In VitroCell.Dev.Biol.28A:403-409,June 1992 © 1992TissueCultureAssociation 0883-8364/92 $01.50+0.00
ROLE O F E N D O T H E L I A L OF CULTURED
CELLS IN THE PROLIFERATIVE RESPONSE PULMONARY VASCULAR SMOOTH MUSCLE CELLS T O REDUCED OXYGEN T E N S I O N ROBERT L. VENDER
Department of Medicine, Division of Pubnonary a~d Critical Care Medicine, The Medical Collegeof Pennsylvania, 3300 Henry Avenue, Philadelphia, Pennsylvania 19129 (Received 21 November 1991; accepted 17 December 1991)
SUMMARY The development of pulmonary hypertension in a wide variety of human disease states and experimental animal models characterized by chronic alveolar hypoxia is mediated by two pathologic vascular processes, a) vasoconstriction and b) vasoconstruction (structural remodeling). The anatomic changes seen within the pulmonary circulation include a) increased deposition of collagen and elastin in the adventitial layer and b) aberrant pulmonary vascular smooth muscle cell proliferation and maturation in the medial segments. Despite the demonstrated ability of pharmacologic manipulation in the experimental animal to ameliorate both the structural and hemodynamic changes, the actual etiologic mechanisms are only beginning to be explored. Using tl~e cell culture technique of co-cultivation, we have investigated the potential role of bovine pulmonary arterial endothelial cell-derived factors in mediating abnormal bovine smooth muscle cell growth under conditions of reduced oxygen tension. We have demonstrated that these cultured endothelial cells exposed in vitro to reduced levels of atmospheric oxygen concentrations of 5.0% and 2.5% 02 for durations of 24 to 72 h produce and secrete soluble growth factor(s) which stimulate smooth muscle cell proliferation when compared to cells maintained under standard tissue culture oxygen conditions of 95% room air. This growth-stimulatory effect required the concomitant presence of serum factors (0.5% fetal bovine serum), was inhibited by heparin, was distinct from platelet-derived growth factor, and seemed to have a molecular weight greater than 14 000 Da. We conclude that reduced levels of oxygen tension in vitro can selectively induce pulmonary arterial endothelial cells to release mitogen(s) which can stimulate vascular smooth muscle replication. Furthermore, we speculate that this in vitro finding may be of importance as an etiologic mechanism to explain the accelerated smooth muscle cell growth characteristic of hypoxic pulmonary arteriopathy.
Key words: hypoxia/endothelial cell; derived mitogens. INTRODUCTION The development of structural changes within the puhnonary circulation under conditions of chronic generalized alveolar hypoxia is an established pathologic entity that occurs in a wide variety of human disease states and experimental animal models which eventuate in the hemodynamic complication of pulmonary hypertension. These alterations are thought to be of an irreversible nature and are theorized to herald late or end stage disease sequelae. These characteristics are in contrast to the vasoconstrictive response of the pulmonary artery to hypoxia which represents a potentially reversible physiologic process regardless of disease duration or severity. Under conditions of chronic alveolar hypoxia, the complexity of this pathologic structural remodeling has been demonstrated throughout the entirety of the pulmonary circulation and throughout all the structural layers of the arterial wall. These changes include: a) the appearance of longitudinally oriented smooth muscle cells within the intima, b) the abnormal deposition of increased amounts of collagen and elastin within the adventitia, c) medial smooth muscle cell hyperplasia and hypertrophy, and d) the differentiation of poorly developed precursor smooth muscle cells (termed intermediate cells and pericytes) into mature smooth muscle cells (9,15,19,23). We interpret these latter two findings to represent
alterations in smooth muscle cell growth. Although pharmacologic manipulation has been shown to ameliorate some of these anatomic changes in experimental animal models, the actual etiologic factors have yet to be identified (7,17). Given the known intimate anatomic and physiologic association between endothelial cells and smooth cells within the pulmonary circulation and recognizing the importance of peptide growth factors in the control of smooth muscle cell proliferation, we hypothesized that soluble endothelial cell-derived mitogens, specifically released under conditions of reduced oxygen tension, might be of potential etiologic significance in explaining one aspect of these structural changes, namely, aberrant smooth muscle cell replication. Using the cell culture technique of co-cultivation to study this aspect of endothelial-smooth muscle cell interaction we now present data that suggests the selective release under in vitro conditions of reduced oxygen content of a soluble endothelial cell-derived mediator that acts in the presence of serum components as a mitogen for cultured pulmonary vascular smooth muscle cell proliferation. MATERIALSANDMETHODS Tissue culture medium 199, minimal essential medium (MEM), penicillin-streptomycin, and trypsin were obtained from GIBCO, Grand Island, 403
404
VENDER
NY; fetal bovine serum (FBS) from Hyclone, Logan, UT; Miilicell-HA culture plate inserts from Millipore Corp., Bedford, MA; Cytodex-3 microcarrier beads from Pharmacia, Inc., Piscataway, NJ; bovine insulin, bovine heparin, HEPES, deoxycytidine, and thymidine from Sigma Chemical Co., St. Louis, MO; antibodies directed against human platelet-derived growth factor (PDGF) and bovine fibroblast growth factor (basic) from R&D Systems, Minneapolis, MN. Endothelial cells. Under sterile conditions, endothelial cells were recovered from the large extraparenchymal pulmonary arteries of freshly killed calves by gentle scraping with a scalpel blade (21). The recovered endothelial cells were initially grown in 25-cm ~ tissue culture flasks containing 5.0 ml of medium 199 supplemented with 10% heat-inactivated FBS, deoxycytidine (5 #g/ml), thymidine (5 #g/ml), penicillin (100 U/ml), and streptomycin (100 #g/ml) and grown in a humidified atmosphere of 5% COs plus 95% room air at 37 ° C. Once confluent, the endothelial cells were subcultured using 0.25% trypsin and grown again under similar conditions as described above. Endothelial cells were identified by their typical cobblestone morphologic appearance under phase contrast microscopy. For cocultivation experiments, confluent flasks of endothelial cells were removed with 0.25% trypsin and transferred to siliconized 250-ml magnetic stirring vessels containing sterile prehydrated collagen-coated, dextran-based microcarrier beads (Cytodex-3) suspended in 100 ml of medium 199 supplemented with 10% FBS, deoxycytidine, thymidine, penicillin, streptomycin, and 25 mM HEPES (N-[2-hydoxyethyl] piperazine-N [2-ethane sulfonic acid]). This entire suspension system was then placed in a standard tissue culture incubator containing 5% COs plus 95% room air at 37 ° C for an initial 2 to 4 h without stirring and with the caps to the side arms of the stirring vessel loosened. After pre-incubation, the caps were tightened, and the sealed vessel transferred to a nonhumidified