- Email: [email protected]
Mast cells enhance migration and proliferation of fibroblasts into an in vitro wound
EXPERIMENTAL
CELL
RESEARCH
188,4%49
(1990)
Mast Cells Enhance Migration and Proliferation of Fibroblasts into an in Vitro Wound FRANCESCA LEVI-SCHAFFER’ Department
of Pharmacology,
Hebrew
University-Hadassah
Medical
INTRODUCTION
Mast cells (MC) are high-affinity IgE-receptor-bearing cells that stain metachromatically with cationic dyes. Activated MC release the mediators stored in their cytoplasmic granules (histamine, proteoglycans, proteolytic enzymes, chemotactic factors, etc.) and both synreprint
requests
should
0014~4827/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
POB
1172, Jerusakm
MATERIALS
91010,
Israel
AND
METHODS
Materials. RPMI-1640, nonessential amino acids, L-glutamine, streptomycin, and penicillin were purchased from Grand Island Biological Co. (Grand Island, NY); metrizamide, compound 48/80, fetal
be addressed. 42
Inc. reserved.
School,
thesize and release membrane phospholipid metabolites such as prostaglandins, leukotrienes, and platelet-activating factor. MC play a recognized, pivotal role in allergic reactions [ 11, and their increased presence in a number of chronic inflammatory conditions, often accompanied by fibrosis, would suggest that they are involved in these pathologic situations as well [2,3 1. Wound healing is a very complex event which involves interactions of various cell types such as lymphocytes, monocytes, epithelial cells, and fibroblasts. Three main overlapping phases have been identified in tissue response to injury: inflammation, granulation tissue and matrix formation, and remodeling. During granulation tissue formation fibroblasts proliferate and migrate into the wound space [4]. Both in normal reparative as well as in pathological wound healing, the latter leading to hypertrophic scars and keloids, MC have been found in increased numbers [5-71. However, studies carried out to define a role for MC in any of the stages of wound healing that utilized in viva techniques [6-lo] and more recently in vitro organ cultures [ll, 121 have given no definitive results as yet. We have become interested in the relationships between MC and fibroblasts and have demonstrated that in vitro MC survive, change in phenotype, and regenerate activation when cultured with fibroblasts [13-161. The aim of this study was to investigate the possible involvement of MC in wound healing, specifically by studying the relationships between MC and fibroblasts in an in vitro defined wound model. This system consisted of rat peritoneal MC cultured on confluent 3T3 fibroblast monolayers which were wounded by performing a linear cut along the midline and scraping one half of the monolayer [17]. We have examined the extent of fibroblast migration and proliferation into the cell-free area in the presence of resting and activated MC. The results obtained point to a facilitating role of MC in these fibroblast processes.
The effects of mast cells (MC) in an in vitro wound model were studied. The model consisted of rat peritoneal MC cultured on confluent monolayers of 3T3 fibroblasts (MC/3T3). A linear wound was performed by cutting along the midline and scraping one half of the monolayer. After 42 h fibroblasts were counted in the scraped area of the wound. In the MC/3T3 cocultures 27.6 + 2.1 fibroblasts were found compared to 16.6 + 0.9 in the 3T3 cultures. The most significant increase in the number of fibroblasts was obtained upon activation of the MC with anti-IgE antibodies immediately after wound production (39.9 + 2.1). Stimulation with compound 48/80 had a weaker effect (32.7 + 1.6). Incubation of 3T3 wounded monolayers with supernatants of anti-IgE- or compound 48/80-activated MC, or with sonicated MC, induced an increase in fibroblast number similar to that found in unactivated MC/3T3. [3H]Thymidine incorporation followed by autoradiography was performed to assess fibroblast mitosis. The highest number of labeled fibroblasts beyond the wound line was found in immunologically activated MC/3T3 (29.7 + 4.4), followed by compound 48/80-activated MC/3T3 (18.4& 1.5),MC/3T3(15.1+3.6),and3T3(10.6-t2.6). After addition of aphidicolin, which inhibited fibroblast mitosis, MC were still effective in enhancing fibroblast migration. In all the cocultures MC were observed to have migrated alongside fibroblasts. Thus merely the presence of MC adhering to wounded fibroblast monolayers significantly enhanced migration and proliferation of the fibroblasts. A further increase was achieved by immunological activation of the MC. We therefore suggest that MC have a facilitating role in this in vitro wound model. o 1990 Academic press. IDC.
1 To whom
AND ARI KUPIETZKY
MAST
CELLS
ENHANCE
FIBROBLAST
calf serum, toluidine blue, trypan blue, gelatin, histamine dihydrochloride, and aphidicolin from Sigma Chemicals (St. Louis, MO); 35-mm culture dishes from Falcon, Be&on-Dickinson and Co. (Cockeysville, MD); 14C-labeled methyl-S-adenosyl-L-methionine (41 mCi/mmol) and thymidine (methyl-3H) (1 Ci/mmol) from New England Nuclear (Boston, MA); Fujichrome colored photo films for 100 ASA slides from Fuji Film Co. (Tokyo, Japan); rabbit IgG anti-rat IgE and normal rabbit serum from Bethyl Laboratories (Montgomery, TX); Developer D19, Fixer, NTB2 photographic emulsion gel from Eastman Kodak (Rochester, NY); 0.6-mm stainless-steel blades from KDS-Hi-Cutter (Tokyo, Japan).
3T3 mowe firoblast
cultures.
mouse skin fibroblast cell broblast monolayers were mm culture dishes in 2 ml ing 10% heat-inactivated 100 rg/ml streptomycin, amino acids, pH 7.2, EM) when monolayers reached seeded in some of the 3T3
The contact-inhibited Swiss albino line (ATCC, Rockville, MD) was used. Fiprepared by seeding 1 X 10’ 3T3 cells in 35of enriched medium (RMPI-1640 containfetal calf serum (v/v), 100 U/ml penicillin, 2 mM L-glutamine, 0.1 mM nonessential [13]. These cultures were used 7 days later confluency. At this time point, MC were cultures as described below.
Mast cells/3T3 fibroblasts coculture (MC/3T3). Mast cells (MC) were isolated and purified (99% purity) from the peritoneal cavity of 250- to 300-g male Sabra rats (an outbred strain of the Hebrew University), as previously described [13]. MC were seeded at a density of 8 X 10’ cells/2 ml EM on confluent monolayers of 3T3 fibroblasts and incubated at 37°C in a 5% COP humidified incubator. The MC adhere to the fibroblast monolayer in about 10 min after seeding. Cocultures were fed every other day with 2 ml of fresh medium. Wound production in 3T3 and MCf3T3 cultures.
A wound was produced in both MC/3T3 (48-72 h after MC seeding) and 3T3 cultures by a blade cut essentially as described by Stewart et al. [17]. A 0.6-mm steel blade was cut with a high-speed rotary Carborundum stone, to fit the 35-mm diameter of the culture dish. The blade was hand held with the aid of forceps and a downward pressure was applied ensuring the formation of an evenly marked edge. A cotton swab was used to scrape off the fibroblasts from one side of the blade in the designated cell-free area of the dish. A 191 -gauge needle fit to a syringe was used to aspirate cell debris and medium. The wounded monolayer was then washed once with Tyrode’s buffer containing 0.1% gelatin (TG). Three distinct zones were therefore present in the culture dish: (1) a cell-rich area; (2) a marked edge termed “wound line”; (3) a cell-free area termed “wound” (Ref. [17] and Fig. 1A). All the dishes were routinely examined under an inverted phase-contrast microscope (Olympus IM-2).
