,D) The same pupillary membrane preparations as in A and B, respectively, illuminated for DiO tagged control (C) and toxic (D) liposomes. (E-H) Micrographs of A8 pupillary membranes treated with liposomes at A7. (E,F) DIC illumination; (G,H) fluorescence illumination for CR3 receptor staining. (E,G) Control liposomes; (F,H) toxic liposomes.

To test whether macrophages might be required for the apoptosis occurring during programmed capillary regres- sion, we eliminated macrophages associated with the pupillary membrane and determined the proportion of putative target cells undergoing apoptosis. These experiments were performed using different volumes of toxic liposomes to determine if there was a dose response for macrophage elimination and for endothelial cell death. Toxic liposomes were injected at A7 and the numbers of macrophages and apoptotic endothelial cells quantitated 4 days later. Macrophage numbers decreased in a dose-dependent manner in response to liposome injections (Fig. 4A-C, dashed line). Importantly, in six independent experiments (Fig. 4A-D, solid line), the number of apoptotic endothelial cells also decreased in a dose-dependent manner.

Fig. 4.

Numbers of macrophages (dashed lines) and apoptotic endothelial cells (solid lines) in response to different liposome doses. Injections of liposomes were performed at A7 and pupillary membranes assessed histologically 4 days later. Six different experiments are shown (A-D) with numbers of macrophages determined in the first three (A-C).

Fig. 4.

Numbers of macrophages (dashed lines) and apoptotic endothelial cells (solid lines) in response to different liposome doses. Injections of liposomes were performed at A7 and pupillary membranes assessed histologically 4 days later. Six different experiments are shown (A-D) with numbers of macrophages determined in the first three (A-C).

As a further test of the impact of macrophage elimination on programmed regression, we examined the morphology of pupillary membranes at A14 after liposome injections at A3 and A7. This liposome delivery schedule resulted in an almost complete persistence of the pupillary membrane as assessed at A14 (Fig. 5A). Since only one eye of each animal received liposomes, the contralateral eye was used as an internal control; this invariably showed a pupillary membrane that was almost completely regressed at A14 (Fig. 5B). Comparison of an A7 normal pupillary membrane (Fig. 5C) with the toxic-liposome-treated membrane from A14 (Fig. 5A) indicated that liposomes prevented regression to a large degree. Pupillary membranes treated with control liposomes regressed with the normal schedule as indicated by the amount of the structure remaining at A14 (Fig. 5D).

Fig. 5.

Macrophage elimination results in a persistent pupillary membrane. (A) Pupillary membrane at A14 after injections of toxic liposomes at A3 and A7. (B) The contralateral, untreated pupillary membrane from the animal in (A). (C) A7 normal pupillary membrane. (D) A14 pupillary membrane at after injections of non- toxic liposomes at A3 and A7.

Fig. 5.

Macrophage elimination results in a persistent pupillary membrane. (A) Pupillary membrane at A14 after injections of toxic liposomes at A3 and A7. (B) The contralateral, untreated pupillary membrane from the animal in (A). (C) A7 normal pupillary membrane. (D) A14 pupillary membrane at after injections of non- toxic liposomes at A3 and A7.

Endothelial cells in normal pupillary membranes at A9 showed normal, non-apoptotic morphology (Fig. 6A). While occasional cells in A14 persistent membranes showed the morphological changes typical of apoptosis, the vast majority did not (Fig. 6B). However, given that programmed cell death was assessed by morphological criteria, it was possible that, while the persisting cells did not appear to undergo apoptosis, they were actually non-viable.

Fig. 6.

In the absence of macrophages, target cells in the pupillary membrane remain viable. (A) Hoechst 33258-labeled endothelial cell nuclei in a control A11 pupillary membrane have non-apoptotic morphology. (B) Nearly all the cells in an A14 persistent membrane have normal nuclear morphology. (C) BrdU-labeled A7 control pupillary membrane showing brown-labeled positive nuclei. (D) Persistent (A14) pupillary membrane after 4 BrdU injections over a 24 hour period. (E) Calcein AM-labeled normal A11 pupillary membrane shows bright green fluorescence as does (F), a calcein AM-labeled A15 persistent pupillary membrane. (G) Normal A7 pupillary membrane stained with calcein AM after being killed by fixation. (H) Persistent A14 pupillary membrane in DIC illumination showing capillary lumen containing erythrocytes.

Fig. 6.

