Mononuclear phagocytes can interact with mesenchymal cells and extracellular matrix components that are crucial for connective tissue rearrangement. We asked whether blood monocytes can alter matrix remodeling mediated by human lung fibroblasts cultured in a three-dimensional collagen gel. Blood monocytes from healthy donors (>95% pure) were cast into type I collagen gels that contained lung fibroblasts. Monocytes in coculture inhibited the fibroblast-mediated gel contractility in a time- and concentration-dependent manner. The concentration of PGE2, a well-known inhibitor of gel contraction, was higher (P < 0.01) in media from coculture; this media attenuated fibroblast gel contraction, whereas conditioned media from either cell type cultured alone did not. Three-dimensional cultured monocytes responded to conditioned media from cocultures by producing interleukin-1β and tumor necrosis factor-α, whereas fibroblasts increased synthesis of PGE2. Antibodies to interleukin-1β and tumor necrosis factor-α blocked the monocyte inhibitory effect and reduced the amount of PGE2 produced. The ability of monocytes to block the fibroblast contraction of matrix may be an important mechanism in regulating tissue remodeling.
- prostaglandin E2
- tumor necrosis factor-α
blood monocytes belong to the mononuclear phagocytic system. They originate from a bone marrow stem cell and circulate for 2–3 days in the bloodstream before they migrate into different tissues (33). In the lung, it is believed that blood monocytes become interstitial monocytes/macrophages (30, 35). Monocytes have a variety of functions in the inflammatory response. For example, they can act as antigen-presenting cells and interact with T cells and initiate an immune response. In addition, they have phagocytic functions, and they are capable of secreting a variety of inflammatory mediators such as the proinflammatory cytokines interleukin (IL)-1β and tumor necrosis factor (TNF)-α (3). In addition, these cells can interact with mesenchymal cells and the extracellular matrix components that are crucial for connective tissue rearrangement (5, 9, 25,31).
The chronic inflammatory response is characterized by the derangement of repair processes and the development of scar tissue. This process is complex and consists of recruitment and proliferation of fibroblasts, production of extracellular matrix, and tissue remodeling (27). Part of this process can involve tissue contraction, which can alter tissue function. For example, in asthma and in chronic bronchitis, a subepithelial or peribronchiolar fibrosis that may lead to fixed airway obstruction and chronic lung function impairment can be observed (16, 26, 28). Similar processes take place in inflammatory or fibrotic disorders, including atherosclerosis, cirrhosis of the liver, and rheumatoid arthritis, in many tissues.
Interactions between monocytes and fibroblasts are thought to be of importance in the repair processes that follow inflammation (6,7, 9, 10, 17). In the present study, therefore, we tested the hypothesis that blood monocytes may affect fibroblast contractility, which is one aspect of tissue remodeling. To accomplish this, we utilized the three-dimensional gel contraction assay in which human lung fibroblasts are cultured in a matrix of native collagen fibers. With this system we were able to demonstrate that monocytes can block the ability of fibroblasts to contract their surrounding matrix. In addition, we showed that this effect is mediated by the IL-1β and TNF-α produced by the monocytes that drive the paracrine/autocrine production of fibroblast-derived PGE2.
Type I collagen (rat tail tendon collagen; RTTC) was extracted from rat tail tendons by a previously described method (11, 22). Briefly, tendons were excised from rat tails, and the tendon sheath and other connective tissues were carefully removed. After repeated washes with Tris-buffered saline (0.9% NaCl and 10 mM Tris, pH 7.5) and 95% ethanol, the collagen was extracted in 6 mM acetic acid. Protein concentration was determined by weighing a lyophilized aliquot from each lot of collagen solution. SDS-PAGE routinely demonstrated no detectable proteins other than type I collagen. The RTTC was stored at 4°C until used. PGE2 and indomethacin were purchased from Sigma (St Louis, MO). Neutrophil elastase purified from human sputum was purchased from Elastin Products (Owensville, MO). TNF-α, IL-1β, and transforming growth factor (TGF)-β1 were purchased from R&D Systems (Minneapolis, MN). Polyclonal antibodies to IL-1β and TNF-α were purchased from R&D Systems. Cell culture media, except for FCS, were obtained from GIBCO BRL (Life Technologies, Grand Island, NY). FCS was purchased from Biofluids (Rockville, MD).
