10 – animal models to study for pollutant effects1

10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1 10 – animal models to study for pollutant effects1

Trang 1

10

Animal Models to Study for Pollutant Effects

URMILA R KODAVANTI and DANIEL L COSTA

Pulmonary Toxicology Branch, Experimental Toxicology Division, National Health and Environmental Effects Research Laboratory, US Environmental Protection Agency, Research Triangle Park, NC, USA

This article has been reviewed by the National Health and Environmental Effects Research Laboratory,

US Environmental Protection Agency and approved for publication Approval does not signify that the contents necessarily reflect the views and the policies of the Agency nor mention of trade names or com- mercial products constitute endorsement or recommendation for use

INTRODUCTION

Understanding of human pathobiology can be gained using laboratory animal models that have been developed to reflect human conditions Studies involving animal models provide important information on biological mechanisms of initiation, progression, and resolution of toxicant-induced tissue injury In the context of air pollution health effects studies, models are used to understand pollutant deposition and clearance as well as the mechanisms of biological action (Brain et al., 1988a; Reid, 1980; Stuart, 1976) Depending on the human condition being modeled, appropriate healthy or susceptible laboratory animals can be selected to estimate human health risks from inhaled pollutants (Slauson and Hahn, 1980) Much of our understanding of the toxicity of major air pollutants, such as tropospheric ozone (03), sulfur dioxide (SO2), nitrogen dioxide (NO2), particulate matter (PM), carbon monoxide (CO) and other 'air toxic' pollutants (e.g phosgene and metal/acid aerosols) has derived from studies using laboratory animal

Trang 2

166 IJ.P Kodavanti and D.L Costa

models (Committee of the Environmental and Occupational Health Assembly of the American Thoracic Society, 1996a, b)

Although a variety of inbred or outbred strains of mice, rats, hamsters, guinea pigs, rabbits, dogs and primates have been used in inhalation toxicology studies (Stuart, 1976), most studies have employed healthy animals to examine the basic mechanisms underlying pollutant-induced injury Recently it has become recognized that human susceptibilities to air pollution effects vary dramatically depending on the pre-existent conditions of the host, such as diseases, nutritional deficiencies (Barnes, 1995; Dockery et al., 1993; Hatch, 1995) or genetic differences (Kleeberger, 1995; Kodavanti et al., 1997a) In some instances, susceptibilities to pollutant effects are increased beyond the range covered by the typical uncertainty factors that are considered in health risk estimates to protect susceptible subgroups (Dockery et al., 1993; Pope et a1.,1995; Schwartz, 1994) The experimental studies performed in the past to understand altered responses to air pollutants due to these pre-existing conditions of the host have been spotty and the underlying mechanisms are not clear Thus, as risk-based toxicology progresses, the inclusion of animal models of varying susceptibilities (Fig 10.1) will provide sound experimental evidence for estimating the risks and for better understanding the toxic mechanisms

Pulmonary structure and pattern of air flow vary widely between animals and human, and therefore, it is critical that the animal model is chosen wisely Selection of the appropriate model depends upon the condition being modeled and how closely the model mimics the human in terms of its biological handling of the pollutant (Warheit, 1989)

In general, a range of human conditions- lung disease, genetic alteration/segregation, age,

b Models

Age Related

Fig 10.1 Varying host conditions that influence the susceptibility to inhaled or ingested toxicant-induced lung injury

Trang 3

10 Animal Models to Study for Pollutant Effects 167

malnutrition, or altered physiological profile due to systemic disease- encompasses most potential susceptible subgroups (Brain et al., 1988a) and these can be modeled in selected laboratory animal species In this chapter, we consider the application of both healthy and susceptible animal models for the evaluation of air pollution health effects Major focus

is given, however, to animal models of varying susceptibilities because the literature of the healthy animal response is large and has been detailed elsewhere (Gardner et aL, 1988; Lee and Schneider, 1995)