In the absence of macrophages, target cells in the pupillary membrane remain viable. (A) Hoechst 33258-labeled endothelial cell nuclei in a control A11 pupillary membrane have non-apoptotic morphology. (B) Nearly all the cells in an A14 persistent membrane have normal nuclear morphology. (C) BrdU-labeled A7 control pupillary membrane showing brown-labeled positive nuclei. (D) Persistent (A14) pupillary membrane after 4 BrdU injections over a 24 hour period. (E) Calcein AM-labeled normal A11 pupillary membrane shows bright green fluorescence as does (F), a calcein AM-labeled A15 persistent pupillary membrane. (G) Normal A7 pupillary membrane stained with calcein AM after being killed by fixation. (H) Persistent A14 pupillary membrane in DIC illumination showing capillary lumen containing erythrocytes.

To address this issue, we used several methods for assessing cell viability. First, we determined whether persistent cells were able to synthesize DNA and thus incorporate bromodeoxyuridine (BrdU). Normal A9 membranes labeled with two injections of BrdU over a 3 hour period showed that 3.8±0.3% (n=3) of endothelial cells were in S-phase (Fig. 6C). Target cells persisting at A14 due to macrophage elimination incorporated BrdU at a similar level (4.4±1.1%, n=2) using the same labeling protocol; a 24 hour labelling period with 4 injections of BrdU resulted in 14.6±4.3% (n=3) BrdU-positive endothelial cells (Fig. 6D).

As a stringent test of viability for the entire persistent population, we labeled both normal and persistent cells with the vital dye and esterase substrate calcein AM (Papadopoulos et al., 1994). Characteristic yellow-green fluorescence was observed in the cells of normal pupillary membranes at A7 (Fig. 6E) as well as in persistent membranes at A16 (Fig. 6F). A calcein AM-labeled A7 membrane, killed through fixation, showed only low-level autofluorescence of endothelial cells and red blood cells (Fig. 6G). Capillaries from A16 persistent membranes also showed a distinctive lumen with erythrocytes present immediately after dissection (Fig. 6H). This suggested that persistent capillaries maintained blood flow, and this was confirmed in the live animal using a vital cell imaging system (Meeson et al., 1996; data not shown). These observations show that persistent cells of the pupillary membrane are cycling, are active in metabolism and maintain capillary patency. Together, these observations indicate that persistent cells are viable.

To test whether a purified population of macrophages could re-establish regression, we performed a reconstitution experiment. Toxic liposomes were first injected into the anterior chamber at A7 to effect macrophage elimination. At A12, 2×103 bone-marrow-derived macrophages in serum- free medium were injected and the consequences for pupillary membrane regression assessed. Controls included untreated membranes (from the contralateral eye) and membranes that received liposomes at A7 but medium only at A12. Membranes receiving liposomes but no macrophages were persistent (Fig. 7A) by comparison with untreated membranes from contralateral eyes (Fig. 7B). Liposome- treated membranes that also received macrophages were further regressed (Fig. 7C) than those that received liposomes only (Fig. 7A). Injected macrophages were labelled with fluorescent latex spheres and could be detected adjacent to capillaries of the pupillary membrane after injection into the anterior chamber (Fig. 7C,D). As a measure of the relative complexity of experimental and control membranes, we counted capillary junctions and showed that pupillary membranes receiving macrophages were significantly more regressed than those that did not (Fig. 7E). These experiments provide additional evidence that the macrophage is responsible for induction of programmed cell death and regression.

Fig. 7.

Macrophage injection reestablishes pupillary membrane regression. (A) Persistent pupillary membrane at A14 after liposome-mediated macrophage elimination at A7. (B) Contralateral eye of animal shown in A to indicate the normal degree of regression by this stage. (C) Pupillary membrane at A14 after treatment with liposomes at A7 and macrophages at A12. Exogenous macrophages are labeled with yellow-green fluorescent latex spheres. (D) Higher magnification of a liposome- and macrophage-treated pupillary membrane at A14 showing the association between injected macrophages and capillaries. (E) Quantitation of the number of capillary junctions in pupillary membranes that are treated only with liposomes, those that are treated with liposomes followed by macrophage reconstitution and those that are untreated. The number of capillary junctions in a pupillary membrane is a measure of the degree of regression. Error bars represent standard deviations.

Fig. 7.

Macrophage injection reestablishes pupillary membrane regression. (A) Persistent pupillary membrane at A14 after liposome-mediated macrophage elimination at A7. (B) Contralateral eye of animal shown in A to indicate the normal degree of regression by this stage. (C) Pupillary membrane at A14 after treatment with liposomes at A7 and macrophages at A12. Exogenous macrophages are labeled with yellow-green fluorescent latex spheres. (D) Higher magnification of a liposome- and macrophage-treated pupillary membrane at A14 showing the association between injected macrophages and capillaries. (E) Quantitation of the number of capillary junctions in pupillary membranes that are treated only with liposomes, those that are treated with liposomes followed by macrophage reconstitution and those that are untreated. The number of capillary junctions in a pupillary membrane is a measure of the degree of regression. Error bars represent standard deviations.