MIGRATION
AND
43
PROLIFERATION
Fixation and staining of cultures. Fixation and staining of the 3T3 and MC/3T3 cultures was simultaneously performed by incubation with an ethanolic solution of 0.07% toluidine blue, pH 3.5, for 1 min at RT. Stained cultures were rinsed with absolute ethanol and allowed to dry at RT. Autoradiography of cultures. Following incubation of the wounded cultures with the different supematants, 2 ml EM containing 2 &i [3H]thymidine were added. In some experiments in which we wanted to stop fibroblasts mitosis, aphidicolin (5 pg/ml) was added simultaneously with [3H]thymidine. This aphidicolin concentration was found to stop mitosis of fibroblasts without being cytotoxic for either fibroblasts or MC as detected by the absence of cell staining with trypan blue. The medium was removed after 42 h incubation of the cultures in the CO2 incubator and 2 ml of methanol/acetic acid solution (3/l, v/ v) was added for 15 min at RT. The fixative was then removed and the cultures left to dry. Under dark room conditions, a thin layer of NTBz emulsion gel in water (l/l, v/v) was added and the cultures left to dry for 1 h. Dishes were then wrapped in a light-proof packaging and stored for 24 hat 4°C. Developer D-19 and Fixer were prepared according to the manufacturer’s instructions (Kodak). One milliliter of the developer D19 was added to each plate and left for 2 min. The dishes were washed once with 1 ml of distilled water, and 2 ml of fixer was added for 4 min. The fixer was then removed and the dishes were dipped into water for 1 min. The plates were allowed to dry, stained with toluidine blue in ethanol for 8 min, rinsed with ethanol, and allowed to dry again. Quantitation cultures. Fixed
of fibroblnsts
and
mast cells in the wounded
and stained and/or autoradiographed cultures were examined under a light microscope (Olympus BH-2, Tokyo, Japan) equipped with an eyepiece graticle (graduated grid) to allow cell quantification. The grid contained a 10 by 10 matrix of l.O-mm squares. Samples were counted at a X200 or x400 magnification. In the stained dishes the total number of fibroblasts found beyond 250 pm of the wound line was quantitated in at least six different fields. In the autoradiographed cultures labeled and unlabeled fibroblasts were counted in the cell-rich area adjacent to the wound line and in the wound area, in at least four different fields. MC were counted in the same areas of the cultures. Statistical analysis of the data. Data were analyzed by Student’s t test.
RESULTS
Treatment of wounded cultures.
Immediately after wounding, 3T3 and MC/3T3 cultures were incubated for 10 min at 37°C in the COP incubator with 1 ml of one of the following supernatants: (1) TG++ buffer (TG containing 1.8 mM CaClz and 0.9 mM MgCl,); (2) TG++ containing compound 48/80 (3 fig/ml); (3) TG++ containing rabbit anti-rat IgE (l/40, v/v); (4) TG++ containing normal rabbit serum (l/40, v/v); (5) supernatants of freshly isolated MC activated with compound 48/80; (6) supernatants of freshly isolated MC activated with anti-IgE antibodies; (7) suspensions of freshly isolated and sonicated MC. Supernatants 5 and 6 were prepared, respectively, by incubating 8 X lo4 freshly isolated MC with 1 ml TG++ containing either compound 48/80 (3 &ml) or rabbit anti-rat IgE (l/40) for 10 min at 37°C. After pelleting the MC at 12Og for 5 min, supernatants were collected and immediately added to the wounded 3T3 cultures. The suspension of sonicated MC was prepared by disrupting 8 X lo4 MC/ 1 ml TG+t by continuous sonication for 30 s (output 5, 50% duty cycle, Heat Systems Ultrasonics W380). After incubation of the wounded cultures with the different supernatants, supernatants were collected and the cell monolayers were washed with 1 ml TG++. Two milliliters of EM were added and the cultures were placed in the COP incubator for a period of 42 h (unless otherwise stated) after which they were either fixed and stained or processed for autoradiography as described below.
“Wound” Conditions A wound line was produced in cultures of confluent fibroblast monolayers alone (3T3) and of fibroblasts cocultured with rat peritoneal mast cells (MC/3T3) (Fig. 1A). As can be seen in Fig. 2,26 and 42 h after the wound production there was a significant difference in the number of fibroblasts found beyond the wound line between the 3T3 and the MC/3T3 cultures (P < 0.02 at 26 h, P < 0.01 at 42 h). This difference was no more statistically significant at 50 h. Therefore, we adopted the 42-h time point for all the subsequent experiments. Effect of Mast Cells and Mast Cell Mediators on Fibroblast Numbers beyond the Wound Line The effect of MC on the wounded fibroblast monolayers was checked by counting the number of fibro-
44
LEVI-SCHAFFER
AND
KUPIETZKY
FIG. 1. 3T3 fibroblasts cocultured with rat peritoneal mast cells (MC/3T3) immediately after wounding of the cell monolayer and compound 48/80-induced activation of the mast cells. (A) The wound line (Z), marked by the blade on the plate bottom, divides the undisturbed cell-rich area adjacent to the wound (1) from the cell-free area termed “wound” (3). As shown, no fibroblasts or MC are seen in the cell-free area. Cultures were stained with toluidine blue (X95). (B) In the cell-rich area adjacent to the wound, activated MC exhibit their classical activated raspberry-like appearance. Many MC granules (arrows) appear on the wounded 3T3 monolayer (X190).
blasts found beyond the wound line in 3T3 or MC/3T3 cultures subjected to various treatments. After wound production (time 0), MC/3T3 were incubated for 10 min with buffer, buffer containing compound 48/80, or antiIgE antibodies (Fig. 3). Alternatively, 3T3 cultures were incubated with buffer or supernatants obtained from immunologically or compound 48/80-activated MC or sonicated MC (Fig. 4). As can be seen in Fig. 3a, in the 3T3 cultures incubated with buffer alone the number of fibroblasts in the cell-free area was 16.6 f 0.9 (n = 18, mean zk SE). In experiments in which we compared the effects of buffer alone versus buffer containing com-
pound 48/80 or anti-IgE, we obtained similar values for thesethreecontrolgroups,i.e.,17.3+1.5,17.3+_1.1,and 16.9 + 1.3, respectively (mean + SE, n = 3). Therefore, in all the subsequent experiments we used 3T3 cultures incubated with buffer alone as the only control group. Coculture of MC with fibroblasts significantly increased
Z
+ zi
30-
mz $0 a 3 h s?I
20 -
.. .. .. .:a .. .. ..
m0
ii%
1(-J-
k 10
30
50
HOURS AFTER “WOUNDING” FIG. 2. Time course of the appearance of 3T3 fibroblasts beyond the wound line. Fibroblasts were counted along the 250~pm length of the wound line in 3T3 (0 - - 0) and MC/3T3 (0 - 0) cultures at the indicated time points. Closure of both wounds occurred within 6 days. Data are the mean + SE of a representative experiment performed in triplicate.
s o- jll
a
1:: .. .. .. ... b c
d
FIG. 3. Effect of unactivated and activated MC on fibroblast numbers beyond the wound line. The number of fibroblasts found beyond the 250-pm wound line at 42 h after wounding is shown in the following cultures: 3T3 (a), MC/3T3 (b), 3T3/MC activated by compound 48/80 (c), and 3T3/MC activated by anti-IgE (d).
MAST
CELLS
ENHANCE
FIBROBLAST
MIGRATION
AND
45
PROLIFERATION
not differ from that obtained in MC/3T3 cultures incubated with buffer alone and were lower than those obtained in activated MC/3T3 cocultures (see Fig. 3). Morphological Studies of the Wounded MC/3T3 Cultures
A
B
C
D
FIG. 4. Effect of sonicated MC and the supernatants of activated MC on fibroblast numbers beyond the wound line. The number of fibrohlasts found beyond the 250~pm wound line at 42 h after wounding is shown in 3T3 cultures incubated with: buffer (A); suspension of sonicated freshly isolated MC (B); supernatant of compound 48/80-activated MC (C); supernatant of anti-IgE-activated MC (D).