room air oven at 37 ° C and stirred continuously at a speed of 15 rpm. The suspension growth media was replenished twice weekly by removal of one-half of the vessel volume, followed by replacement of an equal volume of fresh media containing identical supplements. At each refeeding, the entire sequence of events was repeated. On a weekly basis, serial aliquots of the suspension media containing the microcarrier-borne endothelial cells were removed, the beads inspected for confluency, and the endothelial ceils counted manually by hemocytometry after detachment from the beads with 0.25% trypsin. Additional microcarrier beads were added to the system as needed to achieve endothelial cells in appropriate numbers as required for the particular experimental condition. Endothelial cells from Passages 2 to 10 were used for study. Smooth muscle cells. After the recovery of endothelial cells, the entire intimal surface of the vessel was vigorously scraped with a sterile scalpel blade to remove any remaining adherent endothelial cells. The vessel surface was then washed extensively with phosphate buffered saline (PBS) to remove the detached cells and debris. The vessel was then inverted and the adventitial layer peeled off. This surface was also extensively washed with PBS and the remaining specimen transferred to a sterile petri dish with the intimal surface facing upward. Two millimeter explants of the medial segment were then obtained using standard dermatologic punch biopsy instrument and placed into 35-mm diameter, 6-well tissue culture plates containing minimal essential medium plus 24 mM sodium bicarbonate (MEM) supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 #g/ml). The explants were then placed in a fully humidified tissue culture incubator at 37 ° C in an atmosphere of 5% COs plus 95% room air (14). Fresh medium was added dally until smooth muscle cells appeared in sufficient density for passage. At this time, the smooth muscle cells were detached with 0.25% trypsin and subcultured into 25 cm 2 tissue culture flasks containing 5.0 ml of MEM plus 10% FBS and antibiotics and grown at 37 ° C in an atmosphere of 5% CO2:95% room air. Smooth muscle cells were identified by morphologic appearance under phase contrast microscopy showing the characteristic formation of hillocks and valleys postconfluence (20). Smooth muscle cells from Passages 2 to 10 were used for study. Smooth muscle cell growth bioa.~say. Pulmonary vascular smooth muscle cells were initially plated at a constant density into 35-mm diameter, 6-well tissue culture plates. This initial plating density between experiments ranged from 1.0 to 7.0 × 103 cells per well (100 to 700 cells/cm2). These cells were initially incubated in 2.0 ml per well of MEM supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 pg/ml) in an atmosphere of 5% COs plus 95% room air at 37 ° C for 72 to 96 h. This
period of time was chosen to allow the ceils to become attached and stabilized after exposure to trypsin. The original medium was then removed, the cell layer washed twice with serum-free medium and refed with 2.0 ml per well of serum-supplemented medium 199:MEM ( h 1). Random wells were selected and cell number assayed by hemocytometry before experimental manipulations to determine the baseline starting cell number values. Subsequent data will be analyzed in relation to this starting cell density or as a ratio of the negative control cell number value after each specified incubation duration. At the same time, adequate numbers of endothelial cells grown to confluence on microcarrier beads were removed from the suspension apparatus, placed into a 50-ml centrifuge tube, allowed to sediment, and the supernatant aspirated. These endothelial cells were then washed twice with a 10-fold volume excess of serum-free medium 199:MEM ( h l ) , so as to dilute out the high concentrations of FBS in the original suspension medium, and then resuspended in the necessary volume of medium 199:MEM plus specified concentrations of FBS, as required for the particular experimental situation. Mierocarrier-borue endothelial cells ( 1.0 X 106 cells per well) were then added to the inside surface of individual 12-mm diameter culture plate inserts (Millicell-HA:Millipore Corp.) which were previously placed within the 6-well culture plates containing the smooth muscle cells. The entire apparatus was then incubated at 37 ° C for durations of 24 or 72 h in an atmosphere of either a) 5% CO2 plus 95% room air (approximate 02 concentration of 18 to 20%); b) 5% CO2 plus 5.0% 02, balance Nz; or c) 5% COs plus 2.5% 03, balance N2. After each 24-h period, the medium was removed and the cells refed with fresh medium identical to the initial composition. After the desired period and conditions of exposure, the inserts were removed, the adherent smooth muscle cells detached by two consecutive 10-min incubations with 0.25% trypsin, and counted manually by hemocytometry. Pulmonary vascular smooth muscle cells incubated for similar durations at 37 ° C in the absence of inserts, beads, and endothelial cells in medium 199:MEM plus 0.5% FBS in 5% CO2 plus 95% room air served as negative controls. Pulmonary vascular smooth muscle cells incubated for similar durations at 37 ° C in the absence of inserts, beads, and endothelial cells in medium 199:MEM supplemented with 10% FBS in 5% COs plus 95% room air served as positive controls. To determine the modulating effect of bovine heparin on any demonstrated smooth muscle cell proliferation, additional experiments were carried out in the presence of 1.5 to 150 gg/ml heparin for the entire duration of the co-cultivation period (2). In addition, identical experiments were performed in the continuous presence of blocking antibodies directed against humanPDGF (40 ttg/ml) and bovine FGF-b (40 #g/ml). For molecular sizing studies, spectrapor dialysis tubing of a specified exclusion limit was adhered to the inside of the inserts using a nontoxic sealant, and similar growth assays performed. Oxygen exposure conditions. The "normoxic" tissue culture exposure conditions of 5% COs plus 95% room air were accomplished using a standard tissue culture incubator, fully humidified at 37 ° C. The reduced oxygen exposure conditions were performed using a Forma Scientific (model 3159) tissue culture incubator with three gas capability, employing an oxygen sensor which automatically adjusts the inflow of COs and N2 into the incubator within an accuracy of 0.5% of the preset atmospheric condition, namely 5% COs plus 5% 02, balance N2 and 5% COs plus 2.5% O2, balance N2. Periodically, medium was sampled anaerobically into a 1.0-ml syringe and analyzed immediately with Corning 170 pH/blood gas analyzer for measurement of oxygen partial pressure (Po2). Statistical analysis. All values are shown as means + SEM. Experimental data are expressed as a ratio in comparison to appropriately matched control values as indicated. A one sample t test was used for statistical analysis comparing mean values of these ratios to control value of 1.0. A paired-sample t test was used to make comparisons for the growth modulation, molecular sizing, and endothelial cell dilution studies. P < 0.05 was assumed to represent statistically significant differences (3).
