Proliferation of fibroblasts contributes to the adventitial thickening observed during the development of hypoxia-induced pulmonary hypertension. However, whether all or only specific subpopulations of fibroblasts proliferate during this process is unknown. Because lung, skin, and gingiva contain multiple fibroblast subpopulations, we hypothesized that the pulmonary artery (PA) adventitia of neonatal calves is composed of multiple fibroblast subpopulations and that only selective subpopulations expand under chronic hypoxic conditions. Fibroblast subpopulations were isolated from PA adventitia of control calves using limited dilution cloning techniques. These subpopulations exhibited marked differences in morphology, actin expression, and serum-stimulated growth. Only select fibroblast subpopulations demonstrated the ability to proliferate in response to hypoxia. Fibroblast subpopulations were similarly isolated from calves exposed to hypoxia (14 days). With regard to morphology, actin expression, and serum-stimulated growth of subpopulations, there were no obvious differences in fibroblast subpopulations between the hypoxic and the control calves. However, the number of fibroblast subpopulations with about a twofold increase in hypoxia-induced DNA synthesis was significantly greater in the hypoxic calves (26%) compared with control calves (10%). We conclude that the bovine PA adventitia comprises numerous phenotypically and biochemically distinct fibroblast subpopulations and that select subpopulations expand in response to chronic hypoxia.
- vascular disease
- vascular remodeling
- fibroblast growth
- hypoxia-induced proliferation
hypoxia-inducedpulmonary hypertension complicates the clinical course of many important lung diseases in both children and adults (23, 42,46). The pulmonary hypertension and accompanying structural remodeling is particularly severe in infants, and often the earliest and most dramatic structural changes are found in the adventitial compartment of the vessel wall (31, 46). In animal models, the resident adventitial fibroblasts have been shown to exhibit early and sustained increases in proliferation and matrix protein synthesis under hypoxic conditions, and these changes have been associated with luminal narrowing and a progressive decrease in the ability of the vessel wall to respond to vasodilating stimuli (3, 23, 30, 45,46). Furthermore, in response to different stimuli, adventitial fibroblasts isolated from hypoxia-induced pulmonary hypertensive neonatal calves demonstrate greater growth capabilities than those from control animals (8), which could be the result of a generalized or polyclonal activation of resident adventitial fibroblasts. Alternatively, it could be the result of the presence of increased numbers of specific fibroblast subpopulations with distinct growth characteristics as has been suggested in the setting of lung fibrosis and scleroderma (15, 20, 21, 24, 32). However, it is not known whether the pulmonary artery (PA) adventitia is composed of heterogeneous fibroblast subpopulations with unique functional characteristics and, if so, whether these fibroblast subpopulations contribute selectively to the fibroproliferative vascular changes observed in cases of severe hypoxia-induced pulmonary hypertension.
The existence of phenotypic and functional heterogeneity among the resident fibroblast populations of different organs has been well described (6, 19, 26, 34, 36, 37, 39, 43). Evidence of fibroblast heterogeneity within the same anatomical site or organ system has also been well documented (2, 11, 15, 16, 18, 20,34). Fluorescence-activated cell sorting (FACS) on the basis of cell-surface antigen expression and limited dilution cloning techniques have both been utilized to demonstrate the existence of heterogeneous fibroblast populations within the lung with regard to morphology, synthesis of prostaglandin (PG) E2, synthesis of collagen, and growth capability (11, 24, 34, 35, 48). Importantly, utilization of either technique results in the generation of stable subpopulations or clones that maintain unique and distinct functional characteristics for prolonged periods of time in culture. Accumulating evidence also supports the concept that persistent local pathophysiological stimuli may provide a natural selective drive, as a result of which certain fibroblast clones or subpopulations with distinct proliferative or matrix-producing capabilities will emerge and contribute specifically to the disease process in fibroblast-enriched organ systems (15, 20, 32, 49). However, it is unknown whether fibroblast subpopulations that are uniquely susceptible to activation and proliferation in response to hypoxia exist within the pulmonary circulation or any other organ system. Furthermore, it is not known whether chronic hypoxia would provide an environment such that fibroblasts with unique hypoxia-induced growth advantages would emerge and increase in number and thus contribute in select ways to the adventitial thickening and fibrosis during the development of hypoxic pulmonary hypertension.
