Human umbilical venous endothelial cells (HUVECs) were purchased from Lonza (Walkersville, MD) and cultivated in EGM-2 BulletKit supplemented with 2% FBS (growth medium) at 37°C in a humidified incubator saturated with a gas mixture containing 21% O₂ and 5% CO₂ in N₂ (normocapnic normoxia) for up to 5 passages. The initial pH of the EGM-2 medium saturated with 21% O₂ and 5% CO₂ at 37°C was 7.3; the medium was used after the pH was adjusted to 7.4 by the addition of 0.1N NaOH solution. The exposures of the cells to hypoxia and/or hypercapnia were carried out by incubation of the HUVECs in humidified atmospheres containing 5% CO₂ and 1% O₂ (normocapnic hypoxia), 10% CO₂ and 1% O₂ (hypercapnic hypoxia), and 10% CO₂ and 21% O₂ (hypercapnic normoxia), with the balance composed of N₂ in each case, at 37^oC. The pH of the medium saturated with 10% CO₂ and either 1% O₂ or 21% O₂ was 7.1, modeling acute hypercapnia. In some experiments modeling chronic hypercapnia, 0.1N NaOH solution was added to the medium saturated with 10% CO₂ and either 1% O₂ or 21% O₂, to obtain a final pH of 7.4. In some additional experiments modeling metabolic acidosis, 0.5N HCl solution was added to the medium saturated with 5% CO₂ and either 1% O₂ or 21% O₂, to obtain a final pH of 7.1. The medium was presaturated by incubation in the appropriate gas mixture overnight, and then added to the cell cultures, as indicated, in each experiment.

HUVECs were grown to confluence in 8-chamber tissue culture dishes. Thereafter, the monolayers were wounded in a straight line using a sterile 200 μL tip and washed twice with phosphate-buffered saline (PBS). The monolayers were quickly added to the presaturated medium and the incubation was continued in the appropriate gas mixture until 48 hours post injury. The progress of wound healing was monitored with a phase-contrast Olympus BX60 microscope (Olympus Optical Co., Ltd., Tokyo, Japan) equipped with a digital camera. Each wound was photographed at specified time-points, and the area of the wound was measured with an image analysis software (Win Roof Version 3.5; Mitani Corporation, Fukui, Japan) on a Microsoft XP computer. The dishes were returned to the incubator between the measurements. Four measurements were taken from four fields of each well in at least three independent experiments.

HUVECs at subconfluence were trypsinized, resuspended in the medium presaturated with each of the appropriate gas mixtures, and seeded on to 96-well dishes. After 1 hour of incubation in the same gas mixture condition, non-attached cells were removed by a gentle wash with PBS. Then, the attached cells were lysed in 100 µl of PBS containing 0.2% Triton-X, in which the total protein concentration was measured by the Protein DC assay (Bio-rad Laboratories, Tokyo, Japan) according to the manufacturer’s instructions. The adhesion was quantified by dividing the protein concentration in the attached cell lysate by the protein concentration in the total cell lysate.

Cell migration was examined by the Boyden chamber technique using a 48-well chamber (Neuroprobe, Bethesda, MD). Polycarbonate membranes with 8-µm pores (Neuroprobe) were coated with 1% collagen. Vascular endothelial growth factor (VEGF; 500 ng/ml) was used in the bottom wells as an attractant. HUVECs at subconfluence were trypsinized, resuspended in in the presaturated medium, and plated on to each of the top wells above the filter. The chambers were then incubated for 6 hours under each gas mixture condition. After incubation, cells on the top of the filter were removed by scraping. The filter was then stained with a Diff-Quick stain. Endothelial cell migration activity was quantified as the number of migrated cells on the lower surface of the filter in 10 high-power fields using a light microscope at x 1000 magnification.

Proliferating HUVECs were identified by immunocytochemistry for the detection of bromodeoxyuridine (BrdU) incorporation into the cellular DNA. HUVECs at 50% confluence were incubated at different O₂, CO₂, and pH conditions for different periods of time. Cells were loaded with 30 µM BrdU for the final 1 hour and then fixed for immunostaining to detect cell proliferation as described below. Cell proliferation was quantified by counting BrdU-incorporated cells under a light microscope at x400 magnification.