RESULTS The measured dissolved oxygen partial pressure (Po2) as well as the calculated oxygen content (based on an oxygen solubility o f 0 . 0 2 4 4 ml 02 per 1.0 ml H20 at 1.0 atm) of the experimental medium after the time durations of 14 and 2 4 h in atmospheric oxygen concentration of either 9 5 % room air, 5 . 0 % 02 or 2 . 5 % 02
405
HYPOXIC ENDOTHELIAL-SMOOTH MUSCLE INTERACTION TABLE 1 MEASUREMENTS OF TISSUE CULTURE MEDIA OXYGEN CONCENTRATIONS Atmospheric Exposure Conditions
14 Incubation duration, h Range, mmHg Partialpressure (mmHg)~ Content, ml O2/100 ml media~ 24 Incubation duration, h Range, mmHg Partial pressure (mmHg)° Content, ml O J 1 0 0 ml media a
95% Room Air
5.0% Oxygen
2.5% Oxygen
n = 4 144-160 151.9 + 4.4 0.49 + 0.01 n= 4 146-156 150.5 + 2.6 0.48 _+ 0.01
n = 2 71-72 71.7 ± 0.7 0.23 _-+0.01 n = 3 53-60 55.8 + 2.0 0.18 + 0.01
n = 1 49.5 49.5 0.16 n= 4 38-50 46.8 ± 2.5 0.15 ± 0.01
a Mean + SEM.
are shown in Table 1. As can be seen, the reduced atmospheric oxygen conditions of 2.5% and 5.0% O 2 for 24 h resulted in an overall mean Po~ of 50.7 mmHg. Twenty-one separate experimental sets of data were obtained to generate the results recorded in the accompanying tables and figures. The values for the starting cell number (recorded as mean + SEM) ranged from'0.80 + 0.12 to 15.75 + 0.24 X 104 cells. Despite variability in starting cell numbers between experiments, within each individual study the range of values was quite close and in itself could not account for the demonstrated changes in cell number that we observed under the various experimental conditions. Under the negative control conditions of 0.5% FBS, 95% room air oxygen concentration and the absence of endothelial cells, we observed only an 18% increase in smooth muscle cell number after 72 h incubation when compared to starting cell number. Conversely, under the positive control conditions of 10% FBS, 95% room air oxygen concentrations, and the absence of endothelial cells there was a statistically significant increase in all smooth muscle cell measurements at both the 24- and 72-h incubation periods, regardless of whether these values were compared to either matched negative control values for cell number at these same time points or to starting cell numbers (Figs. 1 and 2). This increase in smooth muscle cetl number above starting values after supplementation with 10% FBS amounted to a near doubling (1.89 +_ 0.09) after 24 h and a 5.54 -+ 0.39-fold increase at the conclusion of the 72-h protocol. In separate experiments, where microearrier beads without endothelial cells (naked heads) were handled in a manner identical to studies employing microearrier-borne endothelial cells, no significant difference was observed in smooth muscle cell number between cells grown in 0.5% FBS without the concomitant presence of inserts and naked microcarrier beads compared to cells maintained in 0.5% FBS in the presence of these devices in oxygen concentrations of 95% room air and 5.0% 02. When data were expressed as a ratio of smooth muscle cell number with beads alone divided by cell number without beads after 24 and 72 h incubation, respectively, these values measured 1.07 + 0.05 and 1.01 + 0.24 for the 95% room air conditions and 1.18 +_-0.08 and 0.87 + 0.06 for the 5.0% O2 exposure conditions. The addition of 1.0 X 106 pulmonary arterial endothelial cells to the smooth muscle cells through co-cultivation in 0.5% FBS and
95% room air oxygen concentration elicited a significant increase in smooth muscle cell number after the 24 h incutJation when compared to both negative control values at this same time point and to starting cell number values (1.14 + 0.04 and 1.16 + 0.05, respectively). However, this mitogenic effect of endothelial cells on smooth muscle cell proliferation in the standard 95% room air conditions was not evident after 72 h incubation (Figs. 1 and 2). The discrepancy between the observed increase in smooth muscle cell number at 24 h and its absence at 72 h probably resulted from the previously noted 18% increase in cell number achieved by the addition of 0.5% FBS alone in our negative control conditions, which effectively masked any further increase in smooth muscle cell number achieved by the concomitant addition of endothelial cells in co-culture. However, this data do suggest the spontaneous release of an endothelial cell-derived growth factor into the co-culture media under standard tissue culture conditions of 95% room air:5% CO2, which stimulates smooth muscle cell proliferation of a minor degree. No consistent increase in pulmonary vascular smooth muscle
2.0 O }z
*t
*t
O0
r_.o *t 1.0
z3
20% 02
2,5 - 5.0"/* 0:~
24 HOUR INCUBATION OURATION
El zo% 02
2,s - s.o% o~
7 2 HOUR INCUBATION OUP.ATtOH
FIG. 1. Effect of reduced atmospheric oxygen concentration upon cultured PVSMC proliferation in the absence (-) or presence (+) of PAECs through co-cultivation in 0.5% FBS conditions. Data are expressed as the PVSMC number ratio experimental/negative control and recorded as mean - SEM. Corresponding values in the presence of 10% FBS for 24 and 72 h measured 2.09 ± 0.09 and 4.59 + 0.41, respectively. *P < 0.05 compared to 1.00; ~P < 0.05 compared to cell number ratio in the absence of PAEC under similar conditions of oxygen exposure.