We hypothesized that the normal PA adventitia is composed of heterogeneous subpopulations of fibroblasts and that only select subpopulations possess the capability of proliferating under hypoxic conditions in the absence of exogenous mitogens. Furthermore, we hypothesized that chronic in vivo hypoxia would create a growth advantage for these fibroblast subpopulations and would result in a selective increase in their number in the thickened PA adventitia of chronically hypoxic animals. A positive answer to this question would lend support to the idea that certain fibroblast subpopulations are activated in response to a given pathophysiological condition and may thus selectively contribute to specific disease processes. Because no markers have been developed that allow FACS sorting of bovine fibroblasts and because we did not want to “preselect” fibroblast subpopulations based on any one marker of fibroblast phenotype, our approach was to isolate fibroblast subpopulations using limited dilution cloning techniques and then to evaluate these subpopulations for differences in morphology, contractile protein expression, and growth capabilities. In addition, we specifically evaluated the subpopulations for their ability to proliferate in response to hypoxia in the absence of any exogenous mitogens. We also exposed a group of neonatal calves to chronic hypoxia for 14 days to promote the development of severe pulmonary hypertension and adventitial thickening and fibrosis. Adventitial fibroblast subpopulations from these vessels were then generated using the same techniques and also evaluated for changes in morphology, contractile protein expression, and growth in response to hypoxia. Our findings support the idea that the PA adventitia is composed of multiple functionally heterogeneous fibroblast subpopulations. Only specific fibroblast subpopulations exhibit an ability to proliferate under hypoxic conditions. However, these fibroblast subpopulations appear to undergo selective expansion in response to chronic hypoxia and therefore may contribute to the adventitial thickening and fibrosis observed during the development of chronic hypoxia-induced pulmonary hypertension.
MATERIALS AND METHODS
Human platelet-derived growth factor PDGF-BB and basic fibroblast growth factor (bFGF) were purchased from Bachem Fine Chemicals (Torrance, CA) and suspended in Eagle's minimal essential medium (MEM) with 2% fatty acid-free bovine serum albumin (BSA). MEM, trypsin-EDTA 10× suspension, penicillin, streptomycin, amphotericin B, Hanks' balanced salt solution (HBSS) with HEPES (with and without CaCl2), and monoclonal antibody specific for α-smooth muscle (SM) actin (A-2547 from clone 1A4) were from Sigma (St. Louis, MO). Elastase was purchased from Boehringer Mannheim (Indianapolis, IN). Collagenase II and soybean trypsin inhibitor were from Worthington (Lakewood, NJ). Fetal bovine serum (FBS) was purchased from Gemini Bio-Products (Calabasas, CA). [3H]Thymidine was from ICN Biochemicals (Irvine, CA). Rabbit antibodies against bovine aortic SM myosin were kindly provided by Dr. R. S. Adelstein (National Heart, Lung, and Blood Institute, Bethesda, MD) (25). Affinity-purified rabbit antibody against von Willebrand factor was purchased from DAKO. Conditioned media were prepared for cloning experiments by removing medium from rapidly growing culture of bovine fetal fibroblasts and filtering it through a 0.22-μM filter to remove any cell debris.
Isolation and growth of neonatal bovine main PA adventitial fibroblast subpopulations.