PGE2 was dissolved in PBS to a stock solution of 50 mg/ml. Indomethacin was dissolved in ethanol to a stock solution of 10−2 M. TGF-β1 was dissolved to a stock solution of 1 μg/ml in 4 mM HCl and 0.1% BSA. IL-1β and TNF-α were dissolved to a stock solution of 1 μg/ml in NaCl containing 0.1% BSA. Antibodies to IL-1β and TNF-α were dissolved in PBS to a stock solution of 0.1 mg/ml. Neutrophil elastase was dissolved in PBS to a stock solution of 1 mg/ml.
Human fetal lung fibroblasts were purchased from the American Type Culture Collection (Manassas, VA). The cells were cultured in 100-mm tissue culture dishes (Falcon, Becton Dickinson Labware, Lincoln Park, NJ) in DMEM supplemented with 10% FCS, 50 U/ml of penicillin G sodium, 50 μg/ml of streptomycin sulfate (penicillin-streptomycin, GIBCO BRL, Life Technologies), and 1 μg/ml of amphotericin B (Parma-Tek, Huntington, NY). The fibroblasts were passaged weekly. Confluent fibroblasts were trypsinized (trypsin-EDTA; 0.05% trypsin and 0.53 mM EDTA-4Na), resuspended in DMEM without serum, and used for collagen gel culture.
Blood monocytes were isolated by countercurrent centrifugal elutriation of mononuclear leukocyte-rich fractions of blood cells from healthy blood donors undergoing leukophoresis (34). Cell suspensions were >95% monocytes by the criteria of cell morphology on Wright-stained cytosmears.
Preparation of collagen gels.
Collagen gels were prepared by mixing the appropriate amounts of RTTC, distilled water, 4× concentrated DMEM, and cell suspensions so that the final mixture resulted in 0.75 mg/ml of collagen, 3 × 105 fibroblasts/ml, and a physiological ionic strength of 1× DMEM. Fibroblasts were added after all the other ingredients had been mixed. In the coculture experiments, monocytes were added to the fibroblast-collagen mixture before gelation. In general, except in the cell concentration experiments, the monocyte concentration in the mixture was 3 × 105 cells/ml. A 550-μl aliquot of the cell-collagen mixture was then cast into each well of 24-well tissue culture plates with a 2-cm2 growth area (Falcon). Gelation occurred within 15 min at room temperature.
Gel contraction assay.
The ability of cocultured blood monocytes to affect fibroblast-mediated collagen gel contraction was determined by a modification of the slow contraction assay described by Bell et al. (1). Briefly, fibroblasts and monocytes were cast into collagen gels as described inPreparation of collagen gels. After gelation, the gels were released from the surface of the culture well and transferred to 60-mm tissue culture dishes (Falcon) containing 5 ml of serum-free DMEM. Except in the time-course experiments, the floating gels were then incubated at 37°C under 5% CO2 for 2 days. To prevent attachment of the gels to the bottom of the culture dishes, the dishes underwent continuous rocking on a rocker platform (Bellco Biotechnology, Vineland, NJ) during the first 24 h. The areas of the floating gels were measured with an image analyzer (Optomax V, Optomax, Burlington, MA). The images were captured with a video camera with a zoom lens mounted ∼18 inches above a lighted stage. Sixty-millimeter dishes rested on this stage, and the gels were observed directly. No special mounting was required for stability. The zoom lens of the video camera was adjusted so that the capture area of the Optomax V image analyzer was approximately two-thirds filled by the uncontracted gel. The area was then captured and converted to pixels. Each gel was localized within the capture area manually. With this method, each gel could be monitored for area preserving sterility. Thickness is not measured by this method, and contraction in the vertical dimension is not assessed. The media in which the gels were floated were used as conditioned media as described inConditioned media. At appropriate time points, FCS, TGF-β1, neutrophil elastase, PGE2, IL-1β, and TNF-α were added to the media in which the gels floated. Indomethacin was added to the gel and to the medium.
To investigate whether the interactions between fibroblasts and monocytes are due to a soluble component or to required cell-to-cell contact, conditioned media were harvested and used in the experiments. Fibroblasts and monocytes (3 × 105 of each cell type/ml) were cast in a three-dimensional collagen gel. After gelation, the gels were released into 5 ml of serum-free DMEM and cultured for 2 days. After 2 days of culture, the media were harvested and used as conditioned media. Conditioned media were also made by culturing either cell type alone.
Determination of PGE2.
Soluble PGE2 in the culture media from collagen gel cultures was assayed by a commercially available radioimmunoassay (PerSeptive Biosystems, Framingham, MA) (24).
Measurement of IL-1β and TNF-α.
Concentrations of IL-1β and TNF-α in collagen gel culture supernatants were measured by a commercially available ELISA (R&D Systems, Minneapolis, MN).
Determination of the cell source of PGE2, IL-1β, and TNF-α in the coculture system.