HUVECs were fixed with 3% paraformaldehyde and permeabilized with 0.5% Triton X-100. For anti-BrdU staining, cells were exposed to 2N HCl for 20 minutes to denature DNA for easy access of the antibody to antigen. Nonspecific binding was blocked with a blocking solution containing 3% BSA, followed by incubation with mouse monoclonal anti-BrdU antibody (Merck Milipore Co., Billerica, MA) or rabbit polyclonal anti-active caspase-3 antibody (Cell signaling Technology Japan, Tokyo, Japan). The primary antibodies were reacted with the secondary antibodies conjugated with horseradish peroxidase polymer (EnVision+ kit; Dako Japan, Tokyo, Japan). The immunoreactants were detected with a 3,3’-diaminobenzidine solution.

TdT-mediated dUTP nick-end labeling (TUNEL) was performed using the Takara In Situ Apoptosis Detection Kit (Takara Biomedicals, Tokyo, Japan) according to the manufacture’s instructions.

The statistical analyses were performed using the Excel X software program with the add-in software Statcel 3 (OMS, Tokyo, Japan). Data are expressed as mean ± SEM. Statistical differences were analyzed by analysis of variance (ANOVA), and if the results were significant, the Tukey-Kramer test was used as a multiple comparison post hoc test. P values of <0.05 were considered to indicate statistical significance.

A reduced rate of wound healing was observed in HUVEC monolayers at 24 and 48 hours after injury under normocapnic hypoxia (5% CO₂, 1% O₂, pH 7.4) as compared to that under the normocapnic normoxic controls (5% CO₂, 21% O₂, pH 7.4) (Fig. 1). However, hypercapnic acidosis (10% CO₂, 1% O₂, pH 7.1), modeling acute hypercapnia, improved the rate of wound healing in the hypoxic HUVEC monolayers (Fig. 1B, C). Buffered hypercapnia (10% CO₂, 1% O₂, pH 7.4), modeling chronic hypercapnia, also improved the rate of wound healing in the hypoxic HUVEC monolayers (Fig. 1B, C). The degree of improvement in the rate of wound healing under buffered hypercapnia was not different from that seen under hypercapnic acidosis. In contrast, normocapnic acidosis (5% CO₂, 1% O₂, pH 7.1), modeling metabolic acidosis, did not alter the rate of wound healing in hypoxic cultures of HUVEC monolayers (Fig. 1C). Hypercapnia did not stimulate wound healing of the HUVEC monolayers under normoxic conditions (Fig. 1D). These findings suggest that hypercapnia rather than acidosis per se accelerates the wound healing in HUVEC monolayers under hypoxic conditions.

Fig. (1). Hypercapnia accelerates wound healing in HUVEC monolayers exposed to hypoxia. Cells were grown to confluence in 8-chamber tissue culture dishes before wounding. (A) Representative photomicrographs of cell monolayers 48 hours after wounding. Broken lines indicate the leading edge of the wound repopulating cells. Scale bars = 200 µm. (B) Time course of changes in the size of the remaining wound. The data points represent the means ± SEM of 4 wells. **P < 0.01 vs cells cultured at 21% O2/5% CO2, pH 7.4; 1% O2/10% CO2, pH 7.1 and 1% O2/10% CO2 buffered to pH 7.4. (C) Effects of hypoxia, hypercapnia and acidosis on wound healing. Wounded cell monolayers were cultured under various conditions with different O2 (21% or 1%), CO2 (5% or 10%), and pH (7.4 or 7.1) levels. The areas of the wounds were measured at 24 hours. Wounded cell monolayers were also cultured at 1% O2/5% CO2, pH 7.4 in conditioned medium (CM) from cell cultures exposed to 1% O2/10% CO2, pH 7.1 for 48 hours. **P < 0.01 as compared to all other conditions. (D) Effect of hypercapnia on wound healing in HUVEC monolayers cultured under normoxia at 24 hours.

Fig. (2). Hypercapnia does not stimulate cell attachment (A) or migration (B) of HUVECs exposed to hypoxia. Cells were grown to subconfluence before trypsinization and replated on to 96-well dishes for the 1-hour attachment assay (A) or used in the Boyden blind well chamber migration assay for 6 hours, during which VEGF (500 ng/ml) was used as an attractant (B). Both assays were performed under different cell culture conditions as specified. The bars represent the means ± SEM of 6 wells (A) or 3 wells (B) **P < 0.01 as compared to control cells incubated under 21% O2 and 5% CO2, pH 7.4.