the number of fibroblasts beyond the wound line (27.6 + 2.1, mean + SE, n = 9, P < 0.01) (Fig. 3b). When cocultured MC were stimulated to release their mediators by addition of an optimal concentration of compound 48/ 80 (Fig. 3c) similar values were obtained (32.7 f 1.5, mean f SE, n = 10, P < 0.1). Immunological stimulation of the MC/3T3 cultures with anti-IgE antibodies resulted in the highest number of fibroblasts counted over the wound line, i.e., 39.9 f 2.1 (n = 8, mean + SE) (Fig. 3d). This value was significantly higher than both that obtained in unstimulated MC/3T3 (P < 0.01) and that of compound 48/80-activated MC/3T3 (P < 0.05). The effect of rabbit anti-rat IgE antibodies on MC/3T3 was not due to the presence of rabbit serum in the supernatant. In fact, when wounded MC/3T3 were incubated with normal rabbit serum the number of fibroblasts found over the wound line was comparable to that of cultures incubated with buffer alone (27.0 f 2.2 versus 27.6 f 2.1, mean + SE, n = 2). In Fig. 4 the results obtained by incubating 3T3 cultures with different MC supernatants are depicted. Addition of sonicated MC to the wounded 3T3 monolayer (Fig. 4B) caused a slight though significant increase in the fibroblast numbers in comparison to 3T3 cultures incubated with buffer alone (Fig. 4A) (23.8 t- 2.7, versus 16.6 _+0.9, mean f SE, n = 11, P < 0.05). The supernatants of both compound 48/80 (Fig. 4C) and anti-IgE (Fig. 4D)-activated MC caused also an increase in the fibroblast numbers compared to the buffer-incubated 3T3 cultures (26.6 + 1.8 and 29.1 + 2.4, P < 0.01 and P < 0.001, n = 10, mean * SE). However, these values did
Wounded MC/3T3 cultures were stained with acidic toluidine blue and examined under the light microscope 10 min and 42 h after the different incubation procedures. As expected, 10 min after activating the MC/3T3 cultures, MC lost their resting appearance and acquired the classical activated raspberry-like appearance. Many metachromatic MC granules appeared on the wounded monolayer (Fig. 1B). Fibroblast monolayers appeared stretched and “empty spaces” were formed between the fibroblasts. At this time point, no fibroblast and no MC were detected over the wound line (Fig. 1A). Forty-two hours later as expected, the activated MC reassumed their resting appearance, although they were less granulated than in control-unactivated cocultures [ 161. Fibroblast monolayers appeared as before MC activation. At this time point, in all the cultures, fibroblasts were found over the wound line (Figs. 5A-5D). In stimulated MC/3T3 cultures (Figs. 5C and 5D), fibroblasts migrated and proliferated in higher numbers (see also Figs. 3 and 4) than in 3T3 (Fig. 5A) and MC/3T3 (Fig. 5B) control cultures. Both in unstimulated (Fig. 5B) and stimulated (Figs. 5C and 5D) cocultures, MC were detected over the wound line, morphologically associated with fibroblasts. MC numbers were quantitated and compared in unstimulated, compound 48/80-activated and in anti-IgEactivated MC/3T3 cultures. No difference between the numbers of MC adjacent to the wound line and in the wound area was found in these three different cultures. However, the number of MC found in the wound was always much lower than that found in the area adjacent to the wound (2.5 + 0.3/500 PM in the wound line compared to 10.9 + 0.9/500 PM adjacent to the wound line, n = 4 + SE). Determination of Fibroblast Mitosis 3T3 and MC/3T3 Cultures
in the Wounded
After the lo-min incubation procedures, 3T3 and MC/ 3T3 wounded cultures were incubated for 42 h with EM containing [3H]thymidine and were autoradiographed to assess the extent of fibroblast mitosis. Fibroblasts, both near the wound and far from the wound, were heavily labeled under all the culture conditions. In Fig. 6 the number of labeled fibroblast nuclei and the total number of fibroblasts (labeled plus unlabeled) found beyond the wound line are compared in the differently treated cultures. The highest value of labeled fibroblasts was found in the immunologically activated MC/3T3 (29.7 + 4.4, mean f SE, n = 3 (Fig. 6d)), fol-
46
LEVI-SCHAFFER
AND
KUPIETZKY
FIG. 5. After 42 h of incubation, numerous fibroblasts are foundbeyond the wound line. 3T3 fibroblasts alone (A), MC/3T3 (B), compound 48/80-activated MC/3T3 (C), and anti-IgE activated MC/3T3 (D). Cultures were examined after fixation and toluidine blue staining. The maximum number of fibroblasts was found in the immunologically activated cocultured MC/3T3 (D). In all cocultures, MC were detected beyond the wound line (arrows) (X100).
lowed by compound 48/80-activated cultures (18.4 5 1.5, mean + SE, n = 3 (Fig. 6c)), buffer-incubated MC/3T3 (15.1+ 3.6, mean -+_SE, n = 3 (Fig. 6b)) and buffer-incu-
bated 3T3 (10.6 f 2.6, mean k SE, n = 3 (Fig. 6a)). The total number of fibroblasts over the wound line followed the same pattern of the labeled fibroblasts, i.e., the
MAST
CELLS
ENHANCE
FIBROBLAST
MIGRATION
AND
47
PROLIFERATION
TABLE
1
Effect of Aphidicolin on Fibroblast Numbers beyond the Wound Line Culture
Fibroblast 12.3 26.0 3.6 12.9
3T3 MC/3T3 3T3 + aphidicolinb MC/3T3 + aphidicolin Note. Fibroblasts were counted bated with anti-IgE in the absence n The data (n = 5, mean + SE) found beyond the wound line (per ’ Aphidicolin concentration was
FIG. 6. Number of labeled fibroblasts (la) and total fibroblasts (labeled and unlabeled (0) found beyond the wound line (250 pm) in: 3T3 (a), MC/3T3 (b), compound 48/Wactivated MC/3T3 (c), and anti-IgE-activated MC/3T3 (d). [3H]Thymidine was added immediately after wounding and autoradiography was performed 42 h later.
higher the number of fibroblasts in mitosis, the higher the number of fibroblasts found in the cell-free area. When the number of labeled fibroblasts in the cellrich area was evaluated, results similar to those found for the fibroblasts over the wound line were obtained (data not shown). No MC were found to have incorporated [3H]thymidine indicating that under our tissue culture conditions these MC do not undergo mitosis. Also in autoradiographed cultures, MC were found beyond the wound line associated with fibroblasts (Fig. 7).
FIG. 7. Fibroblasts and mast cells beyond the wound line. The culture (MC/3T3) was treated as described in the legend to Fig. 6b. After autoradiography MC/3T3 were stained with toluidine blue to evince the MC (arrows) that were not labeled by [3H]thymidine (X460). MC apparently have migrated alongside with the fibroblasts to which they had originally adhered.
numbeP k 1.4 k 2.3 f 0.9 z!z 1.5
in 3T3 and MC/3T3 cultures incuand presence of aphidicolin. express the number of fibroblasts 250 pm wound line) at 42 h. 5 pg/ml.