406
VENDER
2.0 A
|
1.0 Z-I
24 HOURIr~ICUB~TIONDU~TI~I
72 HOURINCUSATIC~ DU~IlON
FIG. 2. Effect of reduced atmospheric oxygenconcentrationupon cultured PVSMC proliferationin the absence (-) or presence (+) of PAEC through co-cultivationin 0.5% FBS conditions.Data are expressed as the PVSMC ratio experimental/startingcell number and recorded as mean + SEM.Correspondingvaluesin the presenceof 10% FBS for 24 and 72 h measured 1.89 + 0.09 and 5.54 + 0.39 respectively. *P < 0.05 compared to 1.00; "~P< 0.05 compared to cell number ratio in the absence of PAEC under similarconditionsof oxygenexposure.
cells (PVSMC) number was elicited upon exposure to reduced levels of atmospheric oxygen alone, i.e. in the absence of pulmonary arterial endothelial ceils (PAEC). However, the co-cuhivation of pulmonary arterial endothelial cells at a concentration of 1.0 × 106 cells per well with pulmonary vascular muscle cells in 0.5% FBS and their subsequent exposure to reduced levels of oxygen tension (2.5% to 5.0% 02) for both 24 and 72-h incubationsresulted in a reproducible, statistically significant increase in smooth muscle cell number above starting values and when compared to the negative control value at the same points in time (Figs. 1 and 2). This increase in smooth muscle cell number was remarkably constant under all experimental conditions. When expressed in Fig. 1 as a ratio to negative control values of cell number at the same time points, these increases measured 1.45 + 0.09 at 24 h and 1.43 + 0.07 at 72 h. In addition, both the endothelial and smooth muscle cells seemed morphologically normal by phase contrast microscopy under all the above-mentioned culture conditions in 0.5% FBS. An increase in smooth muscle cell number of greater than 20% above negative control values upon co-cultivation with endothelial cells under reduced oxygen conditions was seen in 28/37 (76%) of studies after a 24-h incubation. However, by extending our assay period an additional 48 h to a total incubation of 72 h, an increase in smooth muscle cell number of greater than 20% above negative control values was demonstrated in 11/12 (92%) of experiments. All previously reported data utilized PAEC at a constant density of 1.0 X 106 cells per well. We also performed studies using variable concentrations of endothelial cells to assess a) the minimal number of cells required to elicit a mitogenic effect in our co-cultivation system and b) the differentiation of potential growth inhibitory substances by dilution analysis. Consistent with the previously reported data, PAEC at a concentration of 1.0 )< 106 cells per well effected a statistically significant increase in PVSMC number upon exposure to 5% 02 compared to matched control cells grown in
absence of endothelial ceils. No such increase was seen under 20% 02 conditions. This difference in mitogenic response between the two levels of atmosphere oxygen exposure was also observed at endothelial cell densities of 1.0 )< 104 and 1.0 × 1.03 cells per well (Fig. 3). A growth stimulatory effect was not observed at the lowest endothelial cell density of 1.0 × 102 cell per well. Finally, the concomitant addition of heparin at dosages of 1.5, 15, and 150 #g/ml to the co-cultivation bioassay system significantly inhibited the previously observed increase in SMC number elicited in the presence of endothelial cells and a reduced oxygen level of 5.0% 02. The degree of inhibition was 62 -+ 2.0%, 93 + 4.7%, 59 + 6.6% for the various heparin concentrations, respectively (Table 2). In addition, the demonstrated increase in PVSMC number was not inhibited by the concomitant presence of blocking antibodies directed against PDGF or FGFb, and was specifically excluded by dialysis tubing with a molecular weight cut off less than 14 000 Da (Table 2). However, it should be noted that the addition of anti-FGFb immunoglobulin to our co-cultivation system did cause a mild degree of PVSMC growth inhibition, although this value did not achieve statistical significance. DISCUSSION In this study, utilizing the cell culture technique of co-cultivation, we report the selective induction and release from confluent cultures of PAEC of a soluble mitogen(s) stimulatory for cultured PVSMC proliferation upon in vitro exposure to reduced levels of atmospheric oxygen concentration of 5.0% and 2.5% 02, when compared to cells maintained under standard tissue culture oxygen conditions (5% CO2:95% room air). This SMC growth stimulatory effect was dependent on the presence of serum supplementation (0.5% FBS) and not observed under serum-free conditions. Compared to appropriate control values or in relation to starting cell number, the increase in SMC number was constant regardless of the extent of reduction in oxygen content (5.0% or 2.5% 02) and regardless of the duration of exposure (24 or 72 h). In agreement with other investigators, we were unable to demonstrate any significant increase in PVSMC number upon exposure to reduced levels of oxygen tension alone (1,8). However, we did find a small (14%) but statistically significant increase in SMC number upon co-culture in the presence of endothelial cells after 24 h incubation in standard tissue culture 02 condition of 95% room air. This mitogenic effect was not as evident after 72 h incubation but still supports the concept of the spontaneous basal release under nonmanipulated conditions of a SMC growth factor by the endothelial cells. However, the release of this soluble factor and/or additional factors seems to be enhanced upon exposure to reduce 02 conditions and perceptible at both the 24 and 72-h incubations. This effect occurred in the absence of any grossly visible morphologic changes in the endothelial ceils under phase contrast microscopy. This result contrasts with studies demonstrating enhanced production of PDGF by cultured endothelial ceils only when visibly damaged by incubation with endotoxin and phorbol esters (5). Thus it is unlikely that the observed increased mitogenic activity derived from endothelial cells upon exposure to reduced levels of atmospheric 02 represents a nonspecific reaction associated with cell injury or death, but rather a specific inducible process elicited by reduced 02 alone. The observed difference in SMC number through co-cultivation
HYPOXIC ENDOTHELIAL-SMOOTH MUSCLE INTERACTION LU
zuJ I-
ROLE O F E N D O T H E L I A L OF CULTURED
CELLS IN THE PROLIFERATIVE RESPONSE PULMONARY VASCULAR SMOOTH MUSCLE CELLS T O REDUCED OXYGEN T E N S I O N ROBERT L. VENDER
Department of Medicine, Division of Pubnonary a~d Critical Care Medicine, The Medical Collegeof Pennsylvania, 3300 Henry Avenue, Philadelphia, Pennsylvania 19129 (Received 21 November 1991; accepted 17 December 1991)