Fig. (3). Hypercapnia stimulates proliferation of HUVECs exposed to hypoxia. Cells were grown to 50% confluence and then incubated under various conditions with different O2 (21% or 1%), CO2 (5% or 10%) and pH (7.4 or 7.1) levels for 24 or 48 hours. BrdU was added to the medium 1 hour before paraformaldehyde fixation. (A) Representative photomicrographs of anti-BrdU immunocytochemistry for HUVECs fixed at 24 hours. Distinct dark spots represent cell nuclei with BrdU incorporation. Scale bars = 200 µm. (B) Effect of hypercapnia on cell proliferation. Hypercapnic acidosis and buffered hypercapnia stimulate cell proliferation under hypoxic conditions. Cell proliferation at 24 and 48 hours was quantified as the number of BrdU+ cells/mm2 and data are expressed as percentages of values obtained from the normocapnic normoxic controls (21% O2/5% CO2, pH 7.4) at the corresponding time-points. The bars represent the means ± SEM of 4 wells. **P < 0.01 as compared to cells incubated at 1% O2/5% CO2, pH 7.4. (C) Hypercapnic acidosis but not noromocapnic acidosis stimulates cell proliferation under hypoxic conditions. Data are expressed as the number of BrdU+ cells/mm2. The bars represent the means ± SEM of 4 wells. **P < 0.01 as compared to cells incubated under conditions of 1% O2/5% CO2 with pH 7.4 or pH 7.1. (D) Hypercapnia did not stimulate cell proliferation under noromoxic conditions. Data are expressed as the number of BrdU+ cells/mm2. The bars represent the means ± SEM of 4 wells.

Fig. (4). Hypercapnia does not protect against hypoxia-induced apoptosis of HUVECs. Cells were grown to 50% confluence and incubated under conditions with different O2 (21% or 1%), CO2 (5% or 10%), and pH (7.4 or 7.1) levels for 48 hours. Apoptosis was assessed by TUNEL staining and anti-active-caspase-3 immunocytochemistry. (A) Representative photomicrographs of TUNEL and anti-active-caspase- 3 immunocytochemistry. Scale bars = 200 µm. (B) Quantification of TUNEL+ cells. (C) Quantification of anti-active-caspase-3+ cells. The bars represent the means ± SEM of 4 wells. **P < 0.01 as compared to cells incubated at 21% O2/5% CO2, pH 7.4.

The rate of wound healing in HUVEC monolayers cultured under normocapnic hypoxia was not altered by the conditioned medium from HUVEC monolayers cultured under hypercapnic hypoxia for 24 hours (Fig. 1C). This finding suggests that the stimulatory effect of hypercapnia on wound healing was not due to the secretion of a soluble factor from the hypercapnia-exposed HUVECs.

Wound healing of cell monolayers involves complex processes, including adhesion, migration, and proliferation. Adhesion and migration of HUVECs was inhibited under normocapnic hypoxic conditions as compared with that in the normocapnic normoxic controls (Fig. 2). Neither hypercapnic acidosis nor buffered hypercapnia altered the rates of adhesion or migration of HUVECs exposed to hypoxia. These results suggest that the acceleration of wound healing by hypercapnia under hypoxic conditions was not due to stimulation of cell adhesion or migration.

The effect of hypercapnia on the proliferation of HUVECs was determined by immunocytochemistry for the detection of BrdU incorporation into the cellular DNA. Normocapnic hypoxia inhibited the proliferation of HUVECs as compared with the proliferative activity observed in the normocapnic normoxic controls (Fig. 3). Both hypercapnic acidosis and buffered hypercapnia stimulated proliferation of HUVECs exposed to hypoxia (Fig. 3B). In contrast, normocapnic acidosis did not alter the rate of cell proliferation observed under hypoxic conditions (Fig. 3C). Neither hypercapnic acidosis nor buffered hypercapnia altered the rate of cell proliferation observed under noromoxic conditions (Fig. 3D). These results suggest that hypercapnia stimulates the proliferation of HUVECs exposed to hypoxia.

Another potential mechanism underlying the acceleration of wound healing by hypercapnia was protection of HUVECs from hypoxia-induced apoptosis. Incubation of HUVECs under normocapnic hypoxia for 48 hours increased the rate of apoptosis detected by TUNEL and active caspase-3 immunostaining as compared with that observed in the normocapnic normoxic controls (Fig. 4). Neither hypercapnic acidosis nor buffered hypercapnia altered the rate of apoptosis, suggesting that the effect of hypercapnia on wound healing was not due to modulation of apoptosis.