We harvested adventitia from the main PA of 15-day-old normoxic and hypoxic neonatal calves. Normoxic calves were born and remained at Fort Collins, CO [1,524 m altitude, barometric pressure (Pb) = 650 Torr]. Hypoxic calves were born at Fort Collins altitude, but after they were 1 day old, they were placed into an altitude chamber at simulated altitude (4,570 m, Pb = 445 Torr), where they remained for 2 wk to induce severe pulmonary hypertension as previously described (45). At postmortem examination, adventitial tissue was isolated, carefully dissected free of blood vessels and fat under a dissecting microscope, and cut into small pieces. The tissue pieces were incubated with Ca2+-free HBSS for 30 min at 37°C and then with HBSS containing Ca2+, elastase, collagenase, albumin, and soybean trypsin inhibitor for 90 min at 37°C on a rotator. The tissue was gently triturated with a sterile Pasteur pipette after each 30 min of incubation. The dispersed cells were passed through a 100-μm nylon cell strainer (Falcon) to remove any undigested tissue pieces, diluted in MEM containing 10% FBS to inactivate the enzymes, and centrifuged at 900 rpm for 10 min. Using a light microscope and hemocytometer, we counted the cells and serially diluted this cell suspension with the media containing 30% fetal conditioned media and 10% FBS. Cells were plated at a density of 0.5 cells · well−1 · 0.2 ml−1 in 96-well plates. These cultures were examined at frequent intervals using a light microscope and maintained at 37°C and 5% CO2 with a biweekly change of media containing 30% fetal conditioned media and 10% FBS. The cells were maintained in 96-well plates for 2 wk. Wells with cells were scored to determine cloning efficiency. Once cells reached confluence in the microtiter wells, they were trypsinized and transferred to 24-well culture dishes in MEM containing 10% FBS. When the cells had grown again to confluence, they were transferred serially into 12-well, 6-well culture dishes and 25-mm, 75-mm culture flasks, respectively. In all experiments, fibroblasts were studied between the third and the 15th passages.
Characterization of fibroblast subpopulations by morphology and immunofluorescence staining for α-SM actin, myosin, and factor VIII.
Because our goal was to obtain pure subpopulations of fibroblasts, we first analyzed individual cell populations for their morphological appearance and then examined expression of SM-specific markers in each isolated cell population. Cells were grown to confluence on Tissue-Tek chamber slides in 10% FBS-MEM, fixed with cold methanol for 10–15 min at 4°C, and processed for indirect immunostaining as follows. For double immunofluorescence staining of α-SM actin and SM myosin, fixed cells were incubated with a cocktail of monoclonal α-SM actin and polyclonal anti-SM myosin antibodies (diluted 1:100 and 1:1,000, respectively) for 1 h at room temperature. After three washes in PBS, cells were incubated with a cocktail of biotinylated anti-mouse IgG and FITC-conjugated anti-rabbit IgG (both diluted at 1:100 and both purchased from Sigma) for 1 h at room temperature. The staining for α-SM actin was accomplished by incubation with streptavidin-Texas red (1:50, Amersham). All stained cells were examined with a Nikon Optiphot epifluorescence photomicroscope.
We found that 9% of the total clones were positive for SM myosin heavy chain (MHC). Given the possibility that the positive cells were actually SM cell clones, no further studies were performed on these subpopulations. In addition, we also found that cell clones exhibiting a more epithelioid or cobblestone morphology were occasionally isolated. Cells exhibiting this morphology were always examined for expression of factor VIII antigen. If the cells exhibited factor VIII antigen expression, they were considered endothelial clones and not utilized in this study.
Growth of PA fibroblast subpopulations in the presence of serum.
To examine the in vitro growth characteristics of fibroblast subpopulations, serum-stimulated growth was measured as previously described (9, 10). Cells were sparsely seeded (1.0 × 104 cells/well) in MEM containing 10% FBS in 24-well plates, and cell counts were performed on alternate days betweenday 0 and day 10. Media were supplemented, but not replaced, with fresh media containing 10% FBS on day 4and day 8 to avoid blunting growth of the more rapidly proliferating cells. To assess changes in cell number, the cells were trypsinized for 10 min, gently triturated after addition of an equal volume of 10% FBS-MEM, and counted with a standard hemocytometer under light microscope. Data were expressed as cell number × 104/well. Population doubling time was calculated in exponentially growing cells according to a previously described method (33).
Assessment of hypoxia-induced DNA synthesis.