SUMMARY The development of pulmonary hypertension in a wide variety of human disease states and experimental animal models characterized by chronic alveolar hypoxia is mediated by two pathologic vascular processes, a) vasoconstriction and b) vasoconstruction (structural remodeling). The anatomic changes seen within the pulmonary circulation include a) increased deposition of collagen and elastin in the adventitial layer and b) aberrant pulmonary vascular smooth muscle cell proliferation and maturation in the medial segments. Despite the demonstrated ability of pharmacologic manipulation in the experimental animal to ameliorate both the structural and hemodynamic changes, the actual etiologic mechanisms are only beginning to be explored. Using tl~e cell culture technique of co-cultivation, we have investigated the potential role of bovine pulmonary arterial endothelial cell-derived factors in mediating abnormal bovine smooth muscle cell growth under conditions of reduced oxygen tension. We have demonstrated that these cultured endothelial cells exposed in vitro to reduced levels of atmospheric oxygen concentrations of 5.0% and 2.5% 02 for durations of 24 to 72 h produce and secrete soluble growth factor(s) which stimulate smooth muscle cell proliferation when compared to cells maintained under standard tissue culture oxygen conditions of 95% room air. This growth-stimulatory effect required the concomitant presence of serum factors (0.5% fetal bovine serum), was inhibited by heparin, was distinct from platelet-derived growth factor, and seemed to have a molecular weight greater than 14 000 Da. We conclude that reduced levels of oxygen tension in vitro can selectively induce pulmonary arterial endothelial cells to release mitogen(s) which can stimulate vascular smooth muscle replication. Furthermore, we speculate that this in vitro finding may be of importance as an etiologic mechanism to explain the accelerated smooth muscle cell growth characteristic of hypoxic pulmonary arteriopathy.
Key words: hypoxia/endothelial cell; derived mitogens. INTRODUCTION The development of structural changes within the puhnonary circulation under conditions of chronic generalized alveolar hypoxia is an established pathologic entity that occurs in a wide variety of human disease states and experimental animal models which eventuate in the hemodynamic complication of pulmonary hypertension. These alterations are thought to be of an irreversible nature and are theorized to herald late or end stage disease sequelae. These characteristics are in contrast to the vasoconstrictive response of the pulmonary artery to hypoxia which represents a potentially reversible physiologic process regardless of disease duration or severity. Under conditions of chronic alveolar hypoxia, the complexity of this pathologic structural remodeling has been demonstrated throughout the entirety of the pulmonary circulation and throughout all the structural layers of the arterial wall. These changes include: a) the appearance of longitudinally oriented smooth muscle cells within the intima, b) the abnormal deposition of increased amounts of collagen and elastin within the adventitia, c) medial smooth muscle cell hyperplasia and hypertrophy, and d) the differentiation of poorly developed precursor smooth muscle cells (termed intermediate cells and pericytes) into mature smooth muscle cells (9,15,19,23). We interpret these latter two findings to represent
alterations in smooth muscle cell growth. Although pharmacologic manipulation has been shown to ameliorate some of these anatomic changes in experimental animal models, the actual etiologic factors have yet to be identified (7,17). Given the known intimate anatomic and physiologic association between endothelial cells and smooth cells within the pulmonary circulation and recognizing the importance of peptide growth factors in the control of smooth muscle cell proliferation, we hypothesized that soluble endothelial cell-derived mitogens, specifically released under conditions of reduced oxygen tension, might be of potential etiologic significance in explaining one aspect of these structural changes, namely, aberrant smooth muscle cell replication. Using the cell culture technique of co-cultivation to study this aspect of endothelial-smooth muscle cell interaction we now present data that suggests the selective release under in vitro conditions of reduced oxygen content of a soluble endothelial cell-derived mediator that acts in the presence of serum components as a mitogen for cultured pulmonary vascular smooth muscle cell proliferation. MATERIALSANDMETHODS Tissue culture medium 199, minimal essential medium (MEM), penicillin-streptomycin, and trypsin were obtained from GIBCO, Grand Island, 403
404
VENDER
NY; fetal bovine serum (FBS) from Hyclone, Logan, UT; Miilicell-HA culture plate inserts from Millipore Corp., Bedford, MA; Cytodex-3 microcarrier beads from Pharmacia, Inc., Piscataway, NJ; bovine insulin, bovine heparin, HEPES, deoxycytidine, and thymidine from Sigma Chemical Co., St. Louis, MO; antibodies directed against human platelet-derived growth factor (PDGF) and bovine fibroblast growth factor (basic) from R&D Systems, Minneapolis, MN. Endothelial cells. Under sterile conditions, endothelial cells were recovered from the large extraparenchymal pulmonary arteries of freshly killed calves by gentle scraping with a scalpel blade (21). The recovered endothelial cells were initially grown in 25-cm ~ tissue culture flasks containing 5.0 ml of medium 199 supplemented with 10% heat-inactivated FBS, deoxycytidine (5 #g/ml), thymidine (5 #g/ml), penicillin (100 U/ml), and streptomycin (100 #g/ml) and grown in a humidified atmosphere of 5% COs plus 95% room air at 37 ° C. Once confluent, the endothelial cells were subcultured using 0.25% trypsin and grown again under similar conditions as described above. Endothelial cells were identified by their typical cobblestone morphologic appearance under phase contrast microscopy. For cocultivation experiments, confluent flasks of endothelial cells were removed with 0.25% trypsin and transferred to siliconized 250-ml magnetic stirring vessels containing sterile prehydrated collagen-coated, dextran-based microcarrier beads (Cytodex-3) suspended in 100 ml of medium 199 supplemented with 10% FBS, deoxycytidine, thymidine, penicillin, streptomycin, and 25 mM HEPES (N-[2-hydoxyethyl] piperazine-N [2-ethane sulfonic acid]). This entire suspension system was then placed in a standard tissue culture incubator containing 5% COs plus 95% room air at 37 ° C for an initial 2 to 4 h without stirring and with the caps to the side arms of the stirring vessel loosened. After pre-incubation, the caps were tightened, and the sealed vessel transferred to a nonhumidified room air oven at 37 ° C and stirred continuously at a speed of 15 rpm. The suspension growth media was replenished twice weekly by removal of one-half of the vessel volume, followed by replacement of an equal volume of fresh media containing identical supplements. At each refeeding, the entire sequence of events was repeated. On a weekly