NITRIC OXIDE
Biology and Chemistry
Nitric Oxide 12 (2005) 129–140 www.elsevier.com/locate/yniox
A direct nitric oxide gas delivery system for bacterial and mammalian cell cultures A. Ghaffari a,*, D.H. Neil b, A. Ardakani d, J. Road c, A. Ghahary a, C.C. Miller c a
c
Department of Surgery, Wound Healing Research Group, University of Alberta, Edmonton, Alta., Canada b Laboratory Animal Science, University of Alberta, Edmonton, Alta., Canada Department of Pulmonary and Infectious Disease, Department of Medicine, University of British Columbia, Vancouver, BC, Canada d Pulmonox Medical Inc., Tofield, Alta., Canada Received 21 May 2004; revised 1 December 2004
Abstract Nitric oxide (NO) is the smallest known gaseous signaling molecule released by mammalian and plant cells. To investigate the pathophysiologic role of exogenous NO gas (gNO) in bacterial and mammalian cell cultures, a validated in vitro delivery method is required. The system should be able to deliver gNO directly to bacterial and/or cell cultures in a continuous, predictable, and reproducible manner over a long period of time (days). To accomplish this, a gas delivery system was designed to provide optimal growth conditions for bacteria and/or mammalian cells. Parameters for cell exposure, such as concentration of gNO, nitrogen dioxide (NO2), oxygen (O2), temperature, and relative humidity (RH) were continuously monitored and evaluated. Uptake of gNO into various media was monitored by measuring the nitrite concentration using the Griess reagent technique. A selection of standard growth media [saline, tryptic soy broth (TSB), Middlebrook 7H9 (MB 7H9), and DulbeccoÕs modified EagleÕs medium (DMEM)] exposed to various concentrations of gNO revealed a steady and consistent transfer of gNO into the aqueous phase over a 48-h period. Validation of optimal growth conditions within the device, as compared to a conventional incubator, were accomplished by growing and observing viability of Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and human fibroblast cultures in the absence of gNO. These results indicate that an optimal growth environment for the above tested cells was accomplished inside the proposed delivery system. Dose-dependent toxicological data revealed a significant bacteriostatic effect on P. aeruginosa and S. aureus with continuous exposure to 80 ppm gNO. No toxic effects were observed on dermal fibroblast proliferation at concentrations up to 400 ppm gNO for 48 h. In conclusion, the designed gNO exposure system is capable of supporting cellular viability for a representative range of prokaryote and eukaryotic cells. The exposure system is also capable of obtaining toxicological data. Therefore, the proposed device can be utilized to continuously expose cells to various levels of gNO for up to 72 h to study the in vitro effects of gNO therapy. 2005 Elsevier Inc. All rights reserved. Keywords: Nitric oxide; Cell exposure device; Bacteria; Fibroblasts; Cell viability
Only two decades ago, nitric oxide (NO) was known merely as an air polluting toxic gas found in cigarette smoke and automobile smog. Ironically, in the late 1980s, NO was discovered to be one the smallest, light-
*
Corresponding author. Fax: +1 780 401 3052. E-mail address: abdi.ghaff
[email protected] (A. Ghaffari).
1089-8603/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.niox.2005.01.006
est, and the first gas molecule to act as a biological messenger in mammals. This discovery eventually lead to the Nobel prize in medicine in 1998 [1]. Nitric oxide (NO) is synthesized throughout the body by cells such as endothelial cells [2], macrophages [3], neutrophiles [4], fibroblasts [5], and keratinocytes [6]. As one of the smallest but nevertheless important regulatory molecules in the body, NO is involved in a number of key
130
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
physiological functions such as blood pressure regulation [7], neurotransmission [8], inhibition of platelet adhesion [9], wound healing [10], and nonspecific immune response to infection [11]. The pathophysiology of a number of disorders have been associated with an altered level of NO or expression of its precursor enzymes, nitric oxide synthases (NOS) [12]. Therefore, a better understanding of this ubiquitous molecule and its toxicity at the cellular level is crucial to better explore its potential therapeutic value. Much of the data behind toxicological studies and the biological role of NO is supported by in vitro experiments performed using various NO delivery methods. One of the major obstacles in such studies is to establish a direct and continuous delivery system for exposing various cells to exogenous nitric oxide. Some of the methods described in the literature to deliver NO in cell culture studies include gas permeable membranes, donor compounds, stimulation of NO synthase enzymes (NOSs), and addition of NO-saturated solutions to cellular systems [13–15]. During the application of NO donors, such as spermine NONOate and diethylamine NONOate, stability and consistency of NO release rate is extremely difficult and can be affected by a variety of conditions such as pH of the solution, light, and nature of the aqueous medium. In most cases, NO release rate is not constant, difficult to maintain at a constant level in long-term exposures, and toxicity of residual donor compounds following release of NO is difficult to predict [16]. For instance, NO produced by the stimulation of NOS enzymes can react intracellularly with other nitrogen species generated or oxidants already existing within the cytoplasm, such as superoxide, to form other potentially damaging species like peroxynitrite [17]. These reactions may considerably reduce the availability or even mask the true effects of NO on the cells being studied. With the application of NO-saturated solutions, it is difficult to maintain steady-state NO concentrations over time, as NO levels are continuously depleted with this technique [17]. A more controlled and predictable model for delivery of NO into aqueous solutions was developed by flowing the media on one side of a gas-permeable membrane and then exposing it to NO or NO producing compounds on the other side of the membrane [17,18]. These studies showed that a constant delivery of NO gas led to a steady-state NO concentration in the solution, even in the presence of other species reacting with NO [19]. Although the predicted models in these studies were in agreement with the experimental values and the uptake of NO by aqueous phase was characterized in detail, the design of the exposure chambers did not allow for broad scale use of bacterial and mammalian cells in standard culture plates commonly available in the laboratories. Long et al. [20] and HoehnÕs groups [21] exposed bacterial cells to gNO inside a simple incubation chamber
using a continuous horizontal flow of gNO across the cells. This type of continuous gas flow exposure technique was originally described by Voisin et al. [22] while exposing alveolar macrophages to various toxic gases excluding NO. The gNO incubation chamber described by both Long and Hoehn is not an optimal device for the growth of cell cultures because external variables such as humidity, flow rate, nitrogen dioxide buildup, and oxygen levels cannot be controlled. In recent years, NO has been identified as an antimicrobial agent in the bodyÕs innate and nonspecific line of defense and for its apparent regulatory role in wound healing [10,23]. Infections in humans are often associated with a significant increase in NOS expression and systemic NO production [24] which directly correlates with the hostÕs ability to suppress microbial proliferation and thus contain the infection [25]. Mice lacking the inducible form of NOS, were more susceptible to herpes infection than wild-type counterparts [26]. WellerÕs group had some success in applying an acidified nitrite as a NO releasing compound demonstrating the antimicrobial ability of NO against common cutaneous pathogens [27]. In wound healing regulation, the beneficial role of amino acid arginine supplementation, the substrate for NOS enzymatic activity in NO production, was implicated over 25 years ago in a rat model for incisional repair [28]. The ability of arginine to improve wound healing was later linked to NOS activity in iNOS deficient mice, as loss of NOS function abrogated the beneficial effect of arginine [29]. In a similar animal model for full thickness excisional wound, iNOS knockout mice showed significant delay in wound closure compared with the wild-type control group. More importantly, this delay was reversed by application of adenoviral-mediated expression of human iNOS cDNA at the wound site of the knockout mice [30]. To further investigate the potential therapeutical value of exogenous nitric oxide gas as an antimicrobial agent as well as a mediator in cutaneous wound healing, we report the design of a versatile gNO exposure chamber to test its effect on various cells such as Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and human dermal fibroblasts. The device was optimized for cellular growth conditions, exposure to gNO, and the ability to obtain toxicological data from bacteria and mammalian cells in both solid and aqueous media.