To determine which cell type in the coculture system was responsible for the production of the mediators PGE2, IL-1β, and TNF-α, conditioned media were collected from a 2-day coculture of fibroblasts and monocytes in a three-dimensional collagen gel. Freshly made gels containing either no cells, fibroblasts, monocytes, or both fibroblasts and monocytes were prepared and released into a 1:4 dilution of the conditioned medium in serum-free DMEM. After 2 days of culture, this medium was collected, and the concentrations of PGE2, IL-1β, and TNF-α were assayed as described inDetermination of PGF2 and Measurement of IL-1β and TNF-α.
Blocking of IL-1β and TNF-α by antibodies.
Blocking of IL-1β and TNF-α by antibodies was done to evaluate if this could abolish the inhibitory effect of monocytes on fibroblast-mediated collagen gel contraction. Antibodies to IL-1β and TNF-α (final concentration 0.2 μg/ml) were added to both the collagen gels and the surrounding medium in which the gels were floated. The gels were then cultured for 2 days, the gel area was measured as described in Gel contraction assay, and PGE2 was assayed in the medium. PGE2 was also added exogenously (50 ng/ml) to assess whether it could restore the inhibitory effect of monocytes.
Data are presented as means ± SE of three replicate gels for each condition unless otherwise stated. Because the batch of RTTC, the number of cell passages, and the culture conditions can affect the gel contraction, the data shown in Figs. 1-7 are taken from a representative experiment, each of which was repeated on multiple occasions. Groups of data were evaluated by ANOVA. Data from single time points were compared by Student's t-test. Values ofP < 0.05 were considered significant. A Bonferroni correction was performed to adjust for multiple comparisons of gel areas.
Effect of blood monocyte coculture on fibroblast contractility in a three-dimensional collagen gel.
Blood monocytes, when added to fibroblast-containing collagen gels, attenuated the gel contraction (P < 0.001; Table1). Monocytes alone cultured in a three-dimensional collagen gel did not contract the gel. The inhibition of contraction was time dependent, but coculture of fibroblasts and monocytes resulted in less contractility than fibroblasts alone at all time points for the three days tested (Fig.1). The inhibition of fibroblast contractility was dependent on the number of monocytes added. Adding 1 × 105 monocytes/ml to a gel did not result in additional attenuation of fibroblast-mediated gel contraction (Fig.2).
A number of mediators, including serum, have been shown to augment fibroblast-mediated contraction of collagen gels (12, 14,23). Therefore, we assessed whether addition of blood monocytes had the ability to reverse this augmentation. As shown in Fig.3, monocytes were able to attenuate the increased contractility induced by TGF-β1, neutrophil elastase, and FCS (P < 0.001, P < 0.001, andP < 0.01, respectively). In all cases, when fibroblasts and monocytes were incubated with a procontractile stimulus, the gel area was always intermediate between the area of the gels when the stimulus was added to fibroblasts alone and the area of the cocultured gels.
Effect of conditioned medium on fibroblast-mediated collagen gel contraction.
To investigate whether the inhibition of contractility was due to a soluble mediator or to required cell-cell contact, conditioned medium was tested for its ability to affect fibroblast-mediated collagen gel contraction. Conditioned media were harvested after a 2-day culture from three-dimensional collagen gels containing either fibroblasts, monocytes, or both cells in coculture. Conditioned medium from a coculture of fibroblasts and monocytes attenuated fibroblast contractility in a concentration-dependent manner (P < 0.001; Fig. 4). In contrast, no attenuation was observed when conditioned medium from either cell alone was assayed for fibroblast contractility.
Effect of cyclooxygenase inhibition and exogenous PGE2on fibroblast contractility and PGE2 production in the coculture system.
PGE2 is a potent inhibitor of fibroblast-mediated collagen gel contraction, and its synthesis can be blocked by inhibitors of cyclooxygenase. Thus indomethacin, when added to fibroblasts in a three-dimensional collagen gel, augmented gel contraction (P < 0.01; Table 2). Indomethacin also had the capacity, at least partially yet significantly (P < 0.01), to reverse the inhibition induced by blood monocytes. Addition of exogenous PGE2, as expected, had an inhibitory effect (P < 0.01) that, interestingly, was similar in magnitude to the inhibitory effect observed when fibroblasts and monocytes were cocultured.
Consistent with the findings in the gel contraction assay, the concentration of PGE2 was dramatically increased in the supernatant media from cocultured fibroblasts and monocytes (P < 0.01; Table 2). Indomethacin had the capacity to block PGE2 synthesis from fibroblasts (P < 0.01). More importantly, it was able to almost totally inhibit (P < 0.01) the increased PGE2 production from the coculture system.
