Cytotoxic effects of singlet oxygen.

The toxic effects of gas-phase singlet oxygen, 1O2, on the ciliated respiratory epithelium of hamster trachea have been demonstrated. Tracheal explants treated with 1O2 showed a dose-dependent decrease in cilia beating frequency and focal ciliostasis. A statistically significant decrease in ciliary activity occurred at 1O2 concentrations as low as 154 ppb after a 2-hr exposure. Cytological alterations in the mucociliary epithelium were observed in explants exposed to 235 ppb 1O2 or greater. When cytotoxic effects were related to the time of exposure to 1O2, maximum effects occurred after a 4-hr exposure. In vitro recovery studies indicate that ciliary activity returned to normal between 4 and 8 hr after exposure.


Introduction
An on-going objective of the work in this laboratory is to investigate the role of oxidants in environmental health. Toward this end, a number of studies have been initiated to more fully define the chemical and biological effects of individual oxidant species, especially of singlet molecular oxygen (102). Evidence for the toxicity of singlet oxygen has been based primarily on the observed reactivity of this oxidant with biological substrates and on experiments using sensitizers in which 102 may not be the only toxic species generated (1)(2)(3). A generator for forming 102 in the gas phase at atmospheric pressure was developed in our laboratory several years ago. This generator provides a clean source of 102 for laboratory experiments. We have shown that gas-phase 102 is cytotoxic (4) and induces unscheduled DNA synthesis (5) and sister chromatid exchange (unpublished results). These observations take on particularly important significance in light of the recent report (6) suggesting that prior use of photoradiation therapy may lead to cell damage and consequently resistance to the use of this promising cancer treatment method. It is now widely accepted that photoradiation therapy involves the production of 102 which is responsible for the cytocidal action (7).
In the dose-range study just completed, we show the effect of 102 on tracheal organ cultures of hamster respiratory epithelium over a relatively wide range of 102 concentrations. The time required for the respiratory epithelium to recover from exposure to 102 was also studied.
*Life Sciences Department, IIT Research Institute, 10

Generation of Singlet Oxygen and Exposure System
Singlet oxygen was generated by the method of heterogeneous photosensitization (9,10). Details of the 102 generation and exposure apparatus are described elsewhere (5). The 102 generator consisted of a 13-mm OD x 45-cm water-jacketed Pyrex flow tube lined with a thin film of Rose Bengal that was prepared by evaporation from a methanol solution of the dye. A mixture of 98% N2/2% 02 was passed through the flow tube at 8.2 L/min while the dye film was exposed to strong visible radiation from four 1000-watt projection lamps (G.E. Model DPT) enclosed in an air-cooled reflector. The exit gas was passed directly into the top of a cubeshaped exposure chamber (1 x 1 x 1 ft) and down onto culture dishes supported on a rack in the middle of the chamber.
The concentration of the gas-phase 102 entering the chamber was measured using a liquid nitrogen-cooled germanium photodiode that monitors the 1.27-,um emis- sion of 102 (9,10). The optical detector was calibrated in trapping experiments using the '02-tetramethylene reaction. The limit of detection of 102 in the gas phase using this reaction is 2.5 ppb. Control exposures were performed in which the generator lights were not turned on. These experiments were identical to the 102 exposures except no '02 was present in the 98% N2/2% 02 gas, as evidenced by failure to observe the 1.27 urm emission of '02 during control experiments.* Three uncovered culture dishes, each containing eight ring explants with 2 mL of L-15 medium (GIBCO), were placed in an atmospheric chamber (Bellco Glass, Inc., Vineland, NJ). The chamber was placed on a rocker platform (Bellco Glass, Inc.) that rocked at 10 cycles/ min, allowing the explants to contact both '02 and L-15 medium. Evaporation loss was corrected by the addition ofmedium to the dishes after each hour of exposure.

Experimental Design
In the first series of experiments, we determined the effect of increasing 102 concentration on the ciliary activity and epithelial cytology of hamster tracheal cultures by examining each of the eight ring explants after a culture period of 48 hr. Sets of tracheal cultures were then exposed for a 2-hr period to a mixture of nitrogen and oxygen gas containing '02 concentrations of 97, 154, 235, 266, 290, 345, 403, or 414 ppb. Immediately after exposure, we determined the alterations in the cilia *Since the initial exposure study (4) we have observed a variation Of 102 concentration with gas flow rate through the 102 generator.
Consequently, preliminary studies showing the cytotoxic effects of 102 were performed at a concentration of 827 ppb instead of the 121 ppb reported. beating frequency and cytology by light microscopy. For each singlet oxygen exposure concentration, a control experiment was performed in which eight ring explants were exposed for 2 hr to the nitrogen/oxygen gas stream without 102.
In the second series of experiments, we determined the time required for hamster tracheal cultures to recover from a 2-hr exposure to 401 ppb 102. Eight explants were examined at 1, 2, 4, and 8 hr after exposure for 2 hr. In a control experiment, eight explants were exposed to the 98% N2/2% 02 gas mixture without '02 for 2 hr. The explants were examined identically following completion of the control exposures.
Finally, we determined the effect of 102 exposure on tracheal epithelium by exposing eight explants to 410 ppb '02 for 2, 4, and 8 hr in separate experiments. Control experiments were also performed for each of the exposure periods. All ring explants were examined immediately after exposure.