To test whether fibroblast subpopulations isolated from the PA adventitia of neonatal calves had the ability to replicate in response to low oxygen concentration, cells were seeded at a density of 7.5 × 103/cm2 in media containing 10% FBS, allowed to attach overnight, and then growth arrested for 72 h with 0.1% FBS-MEM. At the end of 72 h, the medium was replaced with fresh MEM only. Quiescent fibroblasts were exposed to normoxia and hypoxia in the presence of [3H]thymidine for 24 h in sealed humidified gas chambers as previously described (7,8). At the end of 24 h, cells were harvested for measurement of thymidine incorporation. Briefly, the assay medium was removed, and cells were rinsed with PBS and fixed with 0.2% perchloric acid. After an additional rinse with PBS, the acid-precipitated cellular material was solubilized with 0.01 N sodium hydroxide-0.1% sodium dodecyl sulfate. The contents of each well were then added to 4 ml of Ecoscint H (National Diagnostics, Atlanta, GA), and radioactivity was measured with a Beckman LS 7500 beta-scintillation counter (Irvine, CA). Cell counts were obtained at the end of the 24-h incubation period. Incorporation of [3H]thymidine into DNA was expressed as counts per million per cell.
[3H]Thymidine incorporation in response to growth factors.
DNA synthesis in response to purified mitogens was measured under serum-free conditions according to the previously described method (8, 9). Cells were seeded at 7.5 × 103/cm2 in MEM-10% serum, allowed to attach for 24 h, and then growth arrested for 72 h with MEM-0.1% serum. At time 0 of the test period, medium was replaced with MEM alone, and purified mitogens (PDGF, 30 ng/ml; bFGF, 40 ng/ml) and [3H]thymidine (0.5 μCi/well) were added to each well and exposed to normoxia and hypoxia for 24 h in sealed humidified gas chambers. At the end of the incubation, cells were processed for the measurement of [3H]thymidine incorporation as mentioned in Assessment of hypoxia-induced DNA synthesis.
All data are presented as arithmetic means ± SE. Each observation was reproduced in cells isolated from at least three different animals. One-way analysis of variance followed by Student-Newman-Keuls multiple comparison test was used for individual comparisons within and between groups of data points. Distribution relationship between the animals and the percentage of “hypoxia-proliferative” fibroblast subpopulations was evaluated by the χ2 method. Data were considered significantly different at P < 0.05.
Isolation of purified fibroblast clones or subpopulations from the PA adventitia of neonatal control calves.
To determine whether clones, or at least highly purified primary fibroblast subpopulations, could be generated from the PA adventitia, we utilized the limited dilution cloning technique. Aggregate fibroblast populations were enzymatically dispersed from cleaned adventitia and suspended in media. Serial dilutions of this cell suspension were made such that ∼0.5 cells were placed into each well of a 96-well plate. Thus some wells likely received no cells, whereas others received one or potentially more cells. For each of eight neonatal control calves, we prepared three 96-well plates. We then determined the number of wells with cells that had grown to confluence at the end of 2 wk. We found in the calf with the highest “cloning” efficiency an average (for the three plates) of 17 ± 0.6 confluent wells, whereas in the calf with the poorest cloning efficiency there were only 6 ± 2.8 such wells. On average for all eight calves, 13% of wells (12.8 ± 1.4 wells in 96-well plates) showed growth of fibroblasts to confluence. Nearly 100% of the wells that demonstrated growth in the 96-well plates survived serial passaging for expansion into quantities sufficient for biochemical and functional characterization. Thus our limited dilution cloning technique could be used to establish purified primary fibroblast subpopulations from the main PA adventitia.
Heterogeneity of morphology and contractile protein expression in fibroblast subpopulations.
We sought to determine whether PA adventitia, like other fibroblast-enriched tissues (2, 11, 12, 15, 16, 18, 19, 35, 37,40), was composed of multiple morphologically, biochemically, and functionally distinct populations by analyzing the fibroblast subpopulations generated from each animal. We found that, when examined by light microscopy, the cells in the >120 subpopulations that were evaluated could be described as having two general morphologies, cells that appeared rounded or rhomboidal in shape (Fig.1, A–C) and cells that were more elongated or spindle shaped (Fig. 1, E–G). Within these two broad classifications, there were differences among subpopulations in characteristics such as cell size, pattern of organization (i.e., whorls vs. linear), and shape (Fig. 1). These morphological findings supported the idea that in neonatal calves, the PA adventitia is composed of a mixed population of fibroblasts.
We also utilized α-SM actin immunoreactivity to differentiate fibroblast subpopulations. Marked differences in α-SM actin expression were observed between fibroblast subpopulations (Fig.2). Some cells expressed α-SM actin in stress fibers (Fig. 2, A and B), whereas others expressed it in the cytoplasm (Fig. 2 C). There were also adventitial fibroblast subpopulations that did not express α-SM actin (Fig. 2 D).