Role of IL-1β and TNF-α.
Monocytes have the capacity to produce IL-1β and TNF-α, and these cytokines can stimulate fibroblasts to produce PGE2(10). To clarify if this is a possible mechanism behind the inhibitory effect, different concentrations of IL-1β and TNF-α were added to a gel culture system containing only fibroblasts. As shown in Fig. 5, there was a concentration-dependent inhibition of fibroblast-mediated collagen gel contraction when IL-1β and TNF-α were added to the culture system. Moreover, there was a similar relationship between the concentration of the cytokines and the induced production of PGE2 by the fibroblasts.
In agreement with the ability of fibroblasts to respond with PGE2 production and inhibition of collagen gel contraction in response to increasing concentrations of the cytokines IL-1β and TNF-α, the levels of these cytokines increased in the coculture system (P < 0.01 for both; Fig.6). The addition of indomethacin to the culture system did not result in a lower concentration of the cytokines.
Cell source of PGE2, IL-1β, and TNF-α in the three-dimensional collagen gel coculture.
To sort out which cell type is responsible for the increased production of PGE2, IL-1β, and TNF-α, we assessed whether either cell type, cultured in a three-dimensional collagen gel, was able to stimulate production in response to conditioned media from a collagen gel coculture. As shown in Table 3, fibroblasts in a three-dimensional collagen gel responded to conditioned media with increasing PGE2 production, but monocytes did not. When fibroblasts and monocytes in a three-dimensional collagen coculture system were stimulated by conditioned media, the production was further increased. Moreover, conditioned media were able to stimulate monocytes to secrete both IL-1β and TNF-α, whereas fibroblasts did not increase secretion of IL-1β and TNF-α in response to conditioned media. The highest concentrations of these cytokines were detected, however, when both fibroblasts and monocytes cultured in a three-dimensional collagen gel were stimulated by the conditioned media (Table 3).
Blocking the inhibitory effect of monocytes by adding antibodies to IL-1β and TNF-α.
Production of IL-1β and TNF-α by monocytes can stimulate fibroblasts to secrete PGE2 and thereby inhibit the contractile activity in a three-dimensional collagen gel. To assess the importance of these cytokines, we blocked them by adding antibodies to the culture system. As shown in Fig. 7, antibodies to IL-1β and TNF-α each partially yet significantly blocked the inhibitory effect of monocytes on fibroblast-mediated collagen gel contraction (P < 0.001 for both). Added together, the antibodies completely blocked the monocyte-induced inhibition (P < 0.001). Irrelevant antibodies (3 μg/ml) slightly augmented contraction of control collagen gels (49.0 ± 0.6 vs. 46.6 ± 0.6%) and had a similar effect on fibroblasts whose contraction had been inhibited by coculture with monocytes (59.2 ± 1.9 vs. 55.2 ± 0.7%). Specific antibodies also blocked monocyte-induced PGE2 production. Compared with the control condition [1,257.5 ± 134.5 (SE) pg/ml], anti-IL-1β blocked PGE2 production 74% (326.5 ± 14.5 pg/ml; P < 0.05), and anti-TNF-α blocked the production by 74% (328.5 ± 71.5 pg/ml;P < 0.05), and the two antibodies added together blocked production 84% (202.5 ± 21.5 pg/ml; P < 0.05). The inhibitory effect of the antibodies appeared to be due to their effect on PGE2 production because the addition of exogenous PGE2 was able to restore (P < 0.01) the inhibitory effect of the monocytes.
In the present study, we demonstrated that human blood monocytes have the capacity to interact with fibroblasts in an in vitro matrix model. Specifically, blood monocytes were able to attenuate fibroblast-mediated contraction of a three-dimensional collagen gel.
The inflammatory response is followed by repair processes to restore a normal tissue architecture. This repair involves a number of events, including recruitment and proliferation of mesenchymal cells, synthesis of extracellular matrix macromolecules, and remodeling. In most cases, the repair process restores tissue without compromise of physiological function. The repair process may, however, result in altered tissue morphology and development of scar tissue (27). One important feature in tissue remodeling is tissue contraction. In the lung, this may lead to shrinkage of the tissue, as frequently seen in idiopathic pulmonary fibrosis. However, tissue contraction also occurs in obstructive lung disorders such as asthma and chronic obstructive pulmonary disease. In these conditions, a deposition of connective tissue and an accumulation of fibroblasts is observed in the subepithelial or peribronchiolar space (15, 26). Combined with narrowing of the airways, this may lead to fixed airflow limitation.