Methods and materials Cell exposure device As shown in Fig. 1, the delivery device consisted of two cylindrical Plexiglas exposure chambers with separate gas entry ports and a common exit port. These
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
chambers were surrounded by an airtight Plexiglas jacket that provided a thermally isolated environment. This jacket housed an electrical heater unit controlled by an internal thermostat (Invensys Appliances Control, Carol Stream, IL), that helped to provide stable temperatures inside the exposure chambers. An electrical fan was placed inside the jacket to circulate the heated air allowing steady-state temperatures to be reached rapidly within the system. A vent, covered by high efficiency particulate air (HEPA) Filter (Sentry Air Systems, Houston, USA) was used to prevent contamination. A hygrometer (VWR Canlab, Mississauga, ON) was used at the exhaust ports to measure relative humidity of the gas mixture exiting the cell exposure chambers. Separate sample lines placed directly over the exposure dishes in each of the two exposure chambers provided samples of the gas mixtures to a nitric oxide/nitrogen dioxide/oxygen electrochemical analyzer (AeroNOx, Pulmonox Medical, Tofield, AB, Canada) to detect the exact composition of each gas in the mixture. The overall apparatus dimensions were 84 · 66 · 44 cm (Fig. 1) and was placed either on a countertop in a Level II biological safety room or in a biological cabinet for the duration of the studies.
131
Gases were supplied from pressurized cylinders. These included 800 parts per million (ppm) medical grade NO, balance nitrogen (ViaNOx-H, Pulmonox Medical, Canada), medical grade air, oxygen, and carbon dioxide (Praxair, Mississauga, ON, Canada) that were delivered at a constant pressure (50 psi) using an appropriate standard Compressed Gas Association (CGA) approved gas regulator. Gases were then mixed together at pre-calculated concentrations using a dilution manifold (Fig. 1A) and a digital mass flowmeter (TSI, Shoreview, MN, USA). The gas manifold allowed for a mixture of up to five different gases at various flow rates. Two identical channels existed in the manifold providing for two different sources of gas mixtures used for either a control chamber or a test treatment chamber. The gas was delivered from the gas manifold through two separate 22 mm (inside diameter) corrugated respiratory tubings, through two humidifiers (MR850, Fisher and Paykel Healthcare, CA, USA). The humidifiers were set to humidify the mixtures up to 90% relative humidity (RH%) using sterile water and a peak temperature setting of 40.0 C. The heated and humidified gas mixtures were then routed independently to each of the exposure chambers at a constant
Fig. 1. Nitric oxide gas delivery. (A) Schematic representation of the gNO exposure chamber. (B) Insulating jacket surrounding the exposure chambers. Glove port (white arrows) in front of the jacket allows access to the chamber without danger of cross contamination. (C) The gas dilution manifold. The gas enters from bottom left of the diagram using stainless steel quick connects. Flow for each gas can be adjusted using valves and two digital flowmeters. The system has the capacity to mix and deliver five different gases simultaneously. The gas mixture is then passed through special filters to prevent possible contamination of system (black arrows). A flow rate of 10.0 L/min was used in all experiments. (D) Placement of cell culture flasks (25 cm2) inside exposure chambers.
132
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
flow rate of 10.0 liter per minute (L/min) for all experiments. Each of the exposure chambers could contain five standard 100 · 15 mm Petri dishes (8-758-03, Fisher Scientific, Canada) or three 6-well plates (6 Well Cell Culture Cluster-3516, Corning, USA) or four 75 cm2 cell culture flasks (Fisher Scientific, USA) each. A vacuum pump was placed at the exhaust port of the exposure chambers to create a negative pressure throughout the system and to prevent the possible back-flow of gases, thus ensuring a one-way flow of gas through the system. The exhaust gas mixtures, after passing through a HEPA filter, comprising mainly of NO, NO2, O2, and CO2, were safely vented to a Class II biosafety cabinet through a vinyl duct. NO, NO2, and O2 levels: chamber temperatures and relative humidity data were monitored and collected over a 72-h period to evaluate and to establish reliable steady-state environmental conditions inside the exposure chambers. Cell culture preparation and exposure Bacterial cell cultures Cultures from E. coli (ATCC 25922), P. aeruginosa (ATCC 27853), and S. aureus (ATCC 25923) strains were used for this study. Strains of each species were prepared one day before each experiment aseptically by adding 0.3–0.4 ml of tryptic soy broth liquid medium (TSB; Dalynn Biologicals, Calgary, Alta., Canada) to the freeze-dried material with a sterile pipette and vortexed. The mixture was then transferred to a test tube containing 5–6 ml of TSB medium. A few drops of inoculate were pipetted onto a tryptic soy agar plate (TSA). The plate was then placed in a conventional incubator at 37 C for 18–24 h and then observed for acceptable bacterial colony growth. On the day of each experiment, 3–5 isolated colonies from the freshly cultured TSA plate were selected. Using a sterile glass rod, colonies were transferred and suspended in 5 ml of TSB solution and labeled. The inoculum was mixed well and then the turbidity was visually compared against a number 0.5 McFarland standard, which approximated a cell count of 2.5 · 108 cfu/ml. If the turbidity exceeded that of the McFarland standard, the inoculum was then diluted with sterile broth or saline. Inoculums were placed in a conventional incubator (UltraTech WJ301D, Baxter, USA) for 20 min to acclimatize. A dilution of 0.1 ml inocula of E. coli, S. aureus, and P. aeruginosa cells were then aseptically pipetted onto clear TSA plates. The 0.1 ml sample was dropped onto the center surface of the agar plate and was then rapidly spread using a sterile (alcohol and flame) culture-spreader metal loop. The plates were maintained in an upright position until the inoculum was absorbed by the agar (approximately 10 min).