Heterogeneity in growth potential of fibroblast subpopulations.
To determine if the isolated fibroblast subpopulations exhibited differences in growth capabilities, we generated growth curves for 42 different fibroblast subpopulations isolated from PA adventitia of neonatal control calves. Marked differences in growth capacity in the presence of 10% serum-containing media were noted between the subpopulations. An example demonstrating the range of growth differences in three subpopulations is shown in Fig.3 A. Using doubling time as a measurement of growth rate, we found that the variability in growth among the subpopulations was great, as indicated by the histogram (Fig.3 B). Doubling times varied from as short as 11 h to as long as 80 h and in some cases over 100 h. The average slope (ratio of cell count/day) of the growth curve of the clones also demonstrated marked heterogeneity of growth between cell populations (Table 1). Finally, the saturation density at confluence also varied remarkably among the fibroblast clones (Table 1) consistent with possible differences in cell size, contact inhibition or lack thereof, or both.
To assess possible relationships between morphology, α-SM actin expression, and growth, 22 fibroblast clones underwent analysis of all variables. We observed that the rounded or rhomboidal cells tended to have shorter rather than longer doubling times, which implied that most of the more slowly growing cells had an elongated morphology. However, the more rapidly growing cells were nearly evenly divided between round and elongated cells. Thus cell morphology was not a reliable marker of the growth capability of fibroblast subpopulations in this study. In addition, there was no consistent predictive relationship between α-SM actin expression and growth capability, since within both α-SM actin-positive and -negative subpopulations, we observed the potential for rapid, moderate, or slow growth (Table 1).
Heterogeneity among fibroblast subpopulations in their growth response to hypoxia.
We considered the possibility that, in addition to differences in morphology, actin staining, and growth capability in serum, fibroblast subpopulations might differ in their proliferative response to hypoxia. In a preliminary experiment, we selected two different fibroblast subpopulations from the same neonatal calf, plated them at 10 × 103 cells/well, and grew them for 7 days in the presence of serum under either normoxic or hypoxic conditions. At the end of 7 days, subpopulation A had cell counts of 79 ± 9 × 103/well in normoxia and 80 ± 11 × 103 in hypoxia, whereas subpopulation B had counts of 58 ± 4 × 103 in normoxia and 90 ± 5 × 103 in hypoxia. These findings suggested that despite slower growth in normoxia, subpopulation Bdemonstrated a significantly higher growth rate under hypoxic conditions, whereas subpopulation A did not. We then growth arrested these same two subpopulations for 72 h with 0.1% FBS containing media and exposed them to hypoxia in absence of any exogeneous mitogens for 24 h and measured DNA synthesis. Hypoxia selectively increased thymidine incorporation in subpopulation B but not in A (Fig. 4), suggesting that hypoxia acts as growth-promoting stimulus for only selected fibroblast subpopulations. The fibroblast subpopulations that had a more than twofold increase in hypoxia-induced proliferation were termed as hypoxia proliferative, and the subpopulations that did not exhibit any increase in DNA synthesis in response to hypoxia were designated as hypoxia nonproliferative.
Because in physiological systems hypoxia is likely to act in concert with other locally produced growth stimuli, we also asked whether there might be differences between the subpopulations when purified peptide mitogens were added in the presence of hypoxia. In hypoxia-nonproliferative subpopulations, we found as expected that PDGF and bFGF stimulate proliferation under normoxic conditions. However, when each of the growth factors was added to the same cell population and then exposed to hypoxia, there was no further increase in mitogen-induced DNA synthesis (Fig.5 A).
In hypoxia-proliferative subpopulations, PDGF- and bFGF-induced proliferation under normoxic conditions, as expected. However, when each of the growth factors was added and the cells were then exposed to hypoxia, purified mitogen-stimulated thymidine incorporation was greatly augmented compared with that observed during normoxia (Fig.5 B). These results suggest that synergistic growth responses between hypoxia and peptide mitogens exist for only select fibroblast subpopulations.