To determine the contribution of fibroblast division in the wound closure, aphidicolin, an antimitotic compound, was added together with [3H]thymidine to 3T3 and MC/3T3 after their incubation with buffer or antiIgE antibodies. Forty-two hours later autoradiography was performed. At the drug concentration used complete inhibition of [3H]thymidine incorporation by 3T3 fibroblasts was achieved. As shown in Table 1, also under these conditions fibroblasts were found over the wound line. In the immunologically activated MC/3T3 cultures, the number of fibroblasts that migrated in the cell-free area in the absence of aphidicolin was nearly double that inits presence (26.0 + 2.3 versus 12.9 f 1.5, mean + SE, n = 5). In 3T3 cultures alone, 12.3 + 1.4 and 3.6 ? 0.9 (mean + SE, n = 5) fibroblasts were found over the wound line in the absence and in the presence of the drug, respectively. DISCUSSION There has been increasing interest in the relationships between mast cells (MC) and fibroblasts since the observation that MC are found in increased numbers in a series of fibrotic conditions such as wound healing, idiopathic pulmonary fibrosis [18], bleomycin-induced fibrosis [ 191, scleroderma [ 201, etc. The present study was undertaken to investigate the effects of the presence and activation of rat peritoneal MC on fibroblasts in an in vitro wound model consisting of wounded monolayers of confluent 3T3 fibroblasts. We have found that MC influence this in vitro wound healing process by increasing fibroblast migration and proliferation. We have previously shown that rat peritoneal MC cocultured with 3T3 fibroblasts rapidly adhere to the monolayer and keep their viability, unique phenotype, and functional activity for over a month [ 131. Moreover, when induced to degranulate, the cocultured MC released their mediators and, in a period of several days regenerated morphologically, functionally, and bio-
48
LEVI-SCHAFFER
chemically [16]. Likewise, the properties of 3T3 fibroblast monolayers are well defined [21,22]. If a wound is made by scratching the cell monolayer, cells migrate into the denuded area, synthesize DNA, and divide, and the wound is thereafter repaired [21,23,24]. Therefore this MC/3T3 culture represented a unique defined in vitro system in which to look for a role of MC in the wound-healing fibrotic stage. We have found that the mere presence of MC on the wounded 3T3 monolayer enhances significantly the migration and proliferation of fibroblasts into the cell-free area (Fig. 2). When the cocultured MC were activated to degranulate by the basic polyamine compound 48/80 these processes were enhanced. Even more impressive results were obtained by stimulating the MC by anti-IgE antibodies (Figs. 3 and 6). Less prominent effects on fibroblast migration and proliferation were observed when the wounded monolayers were incubated with sonicated MC or supernatants from activated MC instead of intact/alive MC (Fig. 4). Sonicated MC suspension contains all the MC preformed mediators such as histamine, proteoglycans, proteolytic enzymes, chemotactic factors, etc., while supernatants of compound 48/80 and anti-IgE-activated MC contain both the preformed mediators and the newly formed mediator, PGD, [25]. It would seem therefore that not only a MC mediator(s) is important in obtaining the enhanced fibroblast migration/proliferation into the wound, but close proximity between MC and fibroblasts can further enhance this process. We have previously shown that MC-fibroblast interactions result in a facilitation of the secretion/degranulation processes of both the rat peritoneal MC and the mouse bone-marrow-derived MC [25, 261. In these studies, including the present one, there is no mechanistic explanation yet available for this obligatory proximity between MC and fibroblasts. Addition of aphidicolin, a specific inhibitor of nuclear DNA synthesis with no detectable effect upon RNA or protein synthesis [ 271, to the wounded 3T3 and MC/3T3 cultures caused a total blockade of fibroblast mitosis. However, fibroblasts still actively migrated beyond the wound line in both untreated and anti-IgE-treated 3T3 and MC/3T3 cultures (Table 1). In these series of experiments migration of fibroblasts was much more prominent in the presence of MC then in 3T3 fibroblast cultures alone, indicating that MC also positively influence this fibroblast property. An interesting observation that might have in uiuo relevance is that in the wounded MC/3T3 cultures, MC were always found over the wound line, apparently having migrated alongside with the fibroblasts (Fig. 7). While the enhancement effect MC had on fibroblast migration in our in vitro wound system is a novel described finding, the mitogenic effect of activation products of MC is a well-described phenomenon at least in. uiuo. For example, Norby and his colleagues have shown
AND
KUPIETZKY
that mesenchymal cell proliferation is caused by a product secreted by MC when stimulated by compound 48/ 80; they concluded that MC histamine is one of several possible important MC-related mitogens [ 111. Roche [28], on the other hand, suggested that heparin contained in the MC granules is a factor that can induce fibroblast proliferation and, being angiogenic, could help in the formation of granulation tissue. In summary, the results presented in this study would suggest that at least one of the possible roles of MC in wound healing might be that of enhancing fibroblast migration and proliferation into the wound. Studies aimed at defining which MC mediator(s) is responsible for both these effects are currently being carried out in our laboratory. We thank Professor S. Shoshan (Department of Oral Biology) for many helpful discussions, Professor J. Kapitulnik (Department of Pharmacology) for the critical reading of the manuscript, and Mrs. Ana Fihach for her skillful secretarial assistance. This work is a partial fullfilment by Mr. Ari Kupietzky of the requirement for the M.Sc. degree in the Hebrew University-Hadassah Medical School. Mr. Ari Kupietzky was supported by a fellowship from the Foulkes Foundation, London. This project was partially supported by a donation to F.L-.S. from the Anticoli Family in memory of their parents.
REFERENCES 1.
Galli, S. J., and Lichtenstein, L. M. andBasophils in Allergy, Principles Reed, C. F., Ellis, E. F., Adkinson, Eds.), 3rd ed., pp. 106-134, Mosby,
2. Bienenstock, Dolovich,
J., Tomioka, J., and Denburg,
(1988) Biology of Mast Cells and Practice (Middleton, E., N. F., and Yunginger, J. W., St. Louis.
M., Stead, R., Ernst, P., Gauldie, J. (1987) Amer. Rev. Respir. Dis.
J.,
135,
55-58. 3. Claman, H. N. (1985) Zmmunol. Today 6,192-195. 4. Clarck, R. A. (1985) J. Amer. Acad. Dermatol. 13,701-725. 5. K&her, C. W., Bunce, H., III, and Shetlar, M. R. (1978) J. Znvest. Dermatol. 70,355-357. 6. Matsuda, H., and Kitamura, Y. (1981) Exp. Hematol. 9,38-43. 7. Fitzpatrick, D. W., and Fisher, H. (1982) Surgery 91,430-434. 8. Franzen, L., and Norby, K. (1983) Acta Pathol. Microbial. Zmmunol.Scand.Sect.A 91,281-289. 9. Persinger, M. A., Lepage, P., Simard, J. P., and Parker, G. H. (1983) Brit. J. Dermatol. 108,179-187. 10. 11. 12.
Schittek, A., Demetriou, A. A., Padawar, enson, S. M. (1984) Agents and Actions
J., Seifter,
E., and Lev-
15,172-176. Norby, K. (1985) Agents Actions 16(3/4), 287-290. Druvefors, P., and Norby, K. (1988) Virchws Arch. I?: Cell Pathol. 55,187-192.
13.
Levi-Schaffer, F., Austen, K. F., Caulfield, J. P., Hein, A., Bloes, W. F., and Stevens, R. L. (1985) J. Zmmunol. 135,3454-3462.
14.
Levi-Schaffer, vallese, P. M.,
F., Austen, K. F., Caulfield, and Stevens, R. L. (1987)
J. P., Hein,
A., Gra-
J. Zmmunol. 139,494-
500. 15. 16.
Levi-Schaffer, R. L. (1986) Levi-Schaffer,
40.
F., Au&en,
K. F., Gravallese,
P. M.,
and Stevens,
Proc. Natl. Acad. Sci. USA 83.6485-6486. F., and Riesel, N. (1989) Cell Zmmurwl. 119,30-
MAST
17.
Stewart,
R. J., Duley,
Exp. Path& 18. 19.
CELLS
ENHANCE
J. A., and Allardyce,
FIBROBLAST
R. A. (1979)
&it.
MIGRATION
J.
60,582-588.
Kawanami, O., Ferrans, V. Y., Fulmer, J. D., and Crystal, (1979) Lab. Invest. 40,717-734. Goto, T., Befus, D., Low, R., and Bienenstock, J. (1984)
Rev. Respir. Dis. 130,797-802. 20. Hawkins, R. A., Claman, H. N., Clark, R. A. F., and Steigerwald, J. C. (1985) Ann. Intern. Med. 102,182-186. 21. Lipton, A., Klinger, I., Paul, D., and Holley, R. W. (1971) Proc. Natl. Acad. Sci. USA 68,2799-2801. 22. Dunn, G. A., and Ireland, G. W. (1984) Nature (London) 312, 63-65. Received August 29,1989 Revised version received
December
11,1989
49
PROLIFERATION
23.
R&F, E. C., and Houck,
24.
Todaro,
R. G.
Amer.
AND
G. J., Lazar,
J. C. (1969)
J. Cell Physiol. 74,235-244. H. (1965) J. Cell Comp.
G. K., and Green,
Physiol. 66,325. 25.
Levi-Schaffer,
F., and Shalit,
M. (1989)
Znt. Arch. Allergy Appl.
Zmmunol. 90,352-357. 26.
Levi-Schaffer, F., Dayton, E. T., Au&en, K. F., Hein, A., Caulfield, J. P., Gravallese, P. M., Liv F. T., and Stevens, R. L. (1987)
J.Zmmurwl. 27.
Spadari,
139.3431-3441.
S., Sala, F., and Pedrali-Noy,
G. (1982)
Sci. 7,29-32. 28. Roche, W. R. (1985) Amer. J. Pathol. 119,57-64.