L A B O R A T O R Y A N I M A L S IN HEALTH EFFECTS S T U D I E S OF AIR

P O L L U T A N T S

The choice of animal species (large or small) in inhalation toxicology studies of air pollutants is frequently driven by factors ranging from their relevance to humans, their availability and expense, to ethical concerns about animal use Recently, greater interest has been shown in the potential to extrapolate from animal to human based on the genetic technology to identify specific host variables At present, rodents are most widely used in inhalation toxicological studies while dogs are largely used for cardiovascular studies, monkeys are preferred for selected chronic inhalation studies of pollutants, e.g the pulmonary effects of ozone (Harkema et al., 1993) and the nervous system effects of man- ganese (Bird et al., 1984) The advantages and disadvantages associated with use of laboratory animals for air pollution studies are highlighted below

The advantage of using mice is the availability of a wide variety of immunological reagents and detailed information on their genetic backgrounds However, the more than

200 strains of laboratory mice vary dramatically in their sensitivity to chemicals Thus, the selection of given strain may determine the 'toxicity' of an inhaled substance or the inter- pretation thereof (Stuart, 1976) Mice have been used in earlier studies involving pulmonary retention of inhaled radionuclides to gain information on comparative toxi- cities However, because the particle deposition pattern in mice can be dramatically different from that in humans, the target organ dose can vary between the mouse and the human (Snipes et al., 1989; Warheit, 1989) Because of their low spontaneous occurrence

of tumors, some mouse strains have been shown to be useful in carcinogenicity studies of radiation, the influenza virus and ozonized gasoline, to which they are more susceptible (Dagle and Sanders, 1984; Pott and Stober, 1983) Mice have also been used widely in allergy and asthma research, not so much because of their similarity to humans in terms

of the pathobiology of allergy and asthma, but because of their ease of use, the ready access

to large numbers of animals (except for special transgenics), and the information available

on their genetics and immunology

Because of their larger size, rats have the advantage of being amenable to more physiological measures and to adequate blood and tissue sampling The rat's longer life-span also

is an advantage in chronic studies (Stuart, 1976); however, as in the case of the mouse, strain-related differences in sensitivity may need to be considered in model selection Unlike the mouse, immunological reagents and detailed genetic information are less available for rats It is hoped that comparable information on rat molecular genetics and immunology will soon become more widely available, because the rat has long been the preferred test species in most inhalation and toxicology studies

Trang 4

168 U.P Kodavanti and D.L Costa

The extensive use of the rat in inhalation studies has provided considerable basic information about its morphologic and physiologic responses to injury, thereby supporting its continued use The rat model has been adopted as the standard for most bioassessments (e.g National Toxicology Program) and remains the species of choice in studies of inhalant- induced fibrosis and carcinoma In the 1970s, a number of studies of inhaled radionuclides yielded important information on lung deposition and dose variability (Bianco et al.,

1980) in the rat Acute and chronic inhalation studies of occupationally hazardous substances have been conducted using rats (Leach et al., 1973) Recent studies addressing the potential long-term consequence of particles such as diesel, soot, carbon black, toner, etc have demonstrated that the rat may be uniquely sensitive to clearance overload and progressive accumulation of particles in the lung over time (Brockmann et al., 1998; Mauderly

et al., 1994; Oberdorster, 1995; Valberg and Watson, 1996) This accumulation occurs at relatively high particle concentrations and appears to predispose the rat to tumorigenesis and fibrosis The mouse and hamster do not appear to respond to the same extent, although they also accumulate the particles Coal miners who have high lung burdens of dust also do not appear to develop tumors The reasons for these discrepancies are unclear and may relate to concomitant epithelial turnover and inflammation in the rat Whatever the explanation, these differences in pathogenesis emphasize the need for cautious inter- pretation of any results and care in the selection of animal species in inhalation studies Nevertheless, most biological responses appear reasonably comparable or can be interpreted

in light of the extensive database and basic understanding of the species differences Because hamsters develop relatively few spontaneous lung tumors and have a high resistance to infection, they were frequently chosen for carcinogenicity studies in the