basis, serial aliquots of the suspension media containing the microcarrier-borne endothelial cells were removed, the beads inspected for confluency, and the endothelial ceils counted manually by hemocytometry after detachment from the beads with 0.25% trypsin. Additional microcarrier beads were added to the system as needed to achieve endothelial cells in appropriate numbers as required for the particular experimental condition. Endothelial cells from Passages 2 to 10 were used for study. Smooth muscle cells. After the recovery of endothelial cells, the entire intimal surface of the vessel was vigorously scraped with a sterile scalpel blade to remove any remaining adherent endothelial cells. The vessel surface was then washed extensively with phosphate buffered saline (PBS) to remove the detached cells and debris. The vessel was then inverted and the adventitial layer peeled off. This surface was also extensively washed with PBS and the remaining specimen transferred to a sterile petri dish with the intimal surface facing upward. Two millimeter explants of the medial segment were then obtained using standard dermatologic punch biopsy instrument and placed into 35-mm diameter, 6-well tissue culture plates containing minimal essential medium plus 24 mM sodium bicarbonate (MEM) supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 #g/ml). The explants were then placed in a fully humidified tissue culture incubator at 37 ° C in an atmosphere of 5% COs plus 95% room air (14). Fresh medium was added dally until smooth muscle cells appeared in sufficient density for passage. At this time, the smooth muscle cells were detached with 0.25% trypsin and subcultured into 25 cm 2 tissue culture flasks containing 5.0 ml of MEM plus 10% FBS and antibiotics and grown at 37 ° C in an atmosphere of 5% CO2:95% room air. Smooth muscle cells were identified by morphologic appearance under phase contrast microscopy showing the characteristic formation of hillocks and valleys postconfluence (20). Smooth muscle cells from Passages 2 to 10 were used for study. Smooth muscle cell growth bioa.~say. Pulmonary vascular smooth muscle cells were initially plated at a constant density into 35-mm diameter, 6-well tissue culture plates. This initial plating density between experiments ranged from 1.0 to 7.0 × 103 cells per well (100 to 700 cells/cm2). These cells were initially incubated in 2.0 ml per well of MEM supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 pg/ml) in an atmosphere of 5% COs plus 95% room air at 37 ° C for 72 to 96 h. This
period of time was chosen to allow the ceils to become attached and stabilized after exposure to trypsin. The original medium was then removed, the cell layer washed twice with serum-free medium and refed with 2.0 ml per well of serum-supplemented medium 199:MEM ( h 1). Random wells were selected and cell number assayed by hemocytometry before experimental manipulations to determine the baseline starting cell number values. Subsequent data will be analyzed in relation to this starting cell density or as a ratio of the negative control cell number value after each specified incubation duration. At the same time, adequate numbers of endothelial cells grown to confluence on microcarrier beads were removed from the suspension apparatus, placed into a 50-ml centrifuge tube, allowed to sediment, and the supernatant aspirated. These endothelial cells were then washed twice with a 10-fold volume excess of serum-free medium 199:MEM ( h l ) , so as to dilute out the high concentrations of FBS in the original suspension medium, and then resuspended in the necessary volume of medium 199:MEM plus specified concentrations of FBS, as required for the particular experimental situation. Mierocarrier-borue endothelial cells ( 1.0 X 106 cells per well) were then added to the inside surface of individual 12-mm diameter culture plate inserts (Millicell-HA:Millipore Corp.) which were previously placed within the 6-well culture plates containing the smooth muscle cells. The entire apparatus was then incubated at 37 ° C for durations of 24 or 72 h in an atmosphere of either a) 5% CO2 plus 95% room air (approximate 02 concentration of 18 to 20%); b) 5% CO2 plus 5.0% 02, balance Nz; or c) 5% COs plus 2.5% 03, balance N2. After each 24-h period, the medium was removed and the cells refed with fresh medium identical to the initial composition. After the desired period and conditions of exposure, the inserts were removed, the adherent smooth muscle cells detached by two consecutive 10-min incubations with 0.25% trypsin, and counted manually by hemocytometry. Pulmonary vascular smooth muscle cells incubated for similar durations at 37 ° C in the absence of inserts, beads, and endothelial cells in medium 199:MEM plus 0.5% FBS in 5% CO2 plus 95% room air served as negative controls. Pulmonary vascular smooth muscle cells incubated for similar durations at 37 ° C in the absence of inserts, beads, and endothelial cells in medium 199:MEM supplemented with 10% FBS in 5% COs plus 95% room air served as positive controls. To determine the modulating effect of bovine heparin on any demonstrated smooth muscle cell proliferation, additional experiments were carried out in the presence of 1.5 to 150 gg/ml heparin for the entire duration of the co-cultivation period (2). In addition, identical experiments were performed in the continuous presence of blocking antibodies directed against humanPDGF (40 ttg/ml) and bovine FGF-b (40 #g/ml). For molecular sizing studies, spectrapor dialysis tubing of a specified exclusion limit was adhered to the inside of the inserts using a nontoxic sealant, and similar growth assays performed. Oxygen exposure conditions. The "normoxic" tissue culture exposure conditions of 5% COs plus 95% room air were accomplished using a standard tissue culture incubator, fully humidified at 37 ° C. The reduced oxygen exposure conditions were performed using a Forma Scientific (model 3159) tissue culture incubator with three gas capability, employing an oxygen sensor which automatically adjusts the inflow of COs and N2 into the incubator within an accuracy of 0.5% of the preset atmospheric condition, namely 5% COs plus 5% 02, balance N2 and 5% COs plus 2.5% O2, balance N2. Periodically, medium was sampled anaerobically into a 1.0-ml syringe and analyzed immediately with Corning 170 pH/blood gas analyzer for measurement of oxygen partial pressure (Po2). Statistical analysis. All values are shown as means + SEM. Experimental data are expressed as a ratio in comparison to appropriately matched control values as indicated. A one sample t test was used for statistical analysis comparing mean values of these ratios to control value of 1.0. A paired-sample t test was used to make comparisons for the growth modulation, molecular sizing, and endothelial cell dilution studies. P < 0.05 was assumed to represent statistically significant differences (3).