Trends Biochem
CELL
RESEARCH
188,4%49
(1990)
Mast Cells Enhance Migration and Proliferation of Fibroblasts into an in Vitro Wound FRANCESCA LEVI-SCHAFFER’ Department
of Pharmacology,
Hebrew
University-Hadassah
Medical
INTRODUCTION
Mast cells (MC) are high-affinity IgE-receptor-bearing cells that stain metachromatically with cationic dyes. Activated MC release the mediators stored in their cytoplasmic granules (histamine, proteoglycans, proteolytic enzymes, chemotactic factors, etc.) and both synreprint
requests
should
0014~4827/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
POB
1172, Jerusakm
MATERIALS
91010,
Israel
AND
METHODS
Materials. RPMI-1640, nonessential amino acids, L-glutamine, streptomycin, and penicillin were purchased from Grand Island Biological Co. (Grand Island, NY); metrizamide, compound 48/80, fetal
be addressed. 42
Inc. reserved.
School,
thesize and release membrane phospholipid metabolites such as prostaglandins, leukotrienes, and platelet-activating factor. MC play a recognized, pivotal role in allergic reactions [ 11, and their increased presence in a number of chronic inflammatory conditions, often accompanied by fibrosis, would suggest that they are involved in these pathologic situations as well [2,3 1. Wound healing is a very complex event which involves interactions of various cell types such as lymphocytes, monocytes, epithelial cells, and fibroblasts. Three main overlapping phases have been identified in tissue response to injury: inflammation, granulation tissue and matrix formation, and remodeling. During granulation tissue formation fibroblasts proliferate and migrate into the wound space [4]. Both in normal reparative as well as in pathological wound healing, the latter leading to hypertrophic scars and keloids, MC have been found in increased numbers [5-71. However, studies carried out to define a role for MC in any of the stages of wound healing that utilized in viva techniques [6-lo] and more recently in vitro organ cultures [ll, 121 have given no definitive results as yet. We have become interested in the relationships between MC and fibroblasts and have demonstrated that in vitro MC survive, change in phenotype, and regenerate activation when cultured with fibroblasts [13-161. The aim of this study was to investigate the possible involvement of MC in wound healing, specifically by studying the relationships between MC and fibroblasts in an in vitro defined wound model. This system consisted of rat peritoneal MC cultured on confluent 3T3 fibroblast monolayers which were wounded by performing a linear cut along the midline and scraping one half of the monolayer [17]. We have examined the extent of fibroblast migration and proliferation into the cell-free area in the presence of resting and activated MC. The results obtained point to a facilitating role of MC in these fibroblast processes.
The effects of mast cells (MC) in an in vitro wound model were studied. The model consisted of rat peritoneal MC cultured on confluent monolayers of 3T3 fibroblasts (MC/3T3). A linear wound was performed by cutting along the midline and scraping one half of the monolayer. After 42 h fibroblasts were counted in the scraped area of the wound. In the MC/3T3 cocultures 27.6 + 2.1 fibroblasts were found compared to 16.6 + 0.9 in the 3T3 cultures. The most significant increase in the number of fibroblasts was obtained upon activation of the MC with anti-IgE antibodies immediately after wound production (39.9 + 2.1). Stimulation with compound 48/80 had a weaker effect (32.7 + 1.6). Incubation of 3T3 wounded monolayers with supernatants of anti-IgE- or compound 48/80-activated MC, or with sonicated MC, induced an increase in fibroblast number similar to that found in unactivated MC/3T3. [3H]Thymidine incorporation followed by autoradiography was performed to assess fibroblast mitosis. The highest number of labeled fibroblasts beyond the wound line was found in immunologically activated MC/3T3 (29.7 + 4.4), followed by compound 48/80-activated MC/3T3 (18.4& 1.5),MC/3T3(15.1+3.6),and3T3(10.6-t2.6). After addition of aphidicolin, which inhibited fibroblast mitosis, MC were still effective in enhancing fibroblast migration. In all the cocultures MC were observed to have migrated alongside fibroblasts. Thus merely the presence of MC adhering to wounded fibroblast monolayers significantly enhanced migration and proliferation of the fibroblasts. A further increase was achieved by immunological activation of the MC. We therefore suggest that MC have a facilitating role in this in vitro wound model. o 1990 Academic press. IDC.
1 To whom
AND ARI KUPIETZKY
MAST
CELLS
ENHANCE
FIBROBLAST
calf serum, toluidine blue, trypan blue, gelatin, histamine dihydrochloride, and aphidicolin from Sigma Chemicals (St. Louis, MO); 35-mm culture dishes from Falcon, Be&on-Dickinson and Co. (Cockeysville, MD); 14C-labeled methyl-S-adenosyl-L-methionine (41 mCi/mmol) and thymidine (methyl-3H) (1 Ci/mmol) from New England Nuclear (Boston, MA); Fujichrome colored photo films for 100 ASA slides from Fuji Film Co. (Tokyo, Japan); rabbit IgG anti-rat IgE and normal rabbit serum from Bethyl Laboratories (Montgomery, TX); Developer D19, Fixer, NTB2 photographic emulsion gel from Eastman Kodak (Rochester, NY); 0.6-mm stainless-steel blades from KDS-Hi-Cutter (Tokyo, Japan).
3T3 mowe firoblast
cultures.
mouse skin fibroblast cell broblast monolayers were mm culture dishes in 2 ml ing 10% heat-inactivated 100 rg/ml streptomycin, amino acids, pH 7.2, EM) when monolayers reached seeded in some of the 3T3
The contact-inhibited Swiss albino line (ATCC, Rockville, MD) was used. Fiprepared by seeding 1 X 10’ 3T3 cells in 35of enriched medium (RMPI-1640 containfetal calf serum (v/v), 100 U/ml penicillin, 2 mM L-glutamine, 0.1 mM nonessential [13]. These cultures were used 7 days later confluency. At this time point, MC were cultures as described below.
Mast cells/3T3 fibroblasts coculture (MC/3T3). Mast cells (MC) were isolated and purified (99% purity) from the peritoneal cavity of 250- to 300-g male Sabra rats (an outbred strain of the Hebrew University), as previously described [13]. MC were seeded at a density of 8 X 10’ cells/2 ml EM on confluent monolayers of 3T3 fibroblasts and incubated at 37°C in a 5% COP humidified incubator. The MC adhere to the fibroblast monolayer in about 10 min after seeding. Cocultures were fed every other day with 2 ml of fresh medium. Wound production in 3T3 and MCf3T3 cultures.
A wound was produced in both MC/3T3 (48-72 h after MC seeding) and 3T3 cultures by a blade cut essentially as described by Stewart et al. [17]. A 0.6-mm steel blade was cut with a high-speed rotary Carborundum stone, to fit the 35-mm diameter of the culture dish. The blade was hand held with the aid of forceps and a downward pressure was applied ensuring the formation of an evenly marked edge. A cotton swab was used to scrape off the fibroblasts from one side of the blade in the designated cell-free area of the dish. A 191 -gauge needle fit to a syringe was used to aspirate cell debris and medium. The wounded monolayer was then washed once with Tyrode’s buffer containing 0.1% gelatin (TG). Three distinct zones were therefore present in the culture dish: (1) a cell-rich area; (2) a marked edge termed “wound line”; (3) a cell-free area termed “wound” (Ref. [17] and Fig. 1A). All the dishes were routinely examined under an inverted phase-contrast microscope (Olympus IM-2).