Examination of Tracheal Cultures
Of the 24 tracheal explants exposed per 102 concentration, 8 explants were used for ciliary activity measurements and cytological observations (11). Cilia beating frequency at the periphery of the mucosa of each ring was determined using a Type 1531-AB electronic stroboscope (General Radio, Concord, MA) as the light source and a Nikon inverted microscope. Beating frequencies were measured at each quadrant of the lumen and the mean recorded as beats per minute. The percentage of normal respiratory epithelium, i.e., a smooth luminal surface with beating cilia, was also determined for each tracheal ring. Focal ciliostasis was defined as

Results
Organ culture studies that involve measurements of ciliar activity require tracheal rings with healthy ciliated epithelium. By allowing the cultures to stabilize in vitro for 48 hr prior to exposure to 102, we were able to minimize nonspecific damage. The tracheal ring explant system described in this work functioned well and gave consistent and repeatable responses in the background (before exposure) and control (exposure to the flowing gas stream without 102) experiments. The data presented in Table 1 show that tracheal rings exposed to concentrations of 154 to 414 ppb 102 exhibited a significant decrease in cilia beating frequency. At concentrations as low as 154 ppb, 102 produced a dosedependent decrease in ciliary activity. At 414 ppb '02, we observed a 35% decrease from the background measurement and from the control postexposure values.
Focal ciliostasis also increased with increased 102 concentrations. Cultures exposed to 102 at 266 ppb or greater produced 50 to 88% focal ciliostasis (Fig. 1). The appearance of the epithelium from explants exposed to 102 was different from the control ( Table 1).
Observations made on eight separate explants with phase optics showed sloughing of epithelial cells, as well as cells protruding into the lumen. The extent of these cytological changes depended on the concentration of 102 and were first observed at a concentration of 235 ppb. The time required for the tracheal explants to recover from a 2-hr exposure to 401 ppb 102 was studied ( Table  2). Observations made at 1 and 2 hr after exposure still showed a statistically significant reduction in cilia beating frequency and cytological alterations. None of the cultures showed focal ciliostasis after 1 hr recovery. After 4 hr recovery, the mean cilia beating frequency (1163) is significantly different than the control value (1197), and the mean percentage of normal epithelium is at 79%, which is also statistically different from the control value (87%). By 8 hr, both the cdliary activity and cytological observations approached the normal ranges. Table 3 shows the effects of 2-, 4-, and 8-hr exposure of tracheal rings to an 102 concentration of 410 ppb. Significant changes in ciliary activity and cytopathology were observed after 2 hr exposure. At the end of 4-hr exposure, the values for both the mean cilia beating frequency and the mean percent normal epithelium continued to decrease. The 8-hr exposure values show a cilia beating frequency of 663 beats/min with 40% of the normal epithelium remaining. These values were not significantly different from the 4-hr treatment. No significant changes were observed in separate control experiments conducted for 2, 4, and 8 hr.

Discussion
Hamster tracheal organ cultures have been used extensively to study in vitro effects of environmental agents on mucociliary activity (4,(13)(14)(15)(16)(17)(18). Results with this model system have shown a close correlation with in vivo studies (17,18), suggesting that hamster tracheal epithelium is an excellent model for measuring the effects of inhaled environmental agents on the upper respiratory tract. The system and methods for treating tracheal explants with gas-phase 102 appeared to func-  tion well and gave consistent results in both test and control cultures.
The results obtained in the present study show that gas-phase 102 produces a significant decrease in ciliary activity and that higher concentrations give a more pronounced response. The epithelium of '02-treated explants was moderately affected, and the percentage of tissue affected increased with increasing 02 concentrations. At an 102 concentration of 414 ppb, focal ciliostasis, i.e., no measurable ciliary movement at various sites along the lumen, was also pronounced. When cultures were allowed to recover in maintenance medium following exposure to a concentration of 410 ppb 102, ciliary activity increased gradually with time. After 4 hr, the beating frequency of the cultures exposed to 102 was indistinguishable from control cultures.
At the present time, it is impossible to determine whether singlet oxygen or a secondary toxic oxygen species, presumably a free radical, is responsible for the observed toxic effects. Studies in our laboratory with model membrane systems indicate that 102 reacts with phospholipids to form hydroperoxides. These hydroperoxides could, in turn, decompose to give free radicals. We are planning experiments to determine wheth-er 102 can pass through a model membrane and thus potentially be responsible for the observed effects. The involvement of other toxic oxygen species will also be investigated. In addition, we plan to identify intracellular sites of damage such as the mitochondria by looking at enzyme inactivation and cellular ATP levels.
The development of methods for the generation of and exposure to 102 has led to significant advances in demonstrating the action of 102 on the respiratory epithelium. The clearly observed concentration-dependent response of ciliary activity along the concentration range studied is evidence that the in vitro system we used is sensitive for measuring the effects of 102 on mucociliary respiratory epithelium.