Time--dose response for nitrogen dioxide exposure in an infectivity model system.

The concentration of NO2 in polluted atmosphere is subject to wide variation, according to peak traffic load, industrial productivity, intensity of sunlight and meteorological conditions. Normally NO2 has a low basal concentration with superimposed spikes when the above conditions are optimal for its production. Thus, it is important to determine the relative importance of a short-term, relatively high concentration of NO2 versus exposure for longer periods of minimal dose levels. This problem was approached experimentally by measuring the effect of NO2 on an animal's resistance to the induction of bacterial pneumonia. The data collected indicate that: (1) in short-term dose-response studies using the same Ct (concentration x time) product of 7, the actual concentration exerts a greater influence on NO2 effect than does the duration of exposure; (2) when concentration is held constant and the time increased, the average difference in mortality from controls can be seen after only 1 hr exposure to 3.5 ppm and after 3 weeks of exposure to 0.5 ppm; and (3) the relative mean survival time at 3.5 ppm for 1 hr was 18--36 hr less than that of the control animals.


Introduction
A number of atmospheric chemicals are characterized by very uneven concentrations from day to day or even hour by hour, as a result of the rate of their production and meteorological influences. This is particularly amplified in the case of chemicals formed secondarily through interactions occurring in the atmosphere. For instance, the level of atmospheric NO2 is governed not only by the rate of production of NO, but also by factors existing independently in the environment, favoring its conversion to NO2. This generally results in a low basal atmospheric concentration on which are superimposed peaks of higher concentrations, usually of very short duration and irregular occurrence. Thus, it is of great value to determine the relative importance of the more uniform low basal concentration (as might be modeled from a weekly, monthly, or *U.S. Environmental Protection Agency, National Environmental Research Center, Research Triangle Park, North Carolina 27711. yearly average) in comparison to the high, short peaks or vice versa. Furthermore, it is of importance to know what biological interactions may occur in both systems if applied in the same regimen of exposure.
The problem defined above was approached experimentally by conducting inhalation experiments in animals at various exposure regimens designed to elucidate the relative toxicologic importance of these factors, i.e., continuous basal dose versus short periods of exposure at higher levels. There is only a little information available in the literature concerning time-dose studies of NO2. Gray et al.
(1) found a sharp gradient in LCso at exceedingly high doses in favor of concentration from exposure ranging from 2 to 240 minutes; for instance, 1445 ppm for 2 min and 88 ppm for 240 min. Thus, computed on the basis of the product of concentration and time (ppm x min) the values are approximately 2,890 and 21,120 respectively, a difference of 8-fold in the product Ct.
Wagner et al. (2) noted no pulmonary emphysema after exposure of rats to 25 ppm NO2 February 1976 for 8 hr/day, 5 days per week. Similar negative results were reported by Hine et al. (3) and Boren (4) after intermittent exposure of rats to 25 ppm NO2 for a duration of 4 months. On the contrary, pulmonary emphysema was noted by Freeman et al. (5) after continuous exposure of rats to comparable concentrations. Slower repair was observed by Kleinerman and Wright (6) after cessation of exposure of animals to NO2 on a continuous basis as compared to similar experiments carried out on an intermittent schedule. Ehrlich and Henry (7) found emphysema in mice after both continuous and intermittent exposure at 0.94 mg/m3 (0.5 ppm) NO2. They also observed increased susceptibility to bacterial pneumonia when mice were subjected to the same exposure conditions. Ehrlich et al. (8) reported that fluctuating concentrations of NO2 represent a more significant factor in influencing immune response than exposure to a constant but higher concentration of NO2. Thus, continuous exposure to 3.8 mg/m3 (2 ppm) NO2 did not influence the formation of antibodies or levels of immunoglobulins, while exposure to 0.94 mg/m3 (0.5 ppm) with a daily 1-hr peak of 3.8 mg/m3 (2.0 ppm) NO2 depressed the ability to form serum neutralizing antibodies and significantly alter the levels of immunoglobulins IgM, IgGi, and IgG2.
Due to the ambiguities in the literature concerning dose response for NO2, more definitive work remains to be done to determine the relative importance of various dose regimens and the possible influence of threshold, tolerance, healing, and adaptation in biological reactions to NO2. Because of the interest in dose-response and the possibility that several mechanisms might be operating concurrently in NO2 toxicity, this problem was incorporated in the U.S.-U.S.S.R. Health Exchange cooperative agreement.
In this program the following studies were to be undertaken: short acute exposure versus the same Ct over a longer period; continuous dose versus the intermittent application; superimposition of spike on a low basal dose versus continuous and intermittent at a single concentration. Preliminary data from some of these experiments are presented here.
The model that has been employed in this work involves primarily resistance to the induction of bacterial pneumonia. This parameter was selected because of its demonstrated effects at ambient concentrations of air pollutants (7). This model also probably reflects a summation of the total toxic assaults on the deep lung such as edema, inflammation, cellular necrosis, reduced macrophage function, and upper airway effects, such as ciliostasis (9). Briefly, this model consists of the superimposition of a pathogenic bacterial aerosol (Streptococcus pyogenes, Group C, isolated from a pharyngeal abscess of a guinea pig) following exposure to the gaseous toxicant. Control animals receive bacteria aerosol only. This system has been efficiently utilized in environmental toxicologic studies of ozone (9, 10), nitrogen dioxide (11), irradiated auto smog (12), and trace metals, such as NiO and MnO2 (13,14). This model has been successfully employed to enhance the pulmonary infectivity in mice (10,11,13,14), rats (15), hamsters (11), and squirrel monkeys (16). Among the infectious agents eliciting this reaction in the above animals are Streptococcus pyogenes, Klebsiella pnedmoniae, Diplococcus pneumoniae, and influenza PR-8 virus. Thus, this model is an exquisitely sensitive indicator of biological effects in vivo. In addition, the model is useful in studying the mechanism of action of the pollutants (17,18).