1960s and 1970s (Heinrich et al., 1986) Hamsters also have been used in chronic studies of emission by-products such as diesel exhaust and most notably, cigarette smoke (Stuart, 1976) As the rat grew in favor as the standard animal for inhalation exposure studies (in large part due to the availability of standardized inbred strains), the hamster was used primarily for the development of animal models of clinical disease (e.g emphysema) and studies of drug-induced lung pathobiology (e.g bleomycin-induced fibrosis) (Gurujeyalakshmi et al., 1998; Qian and Mitzner, 1989) However, the use of hamsters

in studying these disorders with air pollution has been limited, possibly due to the small database on the basic biology and health effects of air pollution in this species

Guinea pigs also have a long history of use in inhalation toxicology Their sensitive bronchoconstrictive response to irritant inhalants and antigens has sustained their use in acute and subacute studies of episodic air pollution exposures (Hatch et al., 1986a, b; Hegele et al., 1993) However, because they grow quickly to adult size, their use has been limited in chronic inhalation studies Like humans, they have eosinophils in bronchoalveolar lavage fluid, and thus their inflammatory responses to inhaled pollutants are comparable to those of humans, although the eosinophilia is somewhat overexpressed (Hernandez et al., 1994) Also, the guinea pig model of ascorbate deficiency closely resembles the human situation because, unlike other rodents, it requires dietary ascorbate supplementation (Hatch et al., 1986b) The unique placement of the guinea pig on the evolutionary tree has raised questions about its appropriateness However, some of its fea- tures make it an ideal species for acute and allergic pulmonary reactions The greatest limitation to the use of the guinea pig is the virtual lack of immunologic reagents if one wishes to conduct state-of-the-art molecular biology studies

Larger laboratory animal species such as the dog have been used in cardiopulmonary

Trang 5

studies of air pollution because their lung structure and size resemble that of humans (Plate 1), and they rarely develop spontaneous lung tumors (Heyder and Takenaka, 1996; Reif et aL, 1970) The dog's basic lung physiology also is known to be similar to that of the human The mechanism by which pulmonary interstitial edema occurs in dogs following myocardial infarction also appears to be much like that in the human (Slutsky et al.,

1983) The hematopoietic system of the beagle dog and its development of humoral and cell-mediated immunity also parallels that of humans (Bloom et al., 1987; Hahn et al.,

1991) Both the size and the anatomy of the dog lung provide advantages for studies of particle deposition, especially when attempting to address questions of size-dependent distal lung injury (Cohen, 1996; Fang et aL, 1993) The dog also has been used to model chronic human bronchitis induced by SO 2 inhalation; the model has been shown to closely resemble the human disease (Drazen et al., 1982; Greene et al., 1984)

For many years, the dog has been used to assess acute and longer-term effects, including radionuclides and particle-induced tumor formation (Heyder and Takenaka, 1996) Recently, the dog also has been employed to study the cardiological effects of environmental particulate exposure to support epidemiological findings in humans (Godleski et al., 1997) However, despite these advantages, the use of the dog in toxicological research has declined over the years because of stricter husbandry regulations, difficulties in pro- curement, greater expense, handling difficulties and the arousal of public sentiment against its use in experiments

Inert dust aerosol deposition and translocation studies in miniature swine have yielded data similar to that for humans This species is similar to humans in many regards, including their size, diet, gastrointestinal tract, skin characteristics, and their long life span (Stuart, 1976) Swine and bovine tissues also can be available for development of in vitro

models, because large amounts of tissue can be obtained from the meat industry Bovine pulmonary parenchyma and artery endothelial cells have been used extensively for studies not only of pollutants, but also of pharmaceuticals (Fukui et al., 1996; Madden et al.,