RESULTS The measured dissolved oxygen partial pressure (Po2) as well as the calculated oxygen content (based on an oxygen solubility o f 0 . 0 2 4 4 ml 02 per 1.0 ml H20 at 1.0 atm) of the experimental medium after the time durations of 14 and 2 4 h in atmospheric oxygen concentration of either 9 5 % room air, 5 . 0 % 02 or 2 . 5 % 02
405
HYPOXIC ENDOTHELIAL-SMOOTH MUSCLE INTERACTION TABLE 1 MEASUREMENTS OF TISSUE CULTURE MEDIA OXYGEN CONCENTRATIONS Atmospheric Exposure Conditions
14 Incubation duration, h Range, mmHg Partialpressure (mmHg)~ Content, ml O2/100 ml media~ 24 Incubation duration, h Range, mmHg Partial pressure (mmHg)° Content, ml O J 1 0 0 ml media a
95% Room Air
5.0% Oxygen
2.5% Oxygen
n = 4 144-160 151.9 + 4.4 0.49 + 0.01 n= 4 146-156 150.5 + 2.6 0.48 _+ 0.01
n = 2 71-72 71.7 ± 0.7 0.23 _-+0.01 n = 3 53-60 55.8 + 2.0 0.18 + 0.01
n = 1 49.5 49.5 0.16 n= 4 38-50 46.8 ± 2.5 0.15 ± 0.01
a Mean + SEM.
are shown in Table 1. As can be seen, the reduced atmospheric oxygen conditions of 2.5% and 5.0% O 2 for 24 h resulted in an overall mean Po~ of 50.7 mmHg. Twenty-one separate experimental sets of data were obtained to generate the results recorded in the accompanying tables and figures. The values for the starting cell number (recorded as mean + SEM) ranged from'0.80 + 0.12 to 15.75 + 0.24 X 104 cells. Despite variability in starting cell numbers between experiments, within each individual study the range of values was quite close and in itself could not account for the demonstrated changes in cell number that we observed under the various experimental conditions. Under the negative control conditions of 0.5% FBS, 95% room air oxygen concentration and the absence of endothelial cells, we observed only an 18% increase in smooth muscle cell number after 72 h incubation when compared to starting cell number. Conversely, under the positive control conditions of 10% FBS, 95% room air oxygen concentrations, and the absence of endothelial cells there was a statistically significant increase in all smooth muscle cell measurements at both the 24- and 72-h incubation periods, regardless of whether these values were compared to either matched negative control values for cell number at these same time points or to starting cell numbers (Figs. 1 and 2). This increase in smooth muscle cetl number above starting values after supplementation with 10% FBS amounted to a near doubling (1.89 +_ 0.09) after 24 h and a 5.54 -+ 0.39-fold increase at the conclusion of the 72-h protocol. In separate experiments, where microearrier beads without endothelial cells (naked heads) were handled in a manner identical to studies employing microearrier-borne endothelial cells, no significant difference was observed in smooth muscle cell number between cells grown in 0.5% FBS without the concomitant presence of inserts and naked microcarrier beads compared to cells maintained in 0.5% FBS in the presence of these devices in oxygen concentrations of 95% room air and 5.0% 02. When data were expressed as a ratio of smooth muscle cell number with beads alone divided by cell number without beads after 24 and 72 h incubation, respectively, these values measured 1.07 + 0.05 and 1.01 + 0.24 for the 95% room air conditions and 1.18 +_-0.08 and 0.87 + 0.06 for the 5.0% O2 exposure conditions. The addition of 1.0 X 106 pulmonary arterial endothelial cells to the smooth muscle cells through co-cultivation in 0.5% FBS and
95% room air oxygen concentration elicited a significant increase in smooth muscle cell number after the 24 h incutJation when compared to both negative control values at this same time point and to starting cell number values (1.14 + 0.04 and 1.16 + 0.05, respectively). However, this mitogenic effect of endothelial cells on smooth muscle cell proliferation in the standard 95% room air conditions was not evident after 72 h incubation (Figs. 1 and 2). The discrepancy between the observed increase in smooth muscle cell number at 24 h and its absence at 72 h probably resulted from the previously noted 18% increase in cell number achieved by the addition of 0.5% FBS alone in our negative control conditions, which effectively masked any further increase in smooth muscle cell number achieved by the concomitant addition of endothelial cells in co-culture. However, this data do suggest the spontaneous release of an endothelial cell-derived growth factor into the co-culture media under standard tissue culture conditions of 95% room air:5% CO2, which stimulates smooth muscle cell proliferation of a minor degree. No consistent increase in pulmonary vascular smooth muscle
2.0 O }z
*t
*t
O0
r_.o *t 1.0
z3
20% 02
2,5 - 5.0"/* 0:~
24 HOUR INCUBATION OURATION
El zo% 02
2,s - s.o% o~
7 2 HOUR INCUBATION OUP.ATtOH
FIG. 1. Effect of reduced atmospheric oxygen concentration upon cultured PVSMC proliferation in the absence (-) or presence (+) of PAECs through co-cultivation in 0.5% FBS conditions. Data are expressed as the PVSMC number ratio experimental/negative control and recorded as mean - SEM. Corresponding values in the presence of 10% FBS for 24 and 72 h measured 2.09 ± 0.09 and 4.59 + 0.41, respectively. *P < 0.05 compared to 1.00; ~P < 0.05 compared to cell number ratio in the absence of PAEC under similar conditions of oxygen exposure.