MIGRATION
AND
43
PROLIFERATION
Fixation and staining of cultures. Fixation and staining of the 3T3 and MC/3T3 cultures was simultaneously performed by incubation with an ethanolic solution of 0.07% toluidine blue, pH 3.5, for 1 min at RT. Stained cultures were rinsed with absolute ethanol and allowed to dry at RT. Autoradiography of cultures. Following incubation of the wounded cultures with the different supematants, 2 ml EM containing 2 &i [3H]thymidine were added. In some experiments in which we wanted to stop fibroblasts mitosis, aphidicolin (5 pg/ml) was added simultaneously with [3H]thymidine. This aphidicolin concentration was found to stop mitosis of fibroblasts without being cytotoxic for either fibroblasts or MC as detected by the absence of cell staining with trypan blue. The medium was removed after 42 h incubation of the cultures in the CO2 incubator and 2 ml of methanol/acetic acid solution (3/l, v/ v) was added for 15 min at RT. The fixative was then removed and the cultures left to dry. Under dark room conditions, a thin layer of NTBz emulsion gel in water (l/l, v/v) was added and the cultures left to dry for 1 h. Dishes were then wrapped in a light-proof packaging and stored for 24 hat 4°C. Developer D-19 and Fixer were prepared according to the manufacturer’s instructions (Kodak). One milliliter of the developer D19 was added to each plate and left for 2 min. The dishes were washed once with 1 ml of distilled water, and 2 ml of fixer was added for 4 min. The fixer was then removed and the dishes were dipped into water for 1 min. The plates were allowed to dry, stained with toluidine blue in ethanol for 8 min, rinsed with ethanol, and allowed to dry again. Quantitation cultures. Fixed
of fibroblnsts
and
mast cells in the wounded
and stained and/or autoradiographed cultures were examined under a light microscope (Olympus BH-2, Tokyo, Japan) equipped with an eyepiece graticle (graduated grid) to allow cell quantification. The grid contained a 10 by 10 matrix of l.O-mm squares. Samples were counted at a X200 or x400 magnification. In the stained dishes the total number of fibroblasts found beyond 250 pm of the wound line was quantitated in at least six different fields. In the autoradiographed cultures labeled and unlabeled fibroblasts were counted in the cell-rich area adjacent to the wound line and in the wound area, in at least four different fields. MC were counted in the same areas of the cultures. Statistical analysis of the data. Data were analyzed by Student’s t test.
RESULTS
Treatment of wounded cultures.
Immediately after wounding, 3T3 and MC/3T3 cultures were incubated for 10 min at 37°C in the COP incubator with 1 ml of one of the following supernatants: (1) TG++ buffer (TG containing 1.8 mM CaClz and 0.9 mM MgCl,); (2) TG++ containing compound 48/80 (3 fig/ml); (3) TG++ containing rabbit anti-rat IgE (l/40, v/v); (4) TG++ containing normal rabbit serum (l/40, v/v); (5) supernatants of freshly isolated MC activated with compound 48/80; (6) supernatants of freshly isolated MC activated with anti-IgE antibodies; (7) suspensions of freshly isolated and sonicated MC. Supernatants 5 and 6 were prepared, respectively, by incubating 8 X lo4 freshly isolated MC with 1 ml TG++ containing either compound 48/80 (3 &ml) or rabbit anti-rat IgE (l/40) for 10 min at 37°C. After pelleting the MC at 12Og for 5 min, supernatants were collected and immediately added to the wounded 3T3 cultures. The suspension of sonicated MC was prepared by disrupting 8 X lo4 MC/ 1 ml TG+t by continuous sonication for 30 s (output 5, 50% duty cycle, Heat Systems Ultrasonics W380). After incubation of the wounded cultures with the different supernatants, supernatants were collected and the cell monolayers were washed with 1 ml TG++. Two milliliters of EM were added and the cultures were placed in the COP incubator for a period of 42 h (unless otherwise stated) after which they were either fixed and stained or processed for autoradiography as described below.
“Wound” Conditions A wound line was produced in cultures of confluent fibroblast monolayers alone (3T3) and of fibroblasts cocultured with rat peritoneal mast cells (MC/3T3) (Fig. 1A). As can be seen in Fig. 2,26 and 42 h after the wound production there was a significant difference in the number of fibroblasts found beyond the wound line between the 3T3 and the MC/3T3 cultures (P < 0.02 at 26 h, P < 0.01 at 42 h). This difference was no more statistically significant at 50 h. Therefore, we adopted the 42-h time point for all the subsequent experiments. Effect of Mast Cells and Mast Cell Mediators on Fibroblast Numbers beyond the Wound Line The effect of MC on the wounded fibroblast monolayers was checked by counting the number of fibro-
44
LEVI-SCHAFFER
AND
KUPIETZKY
FIG. 1. 3T3 fibroblasts cocultured with rat peritoneal mast cells (MC/3T3) immediately after wounding of the cell monolayer and compound 48/80-induced activation of the mast cells. (A) The wound line (Z), marked by the blade on the plate bottom, divides the undisturbed cell-rich area adjacent to the wound (1) from the cell-free area termed “wound” (3). As shown, no fibroblasts or MC are seen in the cell-free area. Cultures were stained with toluidine blue (X95). (B) In the cell-rich area adjacent to the wound, activated MC exhibit their classical activated raspberry-like appearance. Many MC granules (arrows) appear on the wounded 3T3 monolayer (X190).
blasts found beyond the wound line in 3T3 or MC/3T3 cultures subjected to various treatments. After wound production (time 0), MC/3T3 were incubated for 10 min with buffer, buffer containing compound 48/80, or antiIgE antibodies (Fig. 3). Alternatively, 3T3 cultures were incubated with buffer or supernatants obtained from immunologically or compound 48/80-activated MC or sonicated MC (Fig. 4). As can be seen in Fig. 3a, in the 3T3 cultures incubated with buffer alone the number of fibroblasts in the cell-free area was 16.6 f 0.9 (n = 18, mean zk SE). In experiments in which we compared the effects of buffer alone versus buffer containing com-
pound 48/80 or anti-IgE, we obtained similar values for thesethreecontrolgroups,i.e.,17.3+1.5,17.3+_1.1,and 16.9 + 1.3, respectively (mean + SE, n = 3). Therefore, in all the subsequent experiments we used 3T3 cultures incubated with buffer alone as the only control group. Coculture of MC with fibroblasts significantly increased
Z
+ zi
30-
mz $0 a 3 h s?I
20 -
.. .. .. .:a .. .. ..
m0
ii%
1(-J-
k 10
30
50
HOURS AFTER “WOUNDING” FIG. 2. Time course of the appearance of 3T3 fibroblasts beyond the wound line. Fibroblasts were counted along the 250~pm length of the wound line in 3T3 (0 - - 0) and MC/3T3 (0 - 0) cultures at the indicated time points. Closure of both wounds occurred within 6 days. Data are the mean + SE of a representative experiment performed in triplicate.
s o- jll
a
1:: .. .. .. ... b c
d
FIG. 3. Effect of unactivated and activated MC on fibroblast numbers beyond the wound line. The number of fibroblasts found beyond the 250-pm wound line at 42 h after wounding is shown in the following cultures: 3T3 (a), MC/3T3 (b), 3T3/MC activated by compound 48/80 (c), and 3T3/MC activated by anti-IgE (d).
MAST
CELLS
ENHANCE
FIBROBLAST
MIGRATION
AND
45
PROLIFERATION
not differ from that obtained in MC/3T3 cultures incubated with buffer alone and were lower than those obtained in activated MC/3T3 cocultures (see Fig. 3). Morphological Studies of the Wounded MC/3T3 Cultures
A
B
C
D
FIG. 4. Effect of sonicated MC and the supernatants of activated MC on fibroblast numbers beyond the wound line. The number of fibrohlasts found beyond the 250~pm wound line at 42 h after wounding is shown in 3T3 cultures incubated with: buffer (A); suspension of sonicated freshly isolated MC (B); supernatant of compound 48/80-activated MC (C); supernatant of anti-IgE-activated MC (D).