A number of preliminary serial dilution studies were performed to determine the optimal dilution to achieve approximately 30–200 colonies per plate. This was done using a micropipetter and aseptically transferring 0.1 ml of freshly prepared inoculum into a labeled 0.9 ml test tube containing 0.85% saline solution. Then 0.1 ml was transferred from the latter test tube (diluted suspension) into a second 0.9 ml of saline solution. This procedure was repeated until the desired dilution level was reached (·105 cfu/ml). The inoculated plates were then transferred and placed in an inverted position inside the exposure chamber. Twelve identical plates were prepared for each study. Four plates were placed in each arm or chamber of the in vitro cell exposure chamber. Four control plates were placed inside a conventional incubator at 37 C for 24 h in each experiment. Following the incubation period, a visual count of colony forming units (cfu) was obtained. Plates were then transferred to a conventional incubator and grown for an additional 24 h to ascertain any difference in colony size and number. It was assumed that each colony forming unit originated from ONE single bacterial cell. Skin cell culture Cultures of human foreskin fibroblasts were established as described previously in the literature by Ghahary et al. [31]. In brief, punch biopsy samples were prepared from human foreskin. The tissue was collected in sterile DulbeccoÕs modified EagleÕs medium (DMEM) with 10% fetal calf serum (FBS) (Gibco, Grand Island, NY), minced into small pieces of less than 0.5 mm in any dimension, washed with sterile medium six times, and distributed into 60 · 15 mm Petri culture dishes (Corning, Corning, NY), four pieces per dish. A sterile glass cover-slip was attached to the dish with a drop of sterile silicone grease to immobilize the tissue fragment. DMEM + Ab (penicillin G sodium 100 U/ml, streptomycin sulfate 100 g/ml, and amphotericin B 0.25 lg/ml) (3 ml) with 10% FBS was added to each dish and incubated at 37 C in a water-jacked humidified incubator in an atmosphere of 5% CO2. The medium was replaced twice weekly. After 4 weeks of incubation, cells were released from dishes by brief (5 min) treatment with 0.1% trypsin (Life technologies, Gaithersburg, MD) and 0.02% ethylenediaminetetraacetic acid (EDTA) (Sigma, St. Louis, MO) in PBS (pH 7.4) and transferred to 75 cm2 culture flasks (Corning, Corning, NY) for incubation. Upon reaching confluency, the cells were released by trypsinization, split for subculture at a ratio of 1:5, and reseeded into 25-cm2 flasks for exposure. Prior to each exposure, the old media (DMEM + 10% FBS) was removed, cell monolayer washed with PBS and fresh DMEM medium with only 2% FBS was added to the cells. Four flasks were prepared with a total depth of liquid of 4.0 mm for each of control and treated group. Fibroblasts from passages 3 to 6 were used for this study.
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
Gas exposure conditions To test the efficacy of the exposure chamber and gNO toxicological data, the growth of various bacterial strains as well as human skin fibroblasts in presence and absence of gNO was compared with identical cells incubated in a conventional incubator (Forma Scientific, Marietta, OH, USA). For validating the exposure chambers, cells were exposed to medical air alone (in absence of gNO) continuously at 10.0 L/min up to 24 h for bacteria and 96 h for fibroblast cells. For the toxicological studies, various doses of gNO were titrated into the medical air at same flow rate for 24 and 48 h periods. For skin fibroblasts, 5.0% CO2 was also added to the gas mixture in each of the control and treatment chambers. Analysis and measurements Cell viability For bacterial cells, visual count of the cfu from each plate was used as a measure of cell growth. Cell counts are expressed as relative percentage of cfu in the treated groups in comparison to the controls. For dermal fibroblasts, the proliferation assays were performed on cells seeded onto 25 cm2 flasks and then compared with the controls. In brief, the fibroblasts were detached after 5 min of exposure to 0.1% trypsin (Life technologies, Gaithersburg, MD) and 0.02% ethylenediaminetetraacetic acid (EDTA) (Sigma, St. Louis, MO) in PBS (pH 7.4), then re-suspended in 1 ml of DMEM with 10% FBS (Gibco, Grand Island, NY) and counted in a hemocytometer by an inverted microscope. The plating efficiency of the fibroblast cultures incubated in gNO chamber was calculated by using the ratio of cell counts on day 4 to the original number of cells plated. To further validate the optimal conditions for growing skin cells and their viability, 1-h culture plate attachment efficiency of the treated fibroblasts was compared with the control cells, immediately following the 24 h incubation in the gNO chamber and the conventional incubator. Fibroblasts were sub-cultured into 25 cm2 flasks and trypsinized for counting after 1 h incubation in a conventional incubator. Morphology, confluency, and contamination of the cells were also evaluated microscopically. Cell proliferation assay To measure the rate of fibroblast proliferation in response to exposure to 20, 200, and 400 ppm gNO, a [3H]thymidine proliferation assay was carried out on the cultured fibroblasts. In brief, [3H]thymidine (Perkin-Elmer Life Sciences, Boston, MA) was added to the conditioned medium of each sample, following gNO exposure for 24 and 48 h, for a final concentration of 2.0 lCi/ml and then incubated for 16 h in a conventional incubator. After this period, fibroblasts were then
133
harvested, washed three times with PBS, dissolved in GITC, and added to the scintillation fluid (Amersham, Oakville, ON, Canada). Radioactive counting was performed using the scintillation counter (Beckman, Fullerton, CA, USA) and measured in counts per min (cpm). Nitric oxide analysis Continuous readings of NO, NO2, and O2 concentrations in the gas mixtures directly above the plates in the exposure chambers were obtained using a calibrated AeroNOx electrochemical NO analyzer (Pulmonox Medical, Tofield, Alta., Canada). The AeroNOx analyzer was calibrated at least every third experiment with NIST referenced nitrogen dioxide (11.9 ppm NO2) and NO (78.2 ppm) calibration gas based on the manufacturerÕs guidelines and specifications (Pulmonox Medical, Tofield, Alta., Canada). To characterize the uptake of gNO into the aqueous phase, cumulative NO 2 and NO3 (NOx) concentrations were measured in the exposed solutions [18]. Three milliliters of saline, TSB, DMEM, and Middlebrook 7H9 mediums was continuously exposed to 20 and 200 ppm gNO for 48 h, under similar conditions. The depth of the liquid media inside a 6-well plate was consistently measured at 4.0 mm. Samples were collected at various time points and then analyzed for NOx concentration using Griess reaction technique as a NO marker that is well described in the literature [32]. In brief, after converting nitrates ðNO 3 Þ to nitrites with nitrate reductase and addition of Griess reagent, the absorbance of the samples were measured at 540 nm using a spectrophotometer (Helios Gamma & Delta, Unicam UV-Visible Spectroscopy, Cambridge, UK). Statistical analysis The results from all of the experiments were analyzed using the unpaired student t test for comparison between any two groups, and by nonparametric equivalents of ANOVA for multiple comparisons. P < 0.05 was considered to indicate statistical significance. Unless otherwise indicated, results are presented as mean ± standard deviation for at least three independent data. A commercially available statistics/graphics package was used to analyze and graph the data (Graphpad-Prism V 3.0, GraphPad Softward, USA).