Our assertion that macrophages actively induce apoptosis in normal cells during programmed tissue regression rests on a number of elements. Specific killing of macrophages by liposomes is indicated by the restricted distribution of the fluorescent tracer in control liposomes and the dose response for macrophage number. Liposomes clearly do not kill endothelial cells; if this were so, the number of dying endothelial cells would be expected to increase with increasing liposome dose and the opposite is observed. Furthermore, it has previously been established that the cytotoxic activity of clodronate is liposome-dependent (Van Rooijen, 1989). The observation that apoptotic endothelial cells decrease in number with increasing liposome dose and that injected bone-marrow macrophages can mediate regression provides the strongest evidence that macrophages induce endothelial cell apoptosis.

These observations indicate that macrophages drive most, if not all aspects of capillary regression. Such an active role for macrophages in pupillary membrane remodeling explains the persistence of these vessels in mice expressing an ocular macrophage-specific toxin-expressing transgene (Lang and Bishop, 1993). The current work, together with experiments characterizing the sequence of events during pupillary membrane regression in vivo, imply that macrophage-induced endothelial cell apoptosis is an early event in regression of the capillary network (Lang et al., 1994; Meeson et al., 1996; Lang, 1997) and suggests that regulation of macrophage function may be one means to control capillary regression.

The results of our experiments reinforce the idea that the macrophage has a role in regulating vessel growth and regression (Sunderkotter et al., 1994). Macrophages produce a number of factors such as bFGF and VEGF that can stimulate proliferation of endothelial cells and it is likely that these have pro-angiogenic activity in situations such as wound healing (Martin, 1997). Macrophages can also produce anti-angiogenic mediators (Sunderkotter et al., 1994) and it is possible that one of the recognised activities might function in regression of the pupillary membrane. The simplicity and accessibility of the pupillary membrane means that it is an appropriate system in which to study the mechanistic details of capillary regression and macrophage-endothelial cell interactions in vivo.

These observations also raise the important possibility that macrophage-induced apoptosis in normal cells might be a more general phenomenon. While an answer to this question awaits further experimentation, it is clear that macrophages are ubiquitous at sites of developmentally programmed regression (for example, during elimination of interdigital tissue (Hopkinson-Woolley et al., 1994)) and it is possible that they may actively induce apoptosis in this setting. Macrophages might also display killing activity during tissue maintenance where there is a requirement for cell death and disposal. One can argue that to avoid the possible deleterious inflammatory effects of secondary necrosis, the macrophage might have evolved to provide a signal essential for cell death. This would ensure that the macrophage and its disposal activity were always present when cell death occurred.

We are indebted to Boehringer Mannheim GmbH for providing the clodronate (dichloromethylene bis-phosphonate) that was used in this study. We thank I. Riviere for assistance with macrophage production, Jon Weider for assistance with all aspects of imaging, Mike Matise for technical advice and Titia de Lange for comments on the manuscript. This work was supported by NIH grant RO1 EY11234 from the National Eye Institute. In addition, we acknowledge research support for the laboratory of R. A. L. from The Searle Scholars Program (grant number 94-F-118) of the Chicago Community Trust. G. D.-R. was supported in part by the Cell and Molecular Biology Graduate Training Program funded by NIH grant number GM07238.