the number of fibroblasts beyond the wound line (27.6 + 2.1, mean + SE, n = 9, P < 0.01) (Fig. 3b). When cocultured MC were stimulated to release their mediators by addition of an optimal concentration of compound 48/ 80 (Fig. 3c) similar values were obtained (32.7 f 1.5, mean f SE, n = 10, P < 0.1). Immunological stimulation of the MC/3T3 cultures with anti-IgE antibodies resulted in the highest number of fibroblasts counted over the wound line, i.e., 39.9 f 2.1 (n = 8, mean + SE) (Fig. 3d). This value was significantly higher than both that obtained in unstimulated MC/3T3 (P < 0.01) and that of compound 48/80-activated MC/3T3 (P < 0.05). The effect of rabbit anti-rat IgE antibodies on MC/3T3 was not due to the presence of rabbit serum in the supernatant. In fact, when wounded MC/3T3 were incubated with normal rabbit serum the number of fibroblasts found over the wound line was comparable to that of cultures incubated with buffer alone (27.0 f 2.2 versus 27.6 f 2.1, mean + SE, n = 2). In Fig. 4 the results obtained by incubating 3T3 cultures with different MC supernatants are depicted. Addition of sonicated MC to the wounded 3T3 monolayer (Fig. 4B) caused a slight though significant increase in the fibroblast numbers in comparison to 3T3 cultures incubated with buffer alone (Fig. 4A) (23.8 t- 2.7, versus 16.6 _+0.9, mean f SE, n = 11, P < 0.05). The supernatants of both compound 48/80 (Fig. 4C) and anti-IgE (Fig. 4D)-activated MC caused also an increase in the fibroblast numbers compared to the buffer-incubated 3T3 cultures (26.6 + 1.8 and 29.1 + 2.4, P < 0.01 and P < 0.001, n = 10, mean * SE). However, these values did
Wounded MC/3T3 cultures were stained with acidic toluidine blue and examined under the light microscope 10 min and 42 h after the different incubation procedures. As expected, 10 min after activating the MC/3T3 cultures, MC lost their resting appearance and acquired the classical activated raspberry-like appearance. Many metachromatic MC granules appeared on the wounded monolayer (Fig. 1B). Fibroblast monolayers appeared stretched and “empty spaces” were formed between the fibroblasts. At this time point, no fibroblast and no MC were detected over the wound line (Fig. 1A). Forty-two hours later as expected, the activated MC reassumed their resting appearance, although they were less granulated than in control-unactivated cocultures [ 161. Fibroblast monolayers appeared as before MC activation. At this time point, in all the cultures, fibroblasts were found over the wound line (Figs. 5A-5D). In stimulated MC/3T3 cultures (Figs. 5C and 5D), fibroblasts migrated and proliferated in higher numbers (see also Figs. 3 and 4) than in 3T3 (Fig. 5A) and MC/3T3 (Fig. 5B) control cultures. Both in unstimulated (Fig. 5B) and stimulated (Figs. 5C and 5D) cocultures, MC were detected over the wound line, morphologically associated with fibroblasts. MC numbers were quantitated and compared in unstimulated, compound 48/80-activated and in anti-IgEactivated MC/3T3 cultures. No difference between the numbers of MC adjacent to the wound line and in the wound area was found in these three different cultures. However, the number of MC found in the wound was always much lower than that found in the area adjacent to the wound (2.5 + 0.3/500 PM in the wound line compared to 10.9 + 0.9/500 PM adjacent to the wound line, n = 4 + SE). Determination of Fibroblast Mitosis 3T3 and MC/3T3 Cultures
in the Wounded
After the lo-min incubation procedures, 3T3 and MC/ 3T3 wounded cultures were incubated for 42 h with EM containing [3H]thymidine and were autoradiographed to assess the extent of fibroblast mitosis. Fibroblasts, both near the wound and far from the wound, were heavily labeled under all the culture conditions. In Fig. 6 the number of labeled fibroblast nuclei and the total number of fibroblasts (labeled plus unlabeled) found beyond the wound line are compared in the differently treated cultures. The highest value of labeled fibroblasts was found in the immunologically activated MC/3T3 (29.7 + 4.4, mean f SE, n = 3 (Fig. 6d)), fol-
46
LEVI-SCHAFFER
AND
KUPIETZKY
FIG. 5. After 42 h of incubation, numerous fibroblasts are foundbeyond the wound line. 3T3 fibroblasts alone (A), MC/3T3 (B), compound 48/80-activated MC/3T3 (C), and anti-IgE activated MC/3T3 (D). Cultures were examined after fixation and toluidine blue staining. The maximum number of fibroblasts was found in the immunologically activated cocultured MC/3T3 (D). In all cocultures, MC were detected beyond the wound line (arrows) (X100).
lowed by compound 48/80-activated cultures (18.4 5 1.5, mean + SE, n = 3 (Fig. 6c)), buffer-incubated MC/3T3 (15.1+ 3.6, mean -+_SE, n = 3 (Fig. 6b)) and buffer-incu-
bated 3T3 (10.6 f 2.6, mean k SE, n = 3 (Fig. 6a)). The total number of fibroblasts over the wound line followed the same pattern of the labeled fibroblasts, i.e., the
MAST
CELLS
ENHANCE
FIBROBLAST
MIGRATION
AND
47
PROLIFERATION
TABLE
1
Effect of Aphidicolin on Fibroblast Numbers beyond the Wound Line Culture
Fibroblast 12.3 26.0 3.6 12.9
3T3 MC/3T3 3T3 + aphidicolinb MC/3T3 + aphidicolin Note. Fibroblasts were counted bated with anti-IgE in the absence n The data (n = 5, mean + SE) found beyond the wound line (per ’ Aphidicolin concentration was
FIG. 6. Number of labeled fibroblasts (la) and total fibroblasts (labeled and unlabeled (0) found beyond the wound line (250 pm) in: 3T3 (a), MC/3T3 (b), compound 48/Wactivated MC/3T3 (c), and anti-IgE-activated MC/3T3 (d). [3H]Thymidine was added immediately after wounding and autoradiography was performed 42 h later.
higher the number of fibroblasts in mitosis, the higher the number of fibroblasts found in the cell-free area. When the number of labeled fibroblasts in the cellrich area was evaluated, results similar to those found for the fibroblasts over the wound line were obtained (data not shown). No MC were found to have incorporated [3H]thymidine indicating that under our tissue culture conditions these MC do not undergo mitosis. Also in autoradiographed cultures, MC were found beyond the wound line associated with fibroblasts (Fig. 7).
FIG. 7. Fibroblasts and mast cells beyond the wound line. The culture (MC/3T3) was treated as described in the legend to Fig. 6b. After autoradiography MC/3T3 were stained with toluidine blue to evince the MC (arrows) that were not labeled by [3H]thymidine (X460). MC apparently have migrated alongside with the fibroblasts to which they had originally adhered.
numbeP k 1.4 k 2.3 f 0.9 z!z 1.5
in 3T3 and MC/3T3 cultures incuand presence of aphidicolin. express the number of fibroblasts 250 pm wound line) at 42 h. 5 pg/ml.