Results Stability of the in vitro delivery system A comparison of 80 ppm gNO concentration between the two chambers for 72 h delivery did not reveal a significant difference (Fig. 2A). Similar results were obtained for 20, 50, and 200 ppm gNO concentrations
134
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
Fig. 2. Determination of the reproducibility of the gNO delivery system. Various parameters of the delivery system were monitored during 72 h of continuous exposure to a pre-set concentration of NO (80 ppm). (A) The gNO concentration in the exposure chamber was monitored with an electrochemical analyzer (AeroNOx, PulmoNOx Medical, Alberta). The maximum variation observed was within ±5% of the pre-set concentration. (B) The buildup of nitrogen dioxide was monitored during 72-h exposure. The NO2 level remained below 3.0 ppm for the duration of this study. (C) The partial pressure of oxygen in the exposure chamber was measured using the AeroNOx device. The partial pressure of O2 in the exposure chambers remained constant for the duration of exposure. (D) The gNO delivery system was able to maintain consistent temperature levels inside exposure chambers at 36.9 ± 0.4 C. Data are presented as single data points from chamber 1 (– –) and chamber 2 (– –).
(data not shown). The delivery system was capable of maintaining a steady NO concentration for the duration of each experiment, averaging 78 ± 1.0 ppm. The buildup of NO2 remained consistently low for the duration of each exposure and averaged 2.0 ± 0.1 ppm (Fig. 2B), well below the standard set for nitric oxide inhalation by Occupational Safety and Health Administration [33]. The partial pressure of O2 within the system was maintained at a consistent level and did not fluctuate more than 0.5% for the duration of the exposure (Fig. 2C). The gNO delivery system maintained consistent temperature levels inside both exposure chambers at 36.9 ± 0.4 C (Fig. 2D). These results confirmed that the cell exposure device consistently maintained a set of environmental variables that are in accordance with the typical specifications for a conventional incubator. In addition, there were no differences in these ambient variables between the two exposure chambers. These results suggest that the designed exposure device allows Petri dishes (solid media), 6-well plates and culture flasks (liquid media containing planktonic bacterial suspension or attached human fibroblasts) to be tested with only one variable change, such as gNO dose, while keeping all other environmental variables constant.
Fig. 3. Nitric oxide gas delivery rate in various culture mediums. The uptake of gNO was estimated by monitoring NOx concentrations in cell-free media. NOx levels were measured using Griess reagent, following continuous exposure to 20 ppm (A) and 200 ppm gNO (B) in saline ( ), TSB ( ), DulbeccoÕs modified EagleÕs medium (DMEM) (–m–), and MB 7H9 (– –). Data are presented as single data points.
Nitric oxide uptake in liquid media As seen in Fig. 3, gNO was transferred in a relatively linear rate in all liquid media at both 20 and 200 ppm
concentrations (Figs. 3A and B, respectively). The transfer or uptake rates of gNO into saline, TSB, MB 7H9, and DMEM are shown in Table 1. The rates were calcu-
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140 Table 1 gNO uptake rate in various media gNO diffusion rate (lM/h)
Saline Broth MB 7H9 DMEM
20 ppm
200 ppm
1.42 ± 0.09 2.34 ± 0.13 2.72 ± 0.28 3.32 ± 0.16
22.5 ± 1.12 75.4 ± 1.85 86.6 ± 2.48 88.4 ± 2.30
Estimated overall uptake rate of gNO in Saline, TSB, MB 7H9, and DMEM solutions. Results are the slopes obtained from best-fit linear regression curves shown in Fig. 3, followed by the standard error of each slope.
lated from the slope of the best-fit linear regression lines and deviations from the best fit are represented by standard error of the slopes. Due to sensitivity of human fibroblasts to pH levels, samples were analyzed during a 48 hour continuous exposure to 200 ppm gNO. The pH in DMEM samples showed a range from 7.20 ± 0.1 to 7.53 ± 0.25, which is a value well tolerated by cultured fibroblasts [31]. Cell growth and viability Bacterial growth The results of three separate experiments, summarized in Fig. 4, indicated that E. coli, P. aeruginosa, and S. aureus grew as well in the designed exposure device as in a conventional incubator. The mean viabilities ± SD of E. coli, P. aeruginosa, and S. aureus grown in the exposure chamber relative to the conventional incubator were 100.2 ± 8.7, 134.9 ± 18.0,
Fig. 4. Bacterial cell viability inside the exposure chamber in the absence of gNO. Three separate experiments were carried out for each organism, in quadruplicate. Treated groups were exposed to medical air inside the gNO exposure chamber for 24 h at 37 C and the control group was placed inside a conventional incubator for 24 h at 37 C. Bacterial viability is expressed as the percent colonies counted in the treated group relative to the control group. On average, E. coli, S. aureus, and P. aeruginosa each expressed growth of 100.2 ± 8.7, 107.5 ± 7.5, and 134.9 ± 18.0% colony count relative to the control, respectively. Results of three independent experiments are expressed as means ± standard deviation.