We then evaluated DNA synthesis using [3H]thymidine incorporation in response to hypoxia in 31 different fibroblast subpopulations isolated from neonatal control calves. We found that, compared with normoxic conditions, hypoxia-induced changes in DNA synthesis ranged from a decrease of 33% to an increase of 265% in these subpopulations (Fig.6 A). These results suggest that PA fibroblast subpopulations differed markedly in their DNA synthetic capabilities in response to hypoxia.
Adventitial fibroblast subpopulations from chronically hypoxic calves.
Because hypoxic pulmonary hypertension is associated with dramatic fibroproliferative changes in the adventitia (45, 46), we considered the possibility that the adventitia of hypoxia-induced pulmonary hypertensive calves might have fibroblast subpopulations that differed from control calves in regard to morphology, α-SM actin expression, and growth capabilities either under normoxic or hypoxic conditions. First, we found that the average cloning efficiency of fibroblasts from hypoxic hypertensive calves (11.3 ± 1.0) was very similar to that of the control calves (12.8 ± 1.4). In 32 fibroblast subpopulations from hypertensive calves that were examined with regard to morphology and α-SM actin expression, 41% exhibited a rounded morphology and 59% of the subpopulations were elongated (Table2), which was not different from the 32 and 68% distribution observed, respectively, in control calves. The microscopic appearances of the round and elongated cells in the subpopulations from hypertensive calves were similar to those from control calves. In hypoxic hypertensive calves, 61% of subpopulations stained positively for α-SM actin expression and 11% did not, which was not different from the values in control calves (68 and 14%, respectively). As in control calves, even for subpopulations that positively expressed actin, the pattern of expression could differ, where, for example, some fibroblasts showed actin expression in stress fibers, whereas in others, it was expressed diffusely within the cytoplasm. Rounded cells could be either actin positive, negative, or mixed, and elongated cells could either be actin positive or mixed (Table 2).
We examined 51 fibroblast subpopulations from nine hypoxic pulmonary hypertensive calves for their serum-stimulated growth capabilities under normoxic conditions. These subpopulations were not different from the subpopulations of control calves with regard to the distribution of cell doubling times under serum-stimulated conditions (i.e., the pattern of distribution was identical to that shown in Fig.3 B).
In response to hypoxia in the absence of exogeneous mitogens, we found essentially the same magnitude of changes from baseline as was observed in cell populations from control calves (Fig. 6 A). However, the hypoxic pulmonary hypertensive calves had more than twice as many hypoxia proliferative (with a more than twofold increase in hypoxia-induced DNA synthesis) fibroblast subpopulations than did the control calves (Fig. 6 B). These results suggested that chronic hypoxia induces selective expansion of hypoxia-proliferative fibroblast subpopulations in the adventitia of hypertensive calves.
The main findings in the present study were that: 1) the adventitia of the normal neonatal bovine PA is composed of numerous phenotypically distinct subpopulations of fibroblasts and 2) chronic hypoxia-induced pulmonary hypertension is associated with a selective increase in the number of resident fibroblast subpopulations with enhanced growth capability under hypoxic conditions. Given the extensive literature showing fibroblast heterogeneity in other tissues (11, 15, 24, 34, 35, 48), it was perhaps not surprising that we would find fibroblast heterogeneity in the PA adventitia, even though this has not been previously reported. Remarkable, however, was the nature, magnitude, and frequency of the differences between the subpopulations. The characteristics which differed between subpopulations included: 1) morphology, 2) the presence or absence of actin staining, 3) actin expression pattern, i.e., cytoplasmic vs. stress fiber distribution, 4) the serum-stimulated growth rate under normoxia and hypoxia,5) the DNA synthesis (in the absence of serum) in normoxia and hypoxia, and 6) growth factor (PDGF and bFGF)-induced proliferative responses under normoxic and hypoxic conditions. Differences in these characteristics were observed among subpopulations generated from several animals. Importantly, the growth rate pattern of the heterogeneic subpopulations obtained from one animal had a striking resemblance to subpopulations generated from other animals, indicating that the observed heterogeneity was a consistent and true characteristic of fibroblast subpopulations of the bovine PA adventitia. Our results are consistent with reports demonstrating heterogeneity of cells composing the endothelium and media (4,14, 27, 38) and, as discussed below, extend the heterogeneity concept to vascular fibroblasts by showing in a single study that numerous functionally distinct subpopulations exist within the adventitia.