1987; Ochoa et al., 1997) The equine lung also has been shown to be very similar to the human lung in terms of gross anatomy and intermediate alveolarization of the distal airways Studies have been conducted in horses to evaluate deposition and clearance of radiolabelled particles, and in donkeys to study the effects of inhaled cigarette smoke and

SO 2 on particle transport (Stuart, 1976)

Monkeys are frequently preferred for research on heart diseases, for vaccine development, and for AIDS and anesthesia research (Ghoniem et al., 1996; Petry and Luke, 1997) Monkeys also have been used to study the pulmonary effects of chronic SO 2, acid mists and 0 3 (Alarie et aL, 1975) The monkey has been shown to be particularly useful for studies of the end airway morphometric changes associated with long-term 0 3 exposure (Harkema et al., 1993) Squirrel monkeys exposed to NO 2 have exhibited responses ranging from slight pathology to mortality after challenges with influenza viruses (Henry

et al., 1970) Of all the laboratory animals, the monkey has the lung structure most similar to that of humans For this reason, studies of disease pathogenesis associated with air pollution exposure in this species provide the most convincing data regarding potential human health effects Although the monkey can be a very useful animal model for many pulmonary studies, their current use in research is generally limited to very specific appli- cations such as infectious diseases and drug testing, because of difficulties associated with their availability and expense, husbandry demands, ethics, and the need for large numbers for statistical power

Trang 6

170 U.R Kodavanti and D.L Costa AIR POLLUTION AND A N I M A L MODELS FOR THE S T U D Y OF VARYING

S USCE PTIBI LITI ES

The Clean Air Act of 1970 mandates that regulations for specific air pollutants be suffi- ciently stringent to protect susceptible subpopulations (US Environmental Protection Agency, 1987) Epidemiological studies over the last decade have suggested that children, asthmatics and elderly people with pre-existing cardiopulmonary diseases may be more susceptible to air pollution-induced injury (Dockery et al., 1993; Pope et al., 1995; Schwartz, 1994) This susceptibility concern has triggered interest among toxicologists to seek a better understanding of biologically plausible mechanisms of cardiopulmonary impairments using laboratory animal models of varying susceptibilities Below are descriptions of a number of animal models that reflect susceptible human subgroups and selected studies involving air pollutants

Age

While it might be expected that young and old humans would react differently to air pollutants, most animal studies have not included these potentially susceptible subgroups Inhalation studies typically focus on the effects in young adult rodents as a standardized model There have been only a few studies where oxidant gases such as 0 3, NO 2 and oxygen ( 0 2 ) have been investigated using animal models from different age groups (Montgomery et al., 1987; Weinstock and Beck, 1988) It is apparent from these studies that very young rats are more tolerant to damage induced by oxidant gases (Mauderly et

al., 1987; Mustafa et al., 1985), whereas very young mice appear to be more susceptible (Sherwin and Richters, 1985) Older rats, on the other hand, have been shown to be more susceptible to O3-induced lung injury (Stiles and Tyler, 1988; Vincent and Adamson, 1995) Some of these studies have yielded equivocal results, thus, there remains controversy in the assessment of age sensitivity, especially with regards to longer- term outcomes It is presumed but not confirmed that structural attributes of the lung and the inductiveness of antioxidative mechanisms play critical roles in determining the oxidant resistance of neonatal and young rodent models (Mustafa et al., 1985; Tyson et al., 1982)

A study reported by Mauderly et al (1987) in which rats were exposed to diesel particles during maturation (from birth through weaning) suggested that adults were more susceptible than the young in terms of degree of pulmonary injury, the efficiency of lung clearance of radiolabeled particles, and collagen accumulation However, it is not known whether young mice respond differently to these pollutants than they do to 0 3 The particle deposition patterns can vary with the stage of lung development, since alveolarization and dimensional changes occur in the respiratory tree after birth in both laboratory animals and humans (Weinstock and Beck, 1988) Age-related susceptibility also may depend upon the type of pollutant and the affected target, since site-specific cellular pro- teins are modified during development and the spectrum of gene expression evolves Since the incidences of spontaneous cancer are increased in most aging animals and humans, it is likely that older animals may show different responses in terms of the carcinogenic effects of inhaled pollutants The knowledge available from existing studies