135
107.5 ± 7.5, respectively. There was no significant difference in growth between the gNO exposure device and the conventional incubator. We concluded that the exposure device could be used to grow bacteria normally in the absence of gNO, and that any toxic effects observed in subsequent experiments were not due to incubation in the exposure chamber. Three independent experiments were performed for each gNO dose of 0, 50, 80, 120, and 160 ppm and the bacterial survival was expressed as a percent of control. As shown in Fig. 6A, the mean P. aeruginosa survival rates for 50, 80, 120, and 160 ppm exposures were 80.7 (7.3), 63.9 (5.6), 22.5 (8.9), and 16.4 (5.6), respectively (SD appears in parentheses). Similarly, the mean survival rate for S. aureus at 50, 80, 120, and 160 ppm exposures were 97.1 (14.3), 66.7 (9.3), 23.4 (1.1), and 2.2 (0.4), respectively (Fig. 6B). There were significant differences between all the bacterial survival means for 50, 80, 120, and 160 ppm gNO concentrations (p < 0.05). There were no significant differences between the bacterial survival means for all plates incubated in the control chamber. There was no difference in the number and size of gNO exposed colonies when they were incubated in a conventional incubator for an additional 24 h. However, control colonies coalesced and could not be counted, following the second 24-h incubation, due to their large size and the number of colonies. These results indicated that control colonies continued to grow whereas the colonies exposed to gNO did not increase in size. These experiments suggest that P. aeruginosa and S. aureus growth is inhibited as the concentration of gNO is increased from 50 to 160 ppm. Skin cell growth Cell proliferation data from three independent experiments are summarized in Fig. 5A. The fibroblasts incubated in the exposure chamber reached an average total cell count of approximately 1.02 · 105 cells per flask. The average total cell count for cells grown in the exposure chamber was very similar (p > 0.05) to that of the control group, which reached about 1.0 · 105 cells per flask. The plating efficiency (expressed as a percentage) was calculated as the ratio of total cells per flask post-incubation to the number of cells originally plated. The 96-h plating efficiency for dermal fibroblasts grown in the exposure chamber and the conventional incubator were 413% and 415%, respectively. To evaluate the fibroblastsÕ ability to reattach to the culture plates as an indication of healthy cell function, their cell attachment capacity in 1-h was evaluated immediately following growth in the exposure chamber or conventional incubator. As shown in Fig. 5B, no significant difference was found in the cell attachment capacity between gNO and the control groups (p > 0.05). Approximately 70% of the trypsinized cells were able to re-attach in the exposure cham-
136
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
Fig. 5. Viability of human fibroblast culture incubated in the exposure chamber in the absence of gNO. (A) Fibroblast cell proliferation was evaluated following 4-day incubation in the exposure chamber (grey bars) or in a conventional cell culture incubator (black bars). (B) The fibroblasts attachment capacity in 1-h was evaluated immediately following 24-h incubation in the exposure chamber or the conventional incubator. (C) Microscopic evaluation of fibroblast morphology for cells grown inside the exposure chamber (in absence of gNO) or in a conventional incubator (control). (A and B) Bars represent the mean of at least three measurements with the corresponding standard deviation.
ber group as compared to 75% of the cells in the control group. A further assessment of normal cell morphology was done using light microscopy after 96 h of incubation in the exposure chamber or the conventional incubator (Fig. 5C). In both groups, spindle-shaped fibroblasts appeared healthy and firmly attached to the flasks. The number of dead cells (white dots in Fig. 5C) floating in the media was minimal in both groups. We therefore concluded that the exposure chamber could be used to grow dermal fibroblasts in the absence of gNO, and the effects observed in subsequent experiments were not due to extraneous variables attributed to the exposure device. To assess the effect of gNO on human dermal fibroblast proliferation, [3H]thymidine incorporation assay was performed as described under Materials and methods. A comparison between 20 and 200 ppm gNO-exposed fibroblasts and the control group at 24 and 48 h did not reveal a significant difference in cell proliferation (Fig. 6C). Proliferation rate is directly proportional to
Fig. 6. Effect of gNO on cells survival. (A and B) P. aeruginosa (A) and S. aureus (B) survival on solid media in the presence of gNO. Bacterial growth was assessed at various delivered NO concentrations and compared to appropriate controls (bacteria grown in absence of gNO). (C) Human dermal fibroblasts exposed to 20, 200, and 400 ppm gNO for 24 and 48 h continuously. Fibroblast proliferation under gNO was assessed with [3H]thymidine incorporation assay and compared with the control group (fibroblast grown in absence of gNO). Data represent the mean and standard deviation of at least three measurements.
the number of radioactive thymidine nucleotide incorporated into the newly synthesized DNA in the cells and is represented as percent count per minute relative to control. However, dermal fibroblast exposure to 400 ppm gNO showed significant toxicity when compared to cells exposed to medical air alone (control). Exposure to 400 ppm gNO for 24 and 48 h inside the chamber reduced the dermal fibroblast proliferation to 66 ± 5.8 and 77.8 ± 7.5%, respectively, of that in the control group. These results demonstrated that the system presented here is capable of detecting dose-dependent toxicological data in cultured dermal fibroblasts following exposure to exogenous NO gas.
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
Discussion To systematically evaluate the effects of gNO on a variety of cell types, a new exposure system was designed, validated, and utilized herein. More specifically, an in vitro cell exposure device was designed to deliver constant levels of gNO, or other complex gas mixtures, from an exogenous source (800 ppm NO cylinder) to a heated, humidified chamber optimized for growth of bacterial and/or mammalian cell cultures. To validate the exposure device, the distribution and reproducibility of several factors such as concentration of NO, O2, CO2, NO2, total NOx levels in solution, chamber temperature, and humidity were monitored and evaluated. This exposure device was found to be capable of maintaining steady-state conditions over several days, for all of the parameters noted above. The exposure device also allowed each of the above parameters to be independently controlled and adjusted. The exposure device incorporated a number of innovative improvements over previous gNO delivery-incubation systems. Specifically, the exposure device provided two independent chambers with identical environmental conditions, wherein cells could be incubated in the presence of gNO in one chamber and in the absence of gNO (control) in the other. As well, each chamber allowed for up to three microtitre dishes or nine Petri dishes to increase the number of samples exposed during each experiment. Additionally, a vacuum pump was incorporated at the exhaust manifold to create a slight negative pressure inside the system, to ensure unidirectional flow of gas, and to minimize cross-contamination due to back-flow of air. To ensure the accuracy of NO, NO2, O2, and CO2 concentrations in the gas mixture, an adjustable gas dilution manifold with digital flowmeters was incorporated into the system. The manifold allows up to five different gases to be mixed simultaneously. Further, continuous samples from the space directly above the culture dishes reflect more accurately the actual concentration of the constituent gases than in other methods. It is well established that oxidation of NO leads to formation of NO2, which can be toxic to cells [34]. The formation of NO2 is dependent on the concentration of initial reactants (NO and O2) as well as the time available for the reaction to take place [21]. As such, the gas flow rate was adjusted in the system to minimize contact time between NO and O2 prior to delivery to exposure chambers. At a flow rate of 10.0 L/min and up to 200 ppm of continuous gNO delivery, the maximum level of NO2 in the exposure chambers was kept below 6.0 ppm. This level of NO2 was noted to be within an acceptable and safe clinical range for inhaled NO therapy in order to avoid toxic effects of NO2 [35]. In addition, previous toxicological studies have not reported any bactericidal effects at these concentrations
137