Aliprantis
,
A. O.
,
Diez Roux
,
G.
,
Mulder
,
L.
,
Zychlinsky
,
A.
and
Lang
,
R.
(
1996
).
Do macrophages kill through apoptosis?
Immunology Today
17
,
573
576
.
Bernasconi
,
S.
,
Peri
,
G.
,
Sironi
,
M.
and
Mantovani
,
A.
(
1991
).
Involvement of leukocyte (β2) integrins (CD18/CD11) in human monocyte tumoricidal activity
.
Int. J. Cancer
49
,
267
273
.
Cui
,
S.
,
Reichner
,
J. S.
,
Mateo
,
R. B.
and
Albina
,
J. E.
(
1994
).
Activated murine macrophages induce apoptosis in tumor cells through nitric oxide-dependent or -independent mechanisms
.
Cancer Research
54
,
2462
2467
.
Hopkinson-Woolley
,
J.
,
Hughes
,
D.
,
Gordon
,
S.
and
Martin
,
P.
(
1994
).
Macrophage recruitment during limb development and wound healing in the embryonic and foetal mouse
.
J. Cell Sci
.
107
,
1159
1167
.
Lang
,
R. A.
(
1997
).
Apoptosis in mammalian eye development: lens morphogenesis, vascular regression and immune privilege
.
Cell Death and Differentiation
4
,
12
20
.
Lang
,
R. A.
and
Bishop
,
M. J.
(
1993
).
Macrophages are required for cell death and tissue remodeling in the developing mouse eye
.
Cell
74
,
453
462
.
Lang
,
R. A.
,
Lustig
,
M.
,
Francois
,
F.
,
Sellinger
,
M.
and
Plesken
,
H.
(
1994
).
Apoptosis during macrophage-dependent tissue remodelling
.
Development
120
,
3395
3403
.
Martin
,
P.
(
1997
).
Wound healing–Aiming for perfect skin regreneration
.
Science
276
,
75
81
.
Matsuo
,
N.
and
Smelser
,
G. K.
(
1971
).
Electron microscopic studies on the pupillary membrane: The fine structure of the white strands of the disappearing stage of this membrane
.
Invest. Ophthalmol
.
10
,
1098
119
.
Meeson
,
A.
,
Palmer
,
M.
,
Calfon
,
M.
and
Lang
,
R. A.
(
1996
).
A relationship between flow and apoptosis during programmed capillary regression is revealed by vital analysis
.
Development
122
,
3929
3938
.
Papadopoulos
,
N. G.
,
Dedoussis
,
G. V.
,
Spanakos
,
G.
,
Gritzapis
,
A. D.
,
Baxevanis
,
C. N.
and
Papamichal
,
M.
(
1994
).
An improved fluorescence assay for the determination of lymphocyte-mediated cytotoxicity using flow cytometry
.
J. Immunol. Methods
177
,
101
111
.
Robinson
,
A. P.
,
White
,
T. M.
and
Mason
,
D. W.
(
1986
).
Macrophage heterogeneity in the rat as delineated by two monoclonal antibodies MRC OX-41 and MRC OX-42, the latter recognizing complement receptor type 3
.
Immunology
57
,
239
247
.
Sulston
,
J. E.
,
Albertson
,
D. G.
and
Thompson
,
J. N.
(
1980
).
The Caenorhabditis elegans male: Postembryonic development of non-gonadal structures
.
Dev. Biol
.
78
,
542
576
.
Sunderkotter
,
C.
,
Steinbrink
,
K.
,
Goebeler
,
M.
,
Bhardwaj
,
R.
and
Sorg
,
C.
(
1994
).
Macrophages and angiogenesis
.
J. Leuk. Biol
.
55
,
410
422
.
Takahashi
,
T.
,
Nowakowski
,
R. S.
and
Caviness
,
V. S.
(
1993
).
Cell cycle parameters and and patterns of nuclear movement in the neocortical proliferative zone of the fetal mouse
.
J. Neuroscience
13
,
820
833
.
Urban
,
J. L.
,
Shepard
,
H. M.
,
Rothstein
,
J. L.
,
Sugarman
,
B. J.
and
Schreiber
,
H.
(
1986
).
Tumor necrosis factor: A potent effector molecule for tumor cell killing by activated macrophages
.
Proc. Natl. Acad. Sci. USA
83
,
5233
5237
.
van de Loosdrecht
,
A. A.
,
Ossenkoppele
,
G. J.
,
Beelen
,
R. H. J.
,
Broekhoven
,
M. G.
,
Drager
,
A. M.
and
Langenhuijsen
,
M. M. A. C.
(
1993
).
Apoptosis in tumor necrosis factor-a-dependent, monocyte-mediated leukemic cell death: a functional, morphologic and flow-cytometric analysis
.
Exp. Hematol
.
21
,
1628
1639
.
Van Rooijen
,
N.
(
1989
).
The liposome-mediated macrophage ‘suicide’ technique
.
J. Immunol. Meth
.
124
,
1
6
.
Vlassara
,
H.
,
Brownlee
,
M.
,
Manague
,
K. R.
,
Dinarello
,
C. A.
and
Pasagian
,
A.
(
1988
).
Cachectin/TNF and IL-1 induced by glucose-modified proteins: Role in normal tissue remodeling
.
Science
240
,
1546
1548
.
Wyllie
,
A. H.
,
Kerr
,
J. F. R.
and
Currie
,
A. R.
(
1980
).
Cell death: The significance of apoptosis
.
Int. Rev. Cytol
.
68
,
251
305
.