Fibroblasts are a key cell in remodeling and in the formation of fibrosis. In this respect, fibroblasts can interact with a variety of cells and matrix components in the extracellular space. Monocytes that are recruited to the site of inflammation can interact with fibroblasts, and this interaction seems to be of importance because mononuclear cells appear to be crucial in the regulation of fibroblast function. For instance, it is known that factors released by mononuclear cells can alter fibroblast growth (9, 10, 18), production of collagens and proteoglycans (17), and fibroblast chemotaxis (25). At least in part, this effect is mediated by the ability of the mononuclear phagocytes to stimulate fibroblast prostaglandin production.
When fibroblasts are cultured in a three-dimensional collagen gel and detached from the underlying surface, the fibroblasts contract the gels over several days in serum-free media. A number of mediators have been shown to affect fibroblast-mediated contraction of collagen gels. For example, serum, TGF-β1, platelet-derived growth factor, and fibronectin enhance contraction of collagen gels (12, 14,23), whereas PGE2, heparin, and β2-adrenergic agonists attenuate contraction (8,21). Although this contraction model, originally described by Bell et al. (1), certainly differs in many aspects from in vivo conditions, it has been considered an in vitro model for tissue remodeling (13). This system has been used by many investigators (2, 20, 29) to explore cellular aspects of remodeling with cells from a variety of tissues.
The aim of the present study was to study interactions between fibroblasts and blood monocytes in a three-dimensional collagen gel with respect to tissue contraction. The inhibition of contraction by monocytes was concentration- and time-dependent. Cyclooxygenase inhibition reversed the monocyte-induced attenuation, and exogenously added PGE2 had a similar inhibitory effect in the coculture system. In the coculture system, monocyte production of the proinflammatory cytokines IL-1β and TNF-α stimulated fibroblasts to produce PGE2. In this regard, the monocyte-induced inhibition of contraction could be blocked by antibodies to IL-1β and TNF-α, also inhibiting fibroblast PGE2 synthesis.
When monocytes were added to fibroblast gels containing other stimulants for gel contraction, the monocyte-induced attenuation was not complete. Moreover, the addition of indomethacin to cocultured gels did not result in a contraction similar in magnitude to the one observed when indomethacin was added to gels containing only fibroblasts. This suggests that other factors besides PGE2may be involved in the mechanism behind inhibition of contraction. When monocytes cultured in a three-dimensional collagen gel were stimulated with conditioned media, they responded with increased IL-1β production. However, the concentration of IL-1β increased nearly 10-fold when both fibroblasts and monocytes were stimulated with the conditioned media. This indicates that the IL-1β production in the coculture system is greatly augmented by cell-cell interactions when stimulated with conditioned media.
Monocytes play a pivotal role in modulating the turnover of extracellular matrix by secretion of proteinases and proteinase inhibitors (4, 5, 36). In addition, they secrete IL-1β and TNF-α, which are potent inducers of matrix metalloproteinases in the resident cells of tissues, e.g., fibroblasts (6, 7,19). Thus these mediators may potentially degrade the collagen. In the short-term culture system used in the present study, it is unlikely that such degradation occurred, because the cytokines attenuated fibroblast contractility. However, the impact of long-term coculture in three-dimensional collagen gels requires further study.
Our findings do not define the in vivo relationship between the monocytes and fibroblasts. However, the ability of monocytes to inhibit fibroblast contractile activity may “loosen” a tissue, thereby facilitating the migration of other cells that participate in repair processes. It may also counteract other augmentative factors. For instance, neutrophils are able to augment fibroblast-mediated collagen gel contraction (32) by virtue of secretion of neutrophil elastase. The recruitment of “inhibiting” monocytes may therefore be a mechanism by which the damaging effects of neutrophil elastase and other factors can be limited.
In conclusion, in the present study, we were able to demonstrate that blood monocytes attenuate fibroblast-mediated contraction of collagen gels by stimulating fibroblast PGE2 production. The finding that monocytes can directly interfere with fibroblasts in a three-dimensional collagen gel may be of importance in the tissue repair that follows inflammation.
We thank Lillian Richards for expert secretarial assistance.
C. M. Sköld is supported by grants from the Swedish Council for Work-Life Research, the Swedish Heart-Lung Foundation, the King Oscar II Jubilee Foundation, the Swedish Society of Medicine, the Karolinska Institute, and Astra-Draco.
Address for reprint requests and other correspondence: S. I. Rennard, Pulmonary and Critical Care Medicine Section, 985125, Nebraska Medical Center, Omaha, NE 68198-5125 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society