The present study examined only a few of many known fibroblast characteristics. From lung tissue, for example, fibroblasts have been shown to be heterogeneous with regard to morphology (15,34), proliferative capacity (6, 20, 24, 34, 48), surface marker (Thy-1, MHC class II, CD4) expression (11,34), type and amount of collagen production (11,16), intracellular metabolic pathways (15), and cytokine production (15). In addition, fibroblasts from nonpulmonary tissues have been shown to be heterogeneous with regard to proliferation (2, 22), size and shape (18,19), presence of lipid droplets (29), collagen synthesis (13, 16, 21, 36, 40, 49, 51), expression of growth factors (16), response to and synthesis of PGE2, and expression of C1q surface receptors, among others. One wonders about heterogeneity in some of the other key functions of fibroblasts, such as capacity for transition to other cell types (i.e., myofibroblast and/or SM cell), migration, protease expression, and propensity to apoptosis. Previous studies have usually focused on a single characteristic or group of characteristics of fibroblast heterogeneity that might account for a particular fibroblast function, response to injury, or disease process. The present study on heterogeneity itself has expanded the prior concepts by showing within a mixed fibroblast population from a single normal neonatal tissue (i.e., the PA adventitia) the marked and frequent variability in numerous characteristics of fibroblast subpopulations.
One important aspect of fibroblast function, the ability to proliferate under hypoxic conditions, has received only limited attention in the literature. We and others have documented that aggregate populations of fibroblasts from the PA demonstrate increased DNA synthesis under hypoxic conditions (7, 8, 44, 50). Similar findings have been reported in aggregate populations of fibroblasts from skin and gingiva (1, 5, 32, 44, 47). Our findings suggest that proliferation under hypoxic conditions is not a response that characterizes all fibroblast subpopulations but, rather, that it is a response limited to only select subpopulations of fibroblasts. It is possible, for instance, that some vascular beds or tissues have a higher proportion of hypoxia-proliferative subpopulations than others. For example, the PA could have a higher proportion of hypoxia-proliferative fibroblast subpopulations than certain systemic vascular beds. This might explain our previously reported findings wherein aggregate fibroblast populations from the pulmonary circulation demonstrate more consistent proliferative responses to hypoxia than aggregate fibroblast populations from the aorta (7). It could also explain the marked variability in hypoxic responses in aggregate fibroblast populations derived from one animal versus another, since mixed populations could vary considerably in the numbers of hypoxia-proliferative cells they contain. Furthermore, the fact that purified populations of fibroblasts with unique hypoxic responses can be generated and maintained in culture will allow investigation into the basic mechanisms that contribute to hypoxia-induced cell proliferation.
It should be noted that we were not able to identify a characteristic that allowed us to predict whether a fibroblast would proliferate under hypoxic conditions or not. This is unlike our findings for hypoxia-responsive and nonresponsive cells within the media of the vessel wall (14). Furthermore, our findings suggest a wide variability of the proliferative responses to hypoxia among fibroblast subpopulations, which is different from our findings of nearly an “all or none” phenomenon in medial smooth muscle cell. It will be important in the future to identify proteins that either confer or associate with hypoxic responsiveness so that more accurate evaluations of select cell expansion can be evaluated in vivo.
Although this study was designed simply to demonstrate that different characteristics exist among the subpopulations of fibroblasts, some comment can be made about the interrelationship of characteristics within a given subpopulation. For example, cells that were smaller and rounded usually (but not always) grew faster in serum than cells that were larger and spindle shaped. But the relationship between morphology and growth rate was weak, because either round-shaped cells or spindle-shaped cells could grow rapidly, and either shape could grow slowly. Nor could morphology be related to the actin expression or the expression pattern. Also, the expression or the pattern of expression of actin did not relate to proliferation. Furthermore, subpopulations that replicated rapidly in normoxia did not necessarily do so in hypoxia. These findings, together with the great spectrum among the subpopulations of growth rates in normoxia and hypoxia, suggest an enormous heterogeneity in aggregate fibroblast populations.