Trang 7

on young adult animals provides an important reference point when deciding what dose levels should be used for comparable studies in different age subgroups

G e n d e r

Because of basic physiological differences between men and women, pollutant health effects may differ (Beck and Weinstock, 1988) Comparative studies involving animal models representing both genders can help us understand whether differences in responsiveness are due to these innate biological dissimilarities For this reason standardized bioassays such as those regulating test requirements incorporate both genders However, past studies of air pollutants have only occasionally included males and females Since human gender differences are known to exist in the case of cigarette smoke and 0 3 effects, the limited prospective provided by animal studies is unfortunate (Beck and Weinstock, 1988; Bush et al., 1996; Seal et al., 1993) It appears from the epidemiology

of cigarette smoke health effects that differences in relative lung size and density of tracheobronchial mucus secretory cells between men and women play a critical role in their susceptibility to smoke-induced bronchitis (reviewed in Beck and Weinstock, 1988) Differences in the structure of the lung and its subcomponents could influence deposition and clearance of inhaled substances in a gender-related manner In general, epidemiological findings suggest the fact that females are more resistant to the harmful effects

of chronic cigarette smoke inhalation than males (Enjeti et al., 1978; Tager and Speizer,

1976) However, acute 0 3 exposure studies designed to evaluate gender differences with regard to effects on forced vital capacity and forced expiratory volume in 1 second, tidal volume and breathing frequency have failed to reveal any differences between men and women (Messineo and Adams, 1990) Analogously, the chronic effects of 0 3 in Fischer

344 male and females rats appear to be similar with regard to resultant pathology and functional changes (Stockstill et al., 1995) The differences in chronic responsiveness of men and women to cigarette smoke and lack thereof in those humans and animals exposed to 0 3 may reflect the complexity of cigarette smoke and its component effects on various cell types of the lung

When tracheal epithelial cells from male and female rats are exposed to cigarette smoke,

it has been shown that cells from females secrete more mucus than those from males Mucus production may relate to more efficient removal of harmful smoke particles from conducting airways and therefore the reduced vulnerability to bronchitis seen in females (Hayashi et al., 1978) This hypothesis has not been tested experimentally using in viva

animal models The production of mucus has been shown to be influenced by hormonal changes in females, e.g postmenopausal women do not exhibit estrous cycle-related changes in the mucus-secreting cells of the airways (Chalon et al., 1971) It is possible that with some pollutants or animals species, the gender-related differences may not be significant enough to make adjustments in regulatory decisions, however, understanding these differences is critical in making meaningful evaluations of health risk

Pregnancy, which can be considered a temporary physiological condition in females, may result in increased susceptibility to pollutant-induced pulmonary injury Associated with pregnancy are fetal growth and development, which may be directly influenced by the pollutant or indirectly affected through decrements in the pulmonary health of the mother Developmental effects of inhaled pollutants, especially 0 3 , have been investigated

Trang 8

172 U.P, Kodavanti and D,L Costa

in mice (Bignami et al., 1994) Moderate effects on selected neurobehavioral tests were noted in newborn mice when dams were exposed to 0 3 during pregnancy and the neonatal period 0 3 also has been shown to be more toxic to pregnant and lactating rats when compared with age-matched non-pregnant controls (Gunnison et al., 1992; Gunnison and Finkelstein, 1997) CO is also well studied in terms of the mechanism by which early fetal mortality and low birthweight occurs in exposed individuals, especially smokers (Acevedo and Ahmed, 1998; Seker-Walker et al., 1997) Studies have shown that placental blood flow can be compromised by cigarette smoke It is not yet known if some of these effects are nicotinic or secondary irritant responses (Economides and Braithwaite, 1994) The limited data on air pollution effects during pregnancy warrants further investigations

to evaluate possible health effects of pollutants in pregnant animal models and develop- ing fetuses