Results
In answer to the question raised in the introduction, the following results have been obtained.
Dose-response studies were conducted utilizing a concentration x time (Ct) product of 7, which was based on the published threshold for NO2 using the infectivity model (3.5 ppm x 2 hr) (11). Figure 1 shows enhancement of mortality from pulmonary infection resulting from the same Ct exposure to NO2 over different time periods. It will be noted that the actual concentration exerted more influence than the duration of exposure. If no interaction existed between concentration and time, (i.e., if they had equal importance), all the test groups should demonstrate an increase in mortality of about 15%, whereas the trend of the combined data indicate a gradient response from a high mortality enhancement of approximately 50% to a low equalling controls.
The interaction of a constant concentration with time over a longer period was also examined. It is the overall plan to construct a family of curves at various concentrations from 0.5 to 14 ppm. At the present moment, three concentrations of NO2 have been studied. A level of 3.5 ppm was used for comparison with the study mentioned above. A lower value of 0.5 ppm was selected, since this concentration represented the lowest figure in the literature yielding positive results for chronic exposure in the infective system (11). Preliminary data at 1.5 ppm Environmental Health Perspectives of NO2 are also included. The data shown in Figure  2 represent 60 individual observations for the 3.5 ppm level, eight for the 0.5 ppm, and six for the 1.5 ppm level. The simple linear regression for the 3.5 ppm and 0.5 ppm of NO2 shows that the mortality increased with time and concentration and was statistically significant at the 0.01 probability level. Estimates which can be generated from these regressions with respect to the true average differences in mortality show that statistical significant change from controls (p < 0.05) can be predicted after 1 hr exposure to 3.5 ppm, and after 3 weeks at the 0.5 ppm concentration. However, the regression for 1.5 ppm was significant only at the 0.25 probability level and is represented by a dashed line. (Additional data will be collected in order to better define the 1.5 ppm response curve.) While it was noted in the first series of experiments ( Fig. 1) that there was a strong interaction in favor of concentration over time, the latter study still illustrates the importance of obtaining toxicological information from long-term exposures to NO2. The data from the same series of experiments were analyzed for influences of exposure time to 3.5 ppm on mean survival period, as calculated according to the equation:

MST=[E(AB) + (DL)]/n
where A is the last day on which any individual mouse was alive, B is the number of mice surviving A days; D is the last day of the experiment (in this case 15); L is the number of mice which were alive on day D; and n is the initial number of mice in the experimental group. A statistically significant (p < 0.05) regression with a negative slope was observed for mean survival of mice challenged with Streptococci pyogenes versus length of exposure to 3.5 ppm for various periods prior to the challenge. From Figure 3, it is evident that there is an inverse relationship between survival and length of exposure to a constant concentration of 3.5 ppm NO2. With this parameter, after 1 hr of exposure to 3.5 ppm, one may be 95% confident that the NO2-exposed mice, on the average, lived between 18   For the same model system (fixed concentration versus time) other concentrations are now being tested for the purpose of developing information which can be used to develop regression curves at other concentrations which then can be used for predicting purposes. Figure 4 presents preliminary data from experiments which were designed to test the effect of concentration and time as above, but utilizing an intermittent exposure model in place of a continuous exposure model. In these experiments, animals were exposed for 7 hr/days, 7 days/week to 3.5 ppm NO2. At various times, animals were removed and given the bacterial aerosol. Figure 4 illustrates that NO2 shows a significant increasing linear relationship with duration of exposure.

Conclusion
In short-term exposure to NO2, the concentration employed has a much greater influence than duration of exposure in terms of the same Ct.
There is a significant increase in toxicity with time during a continuous exposure to a low concentration.
Intermittent response at a constant concentration also shows an increased effect with duration of exposure, although the exact slope at this time is still unclear.  The question arises as to what is the cause of this increase in pulmonary infectivity which we enhance with NO2 It has been demonstrated in previous work at our laboratory that acute exposure to NO2 and O3 have a destructive. action on pulmonary alveolar macrophages (1 7-19), i.e., reduction in number, phagocytic competence, viability, and enzymatic function. These cells are. postulated to be the chief pulmonary defens'e agfainst infectious agents introduced via, the Iairway (20). Thus, this effect on the macrophage co;uld explain the effect seen after a single exposure, to the pollutant. However, it is also known that the alveolar macrophages are renewable resources and that these cells are probably derived from the bone marrow and transported to the lung .in the blood (21). Previous work (17) has indicated that the effect of a single dose of an Environmental Health Perspectives oxidant on the alveolar macrophage lasts for approximately 24 hr. Thus, if the action of NO2 on the infectivity parameter were solely to be a direct effect on the pulmonary macrophages, and that these cells were to be continually renewed, then no increase of toxic action with time would probably occur, particularly with the intermittent model. Therefore, the following hypotheses may be suggested to explain the enhancement of pulmonary infectivity following long term or repeated exposures.
Other damages which are known to occur in the lung, that is, various anatomic and chemical alterations, might have an influence on the role of the infection, and these lesions might increase with time (22,23).
There is an effect on the alveolar macrophage prior to their emergence into the lung which could be mediated through free radicals, nitroxides, or peroxides (24)(25)(26).
A combination of the above two responses could also contribute to the alteration of the host's natural defenses against the inhaled microorganism.
These mechanisms will have to be tested through specific experiments designed to uncover the mode of action of NO2. An amplification of the toxic effects of NO2 may be examined through a combination of acute, chronic, and intermittent NO2 exposures similar to the ones discussed in this paper.