of NO2 [36]. To remove any false positive conclusions in our studies due to the potential bacteriastatic effects of NO2, as apposed to gNO, both S. aureus and P. aeruginosa were exposed solely to 20 ppm NO2 for 24 h without displaying significant bactericidal effect (data not shown). For cells cultured in aqueous media, it is possible that the liquid layer itself may act as a physical barrier through which gNO must pass. To characterize the uptake or transfer of gNO into different culture media, production of NO 2 and NO3 (stable end-products of NO oxidation) were monitored in saline, TSB, DMEM, and MB 7H9 for up to 48 h. Previous studies have demonstrated that the volume and subsequently the depth of aqueous layer affected gas uptake [37,38]. In order the standardize the uptake of the gas into the liquid, the depth of culture medium in the flasks and 6-well plates was consistently kept at a depth of 4.0 mm for all studies regardless of the volume of media used. To characterize the NO uptake rate in liquid media, the reaction between NO and O2 followed by subsequent reactions with water molecules should be considered. The major oxidation reactions of NO by O2 in an aqueous solution is shown below [18]: 2NO+O
2
NO+NO N
2
O
3
!2NO !N
2
+H
2
2
ð1Þ
2
O
ð2Þ
3
O !2NO
2
+2H
þ
ð3Þ
It is important to emphasize that gNO was the only exogenous source of nitrogen added to the gNO exposure chamber when compared to the control exposure chamber (medical air alone). Since nitrite ðNO 2 Þ is a stable end product of NO oxidation, its rate of formation is controlled by the reaction shown in Eq. (1). Under these conditions, as claimed by Wang et al., the reactions can be simplified to: 4NO + O
2
+H
2
O !4NO
2
+ 4H
þ
ð4Þ
Therefore, when NO is the only exogenous nitrogen oxide introduced to the system, the reaction intermediates (NO2 and N2O3) are present in minute amounts and any change in NO 2 concentration should directly correlate with the amount of gNO delivered within the exposure chamber [18]. The NO 3 concentration was not measured separately in our studies, as negligible amounts have been reported in previous studies evaluating NO uptake in aqueous solutions [18,39]. Additionally, the measurement of NO by Griess reagent 2 technique as a marker for NO uptake has been well established in the literature [39–41]. It should be noted that the NO2 present in the gas mixture can also contribute to the total NOx concentration measured in the liquid media [39]. The contribution of NO2 formed in the gas mixture to the total NOx con-
138
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
centration in the liquid samples becomes more significant as the NO2 partial pressure increases because of the higher diffusivity of NO2 than NO (2.2 · 1012 vs. 1.7 · 1013 mol/m/s/Pa) [18]. Using 800 ppm gNO at a high flow rate helped to minimize the contact time between NO and O2. Indeed this is confirmed in Fig. 2, showing that NO2 only constituted 2% of the total gas mixture. In our delivery system, the exposure to pure gNO was avoided as the presence of oxygen-enriched air was essential for growth and function of dermal fibroblasts in culture [31,42]. Under the exposure conditions described above, the average transfer rate of gNO between all the complex media (TSB, DMEM, and MB 7H9) was approximately 2.78 ± 0.71 and 78.2 ± 7.0 lM/h for 20 and 200 ppm, respectively. Curiously, under similar exposure conditions, the rate of gNO uptake by saline solution was about 2–4 times lower than the rate observed in the other media (Table 1). This observation may be explained by Keynes et al. [43] whom recently reported that there are NO-consuming ingredients found in the more complex culture media with the ability to removing NO. They concluded that Hepes buffer and/or the vitamin riboflavin, present in the common culture media, can act independently or synergistically to remove significant amounts of NO generated and released by NO donors in liquid broth media. In addition, the gNO delivery system described herein, unlike NO donors, provides a continuous source of gNO to the liquid inside the exposure chamber. Therefore, the combined effect of an unlimited source of gNO coupled with the binding or apparent removal of NO in aqueous phase by the complex media may result in an aberrant concentration of NOx measured by the Griess technique. The lack of NO-consuming agents in saline solution can offer some explanation for the lower levels of NOx observed following exposure to gNO under identical conditions compared to the more complex media. In future studies, addition of superoxide dismutase to complex biological media may significantly reverse the effect of superoxidedependent consumption of NO in aqueous phase [43]. It was also noted that a 10-fold increase in gNO concentration, from 20 to 200 ppm (Fig. 3), did not lead to a proportionate increase in the NO uptake rate. The total NOx concentration measured following 200 ppm gNO exposure was on average 20 times greater than NOx concentration during the 20 ppm exposure, while other factors remained constant. This observation may be explained by the fact that NO is relatively more stable at lower concentrations, but at higher concentrations in the presence of O2 it may be oxidized to form NO2 within the liquid phase [43,44]. Additionally, as described above, the removal of NO in aqueous phase by NO-consuming agents could possibly explain the disproportionate uptake of gNO into the aqueous solution between the lower and higher NO concentrations.
One key set of tests used to validate the use of the exposure device was its capacity to provide optimal conditions for prokaryotic and eukaryotic cell growth. The viability and growth pattern observed for cells grown inside the exposure device closely resembled the viability and growth pattern observed for cells grown in the commercially available conventional incubator. All three bacterial strains tested inside the exposure device showed equal or better growth compared to the control group in the conventional incubator (Fig. 4). This was not surprising, as conditions such as temperature, humidity, and oxygen were optimized for bacterial growth in the new cell exposure device. Successful outcomes of fibroblast cell attachment, morphology, confluency, and survival, all confirmed the ability of the cell exposure device to function in a manner equivalent to a conventional incubator. The ability of the exposure device to detect toxicological data was validated by testing the effect of gNO on P. aeruginosa, S. aureus survival, and human dermal fibroblasts proliferation in a series of dose–response studies. The bacterial isolates and human dermal cells were selected for these studies based on the potential clinical applications for gNO as an antimicrobial agent as well as a wound repair mediatory factor. We speculate that gNO could be inhaled into the lungs for treating respiratory infections or delivered directly to the surface of infected skin wounds as a topical antimicrobial agent. The in vitro antimicrobial effect of NO against common skin pathogens such as S. aureus was previously demonstrated by using a NO-generating gel on a semi-permeable membrane as well as an acidified nitrite solution [27,45]. However, these techniques are complex, messy and have side-effects that simple gNO may avoid. Both in vitro and in vivo studies of NO on P. aeruginosa in pneumonia or cystic fibrosis have shown a marked reduction in pulmonary bacterial load as well as adherence to human bronchial epithelial cells [46,47]. In our studies, a bacterial effect of 84% and 98% kill, relative to the control, was observed at 160 ppm gNO in P. aureus and S. aureus, respectively. These are encouraging preliminary findings and further study is warranted before any conclusions can be made whether gNO is an effective antimicrobial agent for respiratory or wound infections. To study the effect of gNO on dermal fibroblast proliferation, a [3H]thymidine incorporation assay was carried out for 16 h following exposure to different doses of gNO. Interestingly, dermal fibroblasts cultured in DMEM did not express any toxic effect following 48 h exposure to 200 ppm gNO. However, the exposure device was able to detect significant reduction in fibroblasts proliferation, compared to the control, at 400 ppm gNO (Fig. 6C). No significant difference was observed in fibroblast proliferation between 24 and 48-h exposures. An oxidative DNA damage was previously reported in cultured mouse fibroblasts following expo-
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
sure to high concentrations (>1 mM) of an NO-donor called dipropylenetriamine-NONOate [48]. In conclusion, the results of the experiments presented have characterized and validated the properties of a new gNO exposure device. These data suggest that the gNO can be delivered in a controlled and reproducible manner for long-term exposure experiments. The results confirm that cultured cells can be incubated in the device for at least 96 h while maintaining the cell viability. The exposure system is also capable of detecting toxicological data in dose–response studies on bacteria and mammalian cells. The kinetics of gNO uptake was characterized in the system by measuring the total NOx concentration in the solution, allowing for indirect estimation of gNO delivery. The device allows continuous monitoring of several parameters such as concentrations of gNO, NO2, and O2, temperature, and humidity levels. Another advantage of this device is its mobility and flexibility, which allows the cell cultures to be grown and maintained independent of an external incubator or biosafety cabinet. Finally, the gNO cell exposure device was validated for bacterial and human cell culture growth. This device may increase the ability to carry out future in vitro experiments analyzing the biological and therapeutic effects of gNO.