Given the inherent heterogeneity of fibroblasts within the vascular adventitia, we wondered how these subpopulations would respond to a specific in vivo stimulus, namely chronic hypoxia, which, in the newborn calf, causes a thickened and hypercellular PA adventitia (45). There were at least three possibilities:1) all or nearly all of the fibroblast subpopulations could increase their replicative capacity in response to chronic hypoxia,2) those subpopulations that normally reside in the adventitia and that have the potential for a replicative response to hypoxia might show a heightened response, and 3) subpopulations of fibroblasts that are particularly responsive to hypoxia might become more numerous. Comparing subpopulations from control versus hypoxic calves indicates no statistically significant shift in the spectrum of proliferative responses and a similar range of hypoxia-induced growth responses in normoxic and hypoxic calves. These findings oppose, respectively, the first and second possibilities listed. However, the finding of more subpopulations with greater than twofold increase in hypoxia-induced proliferation in hypoxic calves than in control calves supports the third possibility, namely, an increase in the number of hypoxia-proliferative subpopulations. Furthermore, similar morphologies and α-SM actin expression patterns in fibroblast subpopulations isolated from control and hypoxic calves at least raise the possibility that the hypoxic exposure promoted expansion of these selective subpopulations. Our findings of selective expansion of hypoxia-proliferative fibroblast subpopulations during hypoxia-induced pulmonary hypertension is consistent with the recent report of Panchenko et al. (32) demonstrating selective proliferation of a subset of hypoxia-adapted fibroblasts during skin fibrosis.
In view of the important role of cell-cell interactions in the coordinate control of cell activity, any alteration in the relative proportion of different fibroblast subpopulations within a tissue might be expected to significantly influence the behavior of the entire organ. Expansion of hypoxia-proliferative fibroblast subpopulations under hypoxic conditions could lead, through alterations in the production of paracrine factors as well as extracellular matrix proteins, to changes in SM as well as endothelial function. Fibroblasts are known to significantly influence endothelial and epithelial function (39). For example, fibroblasts from patients with psoriasis induce normal keratinocytes to display many of the aberrations in cell proliferation and differentiation associated with psoriasis (17). Expansion of fibroblasts with unique characteristics could thus have potentially significant and as yet unexplored effects on endothelial and SMC function and could thus contribute in many ways to the functional abnormalities described for hypertensive PA.
On the basis of the results of the present study we propose that the adventitia of the normal PA contains an array of fibroblast subpopulations, differing markedly in numerous biochemical and functional characteristics. One wonders what biological advantage might accrue to such great heterogeneity. A speculative answer for vascular fibroblasts might refer to the large number of normal functions required of these cells, such as providing elasticity for a threefold pressure change within the lumen during each cardiac cycle, structural integrity for a three- to fourfold change in mean pressure during exercise, responding to the hormonal changes during pregnancy, as well as participating in the developmental changes within the arterial wall throughout fetal, neonatal, and adult life. Furthermore, in response to injury, fibroblasts have been shown to play an important role in the structural remodeling of the vascular wall (28, 41). Specific fibroblast subpopulations with certain functions may increase in number in response to specific stimuli and play unique roles in the structural alterations occurring during injury. Although we have suggested that this may occur with chronic hypoxia in the PA adventitia, it needs both confirmation and examination of mechanism through further research.
We thank Steve Hofmeister and Sandi Walchak for harvesting bovine PA tissue, preparing the final figures, and immunofluorescence staining of the cells.
This work was supported by National Institutes of Health Grants HL-64917-01A1, HL-14985, and SCOR-56481. M. Das was supported by a postdoctoral fellowship from the American Heart Association, Arizona, Colorado & Wyoming Affiliate, a Giles Filley Research Award from the American Physiological Society, and a research grant from the American Lung Association.
Preliminary results were presented at Experimental Biology Meetings in New Orleans, LA (8c), and San Francisco, CA (8a); the 10th International Vascular Biology Meeting in Cairns, Australia, 1998; and the American Thoracic Society Annual Meeting in San Diego, CA (8b).
Address for reprint requests and other correspondence: M. Das, Developmental Lung Biology Research Labs, B-131, Univ. of Colorado Health Sciences Center, 4200 E. 9th Ave., Denver, CO 80262 (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 © 2002 the American Physiological Society