S p e c i e s / S t r a i n

A variety of laboratory animal species and strains are used in the assessment of pollutant- induced pulmonary health effects Selection of an appropriate animal species may depend largely upon how relevant it is to the human, because ultimately extrapolation is required

to make a fair evaluation of the human health risks of air pollutants (Brain et al., 1988b; Warheit, 1989) It is recommended that the response of the selected animal model species

to the test material is similar to that of the human (Weil, 1972) In the case of inhaled pollutants, the deposition, clearance, metabolism, absorption, storage, and other potentially species-based physiological aspects of the animal model should be appreciated However,

in most instances our knowledge on all these aspects is not available a priori Ideally, it is recommended that more than one species be used for the initial characterization of any toxic response (Brain et al., 1988b) The obvious disadvantage to using multiple species

is the greater time, effort and expense that is required However, this disadvantage should

be outweighed by the information that can be provided for improving the risk assessment process, and better understanding of the pathobiological mechanisms of injury and disease (Brain et al., 1988b)

There are marked morphological and morphometrical structural differences between the human and laboratory animal lung (reviewed in Warheit, 1989) The branching patterns of the conducting airways differ: human lungs are dichotomous and essentially symmetrical, while in non-primate animals the lungs are highly asymmetrical and monopodal As a result, particle deposition in humans is less uniform, while in animals

a more uniform distribution is achieved (Lippmann and Schlesinger, 1984; Phalen and Oldham, 1983) There are also marked differences in airway structures among animal species that may influence the impact of inhaled materials For example, the distal airway structures of dogs, cats and macaque monkeys are somewhat similar to those in humans

in having respiratory bronchioles; however, in most small rodents the terminal bronchioles terminate directly into alveolar units (Tyler, 1983) Marked species differences are also apparent in the cell populations distributed throughout the lung (Phalen et al., 1989) In the sheep, mucus goblet cells are the predominant secretory cells, but in the mouse, Clara cells are the primary secretory cells (Hopper, 1983)

Descriptions of each animal species and strains that have been used in studies of air pollution are beyond the scope of this chapter The salient issues related to species differences

Trang 9

in the health effects of air pollutants have been presented by Brain et al (1988b) An excel- lent review comparing the lung responses to inhaled particles and gases across many commonly used species has been provided by Warheit (1989) Some examples are presented below in the context of varied species and strain-related responses

Rodent models of genetic or strain-related susceptibility to 0 3 have provided an important tool for understanding the biological basis of variable individual responses to environmental pollutants (Kleeberger, 1995) The mouse strain C57B6 is susceptible to O3-induced neutrophilic inflammation, whereas the C3H/HeJ is not It has been pre- dicted that single gene inheritance at chromosomal locus I n ~ s responsible for the differing susceptibility between these strains (Kleeberger, 1995) Recently we have shown that combustion particles are fibrogenic and cause fibronectin gene expression in Sprague- Dawley rat, while the Fischer 344 rat is relatively less sensitive (Kodavanti et al., 1997a) Species differences have also been noted in a number of studies in terms of their pulmonary antioxidant pools and their responsiveness to air pollutants (Hatch et aL, 1986a; Hatch, 1992) It is likely that the observed differences may reside in genetic strain-related susceptibility or resistance in these rats The more understanding we have about the species-associated genetic susceptibilities of commonly used laboratory animal models and humans, the better our extrapolations will be