To determine the contribution of fibroblast division in the wound closure, aphidicolin, an antimitotic compound, was added together with [3H]thymidine to 3T3 and MC/3T3 after their incubation with buffer or antiIgE antibodies. Forty-two hours later autoradiography was performed. At the drug concentration used complete inhibition of [3H]thymidine incorporation by 3T3 fibroblasts was achieved. As shown in Table 1, also under these conditions fibroblasts were found over the wound line. In the immunologically activated MC/3T3 cultures, the number of fibroblasts that migrated in the cell-free area in the absence of aphidicolin was nearly double that inits presence (26.0 + 2.3 versus 12.9 f 1.5, mean + SE, n = 5). In 3T3 cultures alone, 12.3 + 1.4 and 3.6 ? 0.9 (mean + SE, n = 5) fibroblasts were found over the wound line in the absence and in the presence of the drug, respectively. DISCUSSION There has been increasing interest in the relationships between mast cells (MC) and fibroblasts since the observation that MC are found in increased numbers in a series of fibrotic conditions such as wound healing, idiopathic pulmonary fibrosis [18], bleomycin-induced fibrosis [ 191, scleroderma [ 201, etc. The present study was undertaken to investigate the effects of the presence and activation of rat peritoneal MC on fibroblasts in an in vitro wound model consisting of wounded monolayers of confluent 3T3 fibroblasts. We have found that MC influence this in vitro wound healing process by increasing fibroblast migration and proliferation. We have previously shown that rat peritoneal MC cocultured with 3T3 fibroblasts rapidly adhere to the monolayer and keep their viability, unique phenotype, and functional activity for over a month [ 131. Moreover, when induced to degranulate, the cocultured MC released their mediators and, in a period of several days regenerated morphologically, functionally, and bio-
48
LEVI-SCHAFFER
chemically [16]. Likewise, the properties of 3T3 fibroblast monolayers are well defined [21,22]. If a wound is made by scratching the cell monolayer, cells migrate into the denuded area, synthesize DNA, and divide, and the wound is thereafter repaired [21,23,24]. Therefore this MC/3T3 culture represented a unique defined in vitro system in which to look for a role of MC in the wound-healing fibrotic stage. We have found that the mere presence of MC on the wounded 3T3 monolayer enhances significantly the migration and proliferation of fibroblasts into the cell-free area (Fig. 2). When the cocultured MC were activated to degranulate by the basic polyamine compound 48/80 these processes were enhanced. Even more impressive results were obtained by stimulating the MC by anti-IgE antibodies (Figs. 3 and 6). Less prominent effects on fibroblast migration and proliferation were observed when the wounded monolayers were incubated with sonicated MC or supernatants from activated MC instead of intact/alive MC (Fig. 4). Sonicated MC suspension contains all the MC preformed mediators such as histamine, proteoglycans, proteolytic enzymes, chemotactic factors, etc., while supernatants of compound 48/80 and anti-IgE-activated MC contain both the preformed mediators and the newly formed mediator, PGD, [25]. It would seem therefore that not only a MC mediator(s) is important in obtaining the enhanced fibroblast migration/proliferation into the wound, but close proximity between MC and fibroblasts can further enhance this process. We have previously shown that MC-fibroblast interactions result in a facilitation of the secretion/degranulation processes of both the rat peritoneal MC and the mouse bone-marrow-derived MC [25, 261. In these studies, including the present one, there is no mechanistic explanation yet available for this obligatory proximity between MC and fibroblasts. Addition of aphidicolin, a specific inhibitor of nuclear DNA synthesis with no detectable effect upon RNA or protein synthesis [ 271, to the wounded 3T3 and MC/3T3 cultures caused a total blockade of fibroblast mitosis. However, fibroblasts still actively migrated beyond the wound line in both untreated and anti-IgE-treated 3T3 and MC/3T3 cultures (Table 1). In these series of experiments migration of fibroblasts was much more prominent in the presence of MC then in 3T3 fibroblast cultures alone, indicating that MC also positively influence this fibroblast property. An interesting observation that might have in uiuo relevance is that in the wounded MC/3T3 cultures, MC were always found over the wound line, apparently having migrated alongside with the fibroblasts (Fig. 7). While the enhancement effect MC had on fibroblast migration in our in vitro wound system is a novel described finding, the mitogenic effect of activation products of MC is a well-described phenomenon at least in. uiuo. For example, Norby and his colleagues have shown
AND
KUPIETZKY
that mesenchymal cell proliferation is caused by a product secreted by MC when stimulated by compound 48/ 80; they concluded that MC histamine is one of several possible important MC-related mitogens [ 111. Roche [28], on the other hand, suggested that heparin contained in the MC granules is a factor that can induce fibroblast proliferation and, being angiogenic, could help in the formation of granulation tissue. In summary, the results presented in this study would suggest that at least one of the possible roles of MC in wound healing might be that of enhancing fibroblast migration and proliferation into the wound. Studies aimed at defining which MC mediator(s) is responsible for both these effects are currently being carried out in our laboratory. We thank Professor S. Shoshan (Department of Oral Biology) for many helpful discussions, Professor J. Kapitulnik (Department of Pharmacology) for the critical reading of the manuscript, and Mrs. Ana Fihach for her skillful secretarial assistance. This work is a partial fullfilment by Mr. Ari Kupietzky of the requirement for the M.Sc. degree in the Hebrew University-Hadassah Medical School. Mr. Ari Kupietzky was supported by a fellowship from the Foulkes Foundation, London. This project was partially supported by a donation to F.L-.S. from the Anticoli Family in memory of their parents.
REFERENCES 1.
Galli, S. J., and Lichtenstein, L. M. andBasophils in Allergy, Principles Reed, C. F., Ellis, E. F., Adkinson, Eds.), 3rd ed., pp. 106-134, Mosby,
2. Bienenstock, Dolovich,
J., Tomioka, J., and Denburg,
(1988) Biology of Mast Cells and Practice (Middleton, E., N. F., and Yunginger, J. W., St. Louis.
M., Stead, R., Ernst, P., Gauldie, J. (1987) Amer. Rev. Respir. Dis.
J.,
135,
55-58. 3. Claman, H. N. (1985) Zmmunol. Today 6,192-195. 4. Clarck, R. A. (1985) J. Amer. Acad. Dermatol. 13,701-725. 5. K&her, C. W., Bunce, H., III, and Shetlar, M. R. (1978) J. Znvest. Dermatol. 70,355-357. 6. Matsuda, H., and Kitamura, Y. (1981) Exp. Hematol. 9,38-43. 7. Fitzpatrick, D. W., and Fisher, H. (1982) Surgery 91,430-434. 8. Franzen, L., and Norby, K. (1983) Acta Pathol. Microbial. Zmmunol.Scand.Sect.A 91,281-289. 9. Persinger, M. A., Lepage, P., Simard, J. P., and Parker, G. H. (1983) Brit. J. Dermatol. 108,179-187. 10. 11. 12.
Schittek, A., Demetriou, A. A., Padawar, enson, S. M. (1984) Agents and Actions
J., Seifter,
E., and Lev-
15,172-176. Norby, K. (1985) Agents Actions 16(3/4), 287-290. Druvefors, P., and Norby, K. (1988) Virchws Arch. I?: Cell Pathol. 55,187-192.
13.
Levi-Schaffer, F., Austen, K. F., Caulfield, J. P., Hein, A., Bloes, W. F., and Stevens, R. L. (1985) J. Zmmunol. 135,3454-3462.
14.
Levi-Schaffer, vallese, P. M.,
F., Austen, K. F., Caulfield, and Stevens, R. L. (1987)
J. P., Hein,
A., Gra-
J. Zmmunol. 139,494-
500. 15. 16.
Levi-Schaffer, R. L. (1986) Levi-Schaffer,
40.
F., Au&en,
K. F., Gravallese,
P. M.,
and Stevens,
Proc. Natl. Acad. Sci. USA 83.6485-6486. F., and Riesel, N. (1989) Cell Zmmurwl. 119,30-
MAST
17.
Stewart,
R. J., Duley,
Exp. Path& 18. 19.
CELLS
ENHANCE
J. A., and Allardyce,
FIBROBLAST
R. A. (1979)
&it.
MIGRATION
J.
60,582-588.
Kawanami, O., Ferrans, V. Y., Fulmer, J. D., and Crystal, (1979) Lab. Invest. 40,717-734. Goto, T., Befus, D., Low, R., and Bienenstock, J. (1984)
Rev. Respir. Dis. 130,797-802. 20. Hawkins, R. A., Claman, H. N., Clark, R. A. F., and Steigerwald, J. C. (1985) Ann. Intern. Med. 102,182-186. 21. Lipton, A., Klinger, I., Paul, D., and Holley, R. W. (1971) Proc. Natl. Acad. Sci. USA 68,2799-2801. 22. Dunn, G. A., and Ireland, G. W. (1984) Nature (London) 312, 63-65. Received August 29,1989 Revised version received
December
11,1989
49
PROLIFERATION
23.
R&F, E. C., and Houck,
24.
Todaro,
R. G.
Amer.
AND
G. J., Lazar,
J. C. (1969)
J. Cell Physiol. 74,235-244. H. (1965) J. Cell Comp.
G. K., and Green,
Physiol. 66,325. 25.
Levi-Schaffer,
F., and Shalit,
M. (1989)
Znt. Arch. Allergy Appl.
Zmmunol. 90,352-357. 26.
Levi-Schaffer, F., Dayton, E. T., Au&en, K. F., Hein, A., Caulfield, J. P., Gravallese, P. M., Liv F. T., and Stevens, R. L. (1987)
J.Zmmurwl. 27.
Spadari,
139.3431-3441.
S., Sala, F., and Pedrali-Noy,
G. (1982)
Sci. 7,29-32. 28. Roche, W. R. (1985) Amer. J. Pathol. 119,57-64.
Trends Biochem