Acknowledgments
[7] [8] [9] [10] [11]
[12]
[13]
[14]
[15]
[16] [17]
We thank the team at Pulmonox for their support of this study, in particular, Mr. Robert Lee for his technical assistance in construction of the device. The authors also thank Mr. Rob Cornell from Fisher and Paykel Healthcare, for providing humidifiers and technical support and also University of Alberta Wound Research Group for providing fibroblasts cultures. This study was funded in part by Pulmonox Medical.
References [1] T.N. Raju, The Nobel chronicles. 1998: Robert Francis Furchgott (b 1911), Louis J Ignarro (b 1941), and Ferid Murad (b 1936), Lancet 356 (2000) 346. [2] A. Papapetropoulos, G. Garcia-Cardena, J.A. Madri, W.C. Sessa, Nitric oxide production contributes to the angiogenic properties of vascular endothelial growth factor in human endothelial cells, J. Clin. Invest. 100 (1997) 3131–3139. [3] J. MacMicking, Q.W. Xie, C. Nathan, Nitric oxide and macrophage function, Annu. Rev. Immunol. 15 (1997) 323–350. [4] Y. Yui, R. Hattori, K. Kosuga, H. Eizawa, K. Hiki, S. Ohkawa, K. Ohnishi, S. Terao, C. Kawai, Calmodulin-independent nitric oxide synthase from rat polymorphonuclear neutrophils, J. Biol. Chem. 266 (1991) 3369–3371. [5] C. Nathan, Nitric oxide as a secretory product of mammalian cells, FASEB J. 6 (1992) 3051–3064. [6] D.E. Heck, D.L. Laskin, C.R. Gardner, J.D. Laskin, Epidermal growth factor suppresses nitric oxide and hydrogen peroxide production by keratinocytes. Potential role for nitric oxide in the
[18] [19]
[20]
[21]
[22]
[23] [24]
[25]
[26]
139
regulation of wound healing, J. Biol. Chem. 267 (1992) 21277– 21280. M. Kelm, The L-arginine-nitric oxide pathway in hypertension, Curr. Hypertens. Rep 5 (2003) 80–86. H.Y. Yun, V.L. Dawson, T.M. Dawson, Neurobiology of nitric oxide, Crit. Rev. Neurobiol. 10 (1996) 291–316. D.R. Riddell, J.S. Owen, Nitric oxide and platelet aggregation, Vitam. Horm. 57 (1999) 25–48. M.B. Witte, A. Barbul, Role of nitric oxide in wound repair, Am. J. Surg. 183 (2002) 406–412. S. Sasaki, T. Miura, S. Nishikawa, K. Yamada, M. Hirasue, A. Nakane, Protective role of nitric oxide in Staphylococcus aureus infection in mice, Infect. Immun. 66 (1998) 1017– 1022. D. Bruch-Gerharz, T. Ruzicka, V. Kolb-Bachofen, Nitric oxide in human skin: current status and future prospects, J. Invest. Dermatol. 110 (1998) 1–7. A.R. Amin, P. Vyas, M. Attur, J. Leszczynska-Piziak, I.R. Patel, G. Weissmann, S.B. Abramson, The mode of action of aspirinlike drugs: effect on inducible nitric oxide synthase, Proc. Natl. Acad. Sci. USA 92 (1995) 7926–7930. C.M. Maragos, D. Morley, D.A. Wink, T.M. Dunams, J.E. Saavedra, A. Hoffman, A.A. Bove, L. Isaac, J.A. Hrabie, L.K. Keefer, Complexes of NO with nucleophiles as agents for the controlled biological release of nitric oxide. Vasorelaxant effects, J. Med. Chem. 34 (1991) 3242–3247. D.D. Thomas, K.M. Miranda, M.G. Espey, D. Citrin, D. JourdÕheuil, N. Paolocci, S.J. Hewett, C.A. Colton, M.B. Grisham, M. Feelisch, D.A. Wink, Guide for the use of nitric oxide (NO) donors as probes of the chemistry of NO and related redox species in biological systems, Methods Enzymol. 359 (2002) 84– 105. M. Feelisch, J.S. Stamler, Methods in Nitric Oxide Research, Wiley, New York, 1996. M. Kavdia, S. Nagarajan, R.S. Lewis, Novel devices for the predictable delivery of nitric oxide to aqueous solutions, Chem. Res. Toxicol. 11 (1998) 1346–1351. C. Wang, W.M. Deen, Nitric oxide delivery system for cell culture studies, Ann. Biomed. Eng 31 (2003) 65–79. S. Tamir, R.S. Lewis, W.T. deRojas, W.M. Deen, J.S. Wishnok, S.R. Tannenbaum, The influence of delivery rate on the chemistry and biological effects of nitric oxide, Chem. Res. Toxicol. 6 (1993) 895–899. R. Long, B. Light, J.A. Talbot, Mycobacteriocidal action of exogenous nitric oxide, Antimicrob. Agents Chemother. 43 (1999) 403–405. T. Hoehn, J. Huebner, E. Paboura, M. Krause, J.U. Leititis, Effect of therapeutic concentrations of nitric oxide on bacterial growth in vitro, Crit. Care Med. 26 (1998) 1857–1862. C. Voisin, C. Aerts, E. Jakubczak, A.B. Tonnel, Cell culture in the gas phase: a new experimental model for the in vitro study of alveolar macrophages, Bull. Eur. Physiopathol. Respir. 13 (1977) 69–82. F.Y. Liew, F.E. Cox, Nonspecific defence mechanism: the role of nitric oxide, Immunol. Today 12 (1991) A17–A21. F.C. Fang, Perspectives series: host/pathogen interactions. Mechanisms of nitric oxide-related antimicrobial activity, J. Clin. Invest. 99 (1997) 2818–2825. J.D. Macmicking, R.J. North, R. LaCourse, J.S. Mudgett, S.K. Shah, C.F. Nathan, Identification of nitric oxide synthase are more susceptible to herpes simplex virus infection despite enhanced Th1 cell responses, Proc. Natl. Acad. Sci. USA 94 (1997) 5243–5248. A. Maclean, X.Q. Wei, F.P. Huang, U.A. AlAlem, W.L. Chan, F.Y. Liew, Mice lacking inducible nitric-oxide synthase are more susceptible to herpes simplex virus infection despite enhanced Th1 cell responses, J. Gen. Virol. 79 (1998) 825–830.
140
A. Ghaffari et al. / Nitric Oxide 12 (2005) 129–140
[27] R. Weller, R.J. Price, A.D. Ormerod, N. Benjamin, C. Leifert, Antimicrobial effect of acidified nitrite on dermatophyte fungi, Candida and bacterial skin pathogens, J. Appl. Microbiol. 90 (2001) 648–652. [28] E. Seifter, G. Rettura, A. Barbul, S.M. Levenson, Arginine: an essential amino acid for injured rats, Surgery 84 (1978) 224–230. [29] H.P. Shi, D.T. Efron, D. Most, U.S. Tantry, A. Barbul, Supplemental dietary arginine enhances wound healing in normal but not inducible nitric oxide synthase knockout mice, Surgery 128 (2000) 374–378. [30] K. Yamasaki, H.D. Edington, C. McClosky, E. Tzeng, A. Lizonova, I. Kovesdi, D.L. Steed, T.R. Billiar, Reversal of impaired wound repair in iNOS-deficient mice by topical adenoviral-mediated iNOS gene transfer, J. Clin. Invest. 101 (1998) 967–971. [31] A. Ghahary, Q. Shen, J.A. Rogers, R. Wang, A. Fathi-Afshar, P.G. Scott, E.E. Tredget, Liposome-associated interferon-alpha2b functions as anti-fibrogenic factor for human dermal fibroblasts, J. Invest. Dermatol. 109 (1997) 55–60. [32] L.C. Green, D.A. Wagner, J. Glogowski, P.L. Skipper, J.S. Wishnok, S.R. Tannenbaum, Analysis of nitrate, nitrite, and [15N]nitrate in biological fluids, Anal. Biochem. 126 (1982) 131– 138. [33] NIOSH. Recommendations for Occupational Safety and Health Standard. 37(S-7), 1988, pp. 1–29. Ref Type: Report. [34] D.B. Menzel, Antioxidants in lung disease, Toxicol. Ind. Health 9 (1993) 323–336. [35] J.D. Roberts Jr., J.R. Fineman, F.C. Morin III, P.W. Shaul, S. Rimar, M.D. Schreiber, R.A. Polin, M.S. Zwass, M.M. Zayek, I. Gross, M.A. Heymann, W.M. Zapol, Inhaled nitric oxide and persistent pulmonary hypertension of the newborn. The Inhaled Nitric Oxide Study Group, N. Engl. J. Med. 336 (1997) 605–610. [36] R.L. Mancinelli, C.P. McKay, Effects of nitric oxide and nitrogen dioxide on bacterial growth, Appl. Environ. Microbiol. 46 (1983) 198–202. [37] W. Behnke, K. Levesen, V. Scheer, C. Zetzsch, Interaction of gaseous pollutants with cultured cells and the liquid interface of the lung: an exploratory study on the uptake of ozone, in: U. Mohr (Ed.), Relationships between Respiratory Disease and Exposure to Air Pollution, IlSI, Washington, DC, 1998, pp. 314–319.
[38] D. Ritter, J.W. Knebel, M. Aufderheide, In vitro exposure of isolated cells to native gaseous compounds—development and validation of an optimized system for human lung cells, Exp. Toxicol. Pathol. 53 (2001) 373–386. [39] R.S. Lewis, W.M. Deen, Kinetics of the reaction of nitric oxide with oxygen in aqueous solutions, Chem. Res. Toxicol. 7 (1994) 568–574. [40] D.A. Wink, J.F. Darbyshire, R.W. Nims, J.E. Saavedra, P.C. Ford, Reactions of the bioregulatory agent nitric oxide in oxygenated aqueous media: determination of the kinetics for oxidation and nitrosation by intermediates generated in the NO/ O2 reaction, Chem. Res. Toxicol. 6 (1993) 23–27. [41] S. Goldstein, G. Czapski, Kinetics of nitric oxide autoxidation in aqueous solution in the absence and presence of various reductants, J. Am. Chem. Soc. 117 (1995) 12078–12084. [42] C.B. Lombello, S.M. Malmonge, M.L. Wada, Morphology of fibroblastic cells cultured on poly(HEMA-co-AA) substrates, Cytobios 101 (2000) 115–122. [43] R.G. Keynes, C. Griffiths, J. Garthwaite, Superoxide-dependent consumption of nitric oxide in biological media may confound in vitro experiments, Biochem. J. 369 (2003) 399–406. [44] K. Schmidt, W. Desch, P. Klatt, W.R. Kukovetz, B. Mayer, Release of nitric oxide from donors with known half-life: a mathematical model for calculating nitric oxide concentrations in aerobic solutions, Naunyn-SchmiedebergÕs Arch. Pharmacol. 355 (1997) 457–462. [45] J.B. Hardwick, A.T. Tucker, M. Wilks, A. Johnston, N. Benjamin, A novel method for the delivery of nitric oxide therapy to the skin of human subjects using a semi-permeable membrane, Clin. Sci. Lond. 100 (2001) 395–400. [46] K.E. Webert, J. Vanderzwan, M. Duggan, J.A. Scott, D.G. McCormack, J.F. Lewis, S. Mehta, Effects of inhaled nitric oxide in a rat model of Pseudomonas aeruginosa pneumonia, Crit. Care Med. 28 (2000) 2397–2405. [47] K.E. Darling, T.J. Evans, Effects of nitric oxide on Pseudomonas aeruginosa infection of epithelial cells from a human respiratory cell line derived from a patient with cystic fibrosis, Infect. Immun. 71 (2003) 2341–2349. [48] N. Phoa, B. Epe, Influence of nitric oxide on the generation and repair of oxidative DNA damage in mammalian cells, Carcinogenesis 23 (2002) 469–475.