N u t r i t i o n

Since most air pollutants induce injury through oxidative mechanisms, nutritional deficiencies of antioxidant vitamins constitute a major concern regarding susceptibility (Colditz et al., 1988; Hatch, 1995; Menzel, 1992; Pryor, 1991; Shakman, 1974) Table 10.1 summarizes notable studies of air pollution health effects in nutritionally compromised animal models Animal models of altered nutritional status can be produced in most instances by dietary manipulation In the guinea pig, a deficiency in vitamin C - a critical pulmonary antioxidant - can be achieved by feeding a deficient diet for 2-3 weeks, since guinea pigs, like humans, cannot synthesize their own vitamin C (Hatch et al., 1986b) However, this is not the case for vitamin C deficiency in the rat, since rats produce endogenous vitamin C A rat model of vitamin E deficiency can be developed by dietary restriction (Chow et al., 1979; Goldstein et al., 1970) It has been shown that 0 3 and NO2-induced pulmonary injuries are exacerbated in vitamin C deficient guinea pigs, especially at relatively low concentrations of 0 3 (Kodavanti et aL, 1995a, b, 1996a; Slade

et al., 1989) Similarly, vitamin E deficiency in the rat has been associated with greater pulmonary injury from 0 3 (Goldstein et aL, 1970; Sato et aL, 1976) and from N O 2 (Ayaz and Csallany, 1978; Elsayed and Mustafa, 1982; Menzel, 1979) at relatively lower concentrations This lipid-soluble membrane-bound antioxidant is thought to be critical in scavenging lipid peroxides produced by free radicals at the lung's surface (Hatch, 1995; Pryor, 1991)

Vitamin A is important in the maintenance, differentiation and proliferation of epithelial cells, activities common to both normal lungs and during injury (Takahashi

et al., 1993) Severe vitamin A deficiency alone has been shown to cause bronchiolitis and pneumonia in diet-restricted animal models (Bauernfeind, 1986) Decreased label- ing of alveolar and bronchiolar epithelial cells have been reported following 0 3 exposure

in rats deficient in vitamin A (Takahashi et al., 1993) Similarly, a rat model of dietary

Trang 10

174 U.R Kodavanti and D.L Costa

Table 10.1 Air pollution studies using rodent models of nutritional manipulation a

Pulmonary injury from 0.1 ppm 03

No effect from longer-term 03 Increased lipid peroxidation

03

NO 2 NO2

Mortality due to 03 and no increase

in lysyl oxidase and collagen synthesis in vitamin B6-deficient rats

Increased epithelial damage

vitamin B 6 deficiency has shown impaired collagen cross-linking following 0 3 exposure The mechanism appears to involve the role of vitamin B 6 in the action of lysyl oxidase,

1986)

Glutathione (GSH) deficiency has been achieved by treating rats with buthionine

levels of GSH can also be depleted by treating animals with diethyl maleate without affecting its synthesis pathway Isolated perfused lungs from an animal injected with

(GSH)-deficient rats have been shown to be more susceptible to O3-induced fibrosis, suggesting that GSH may function as an antioxidant in oxidant-induced lung injury

Selenium is another essential nutrient which has been shown to affect the toxicity of

and the removal of free radicals, has four selenium residues that are critical to enzyme

Trang 11

activity It has been shown that selenium-deficient rats are more susceptible to O3-induced

lung injury (Eskew et al., 1986) Likewise, the impact of 0 3 on the activities of GSH

recycling enzymes in the lungs of mice can be directly influenced by dietary selenium defi-

ciency (Hsayed eta/., 1983) It is reasonable to assume that many other pollutants may act

in an analogous manner; however, more studies are needed to understand the role of selenium in oxidative lung damage

Nutritional deficiencies in the human population are common and alone may constitute a major modifier of air pollutant responsiveness (reviewed in Hatch, 1995) Inner city populations under socioeconomic stress are often deficient in many essential antioxidants, while also enduring the highest potential for exposure to urban pollution However, little attention has been paid to the potential health risks of air pollutants and the influence of nutritional status How nutritionally deficient animal models