406
VENDER
2.0 A
|
1.0 Z-I
24 HOURIr~ICUB~TIONDU~TI~I
72 HOURINCUSATIC~ DU~IlON
FIG. 2. Effect of reduced atmospheric oxygenconcentrationupon cultured PVSMC proliferationin the absence (-) or presence (+) of PAEC through co-cultivationin 0.5% FBS conditions.Data are expressed as the PVSMC ratio experimental/startingcell number and recorded as mean + SEM.Correspondingvaluesin the presenceof 10% FBS for 24 and 72 h measured 1.89 + 0.09 and 5.54 + 0.39 respectively. *P < 0.05 compared to 1.00; "~P< 0.05 compared to cell number ratio in the absence of PAEC under similarconditionsof oxygenexposure.
cells (PVSMC) number was elicited upon exposure to reduced levels of atmospheric oxygen alone, i.e. in the absence of pulmonary arterial endothelial ceils (PAEC). However, the co-cuhivation of pulmonary arterial endothelial cells at a concentration of 1.0 × 106 cells per well with pulmonary vascular muscle cells in 0.5% FBS and their subsequent exposure to reduced levels of oxygen tension (2.5% to 5.0% 02) for both 24 and 72-h incubationsresulted in a reproducible, statistically significant increase in smooth muscle cell number above starting values and when compared to the negative control value at the same points in time (Figs. 1 and 2). This increase in smooth muscle cell number was remarkably constant under all experimental conditions. When expressed in Fig. 1 as a ratio to negative control values of cell number at the same time points, these increases measured 1.45 + 0.09 at 24 h and 1.43 + 0.07 at 72 h. In addition, both the endothelial and smooth muscle cells seemed morphologically normal by phase contrast microscopy under all the above-mentioned culture conditions in 0.5% FBS. An increase in smooth muscle cell number of greater than 20% above negative control values upon co-cultivation with endothelial cells under reduced oxygen conditions was seen in 28/37 (76%) of studies after a 24-h incubation. However, by extending our assay period an additional 48 h to a total incubation of 72 h, an increase in smooth muscle cell number of greater than 20% above negative control values was demonstrated in 11/12 (92%) of experiments. All previously reported data utilized PAEC at a constant density of 1.0 X 106 cells per well. We also performed studies using variable concentrations of endothelial cells to assess a) the minimal number of cells required to elicit a mitogenic effect in our co-cultivation system and b) the differentiation of potential growth inhibitory substances by dilution analysis. Consistent with the previously reported data, PAEC at a concentration of 1.0 )< 106 cells per well effected a statistically significant increase in PVSMC number upon exposure to 5% 02 compared to matched control cells grown in
absence of endothelial ceils. No such increase was seen under 20% 02 conditions. This difference in mitogenic response between the two levels of atmosphere oxygen exposure was also observed at endothelial cell densities of 1.0 )< 104 and 1.0 × 1.03 cells per well (Fig. 3). A growth stimulatory effect was not observed at the lowest endothelial cell density of 1.0 × 102 cell per well. Finally, the concomitant addition of heparin at dosages of 1.5, 15, and 150 #g/ml to the co-cultivation bioassay system significantly inhibited the previously observed increase in SMC number elicited in the presence of endothelial cells and a reduced oxygen level of 5.0% 02. The degree of inhibition was 62 -+ 2.0%, 93 + 4.7%, 59 + 6.6% for the various heparin concentrations, respectively (Table 2). In addition, the demonstrated increase in PVSMC number was not inhibited by the concomitant presence of blocking antibodies directed against PDGF or FGFb, and was specifically excluded by dialysis tubing with a molecular weight cut off less than 14 000 Da (Table 2). However, it should be noted that the addition of anti-FGFb immunoglobulin to our co-cultivation system did cause a mild degree of PVSMC growth inhibition, although this value did not achieve statistical significance. DISCUSSION In this study, utilizing the cell culture technique of co-cultivation, we report the selective induction and release from confluent cultures of PAEC of a soluble mitogen(s) stimulatory for cultured PVSMC proliferation upon in vitro exposure to reduced levels of atmospheric oxygen concentration of 5.0% and 2.5% 02, when compared to cells maintained under standard tissue culture oxygen conditions (5% CO2:95% room air). This SMC growth stimulatory effect was dependent on the presence of serum supplementation (0.5% FBS) and not observed under serum-free conditions. Compared to appropriate control values or in relation to starting cell number, the increase in SMC number was constant regardless of the extent of reduction in oxygen content (5.0% or 2.5% 02) and regardless of the duration of exposure (24 or 72 h). In agreement with other investigators, we were unable to demonstrate any significant increase in PVSMC number upon exposure to reduced levels of oxygen tension alone (1,8). However, we did find a small (14%) but statistically significant increase in SMC number upon co-culture in the presence of endothelial cells after 24 h incubation in standard tissue culture 02 condition of 95% room air. This mitogenic effect was not as evident after 72 h incubation but still supports the concept of the spontaneous basal release under nonmanipulated conditions of a SMC growth factor by the endothelial cells. However, the release of this soluble factor and/or additional factors seems to be enhanced upon exposure to reduce 02 conditions and perceptible at both the 24 and 72-h incubations. This effect occurred in the absence of any grossly visible morphologic changes in the endothelial ceils under phase contrast microscopy. This result contrasts with studies demonstrating enhanced production of PDGF by cultured endothelial ceils only when visibly damaged by incubation with endotoxin and phorbol esters (5). Thus it is unlikely that the observed increased mitogenic activity derived from endothelial cells upon exposure to reduced levels of atmospheric 02 represents a nonspecific reaction associated with cell injury or death, but rather a specific inducible process elicited by reduced 02 alone. The observed difference in SMC number through co-cultivation
HYPOXIC ENDOTHELIAL-SMOOTH MUSCLE INTERACTION LU
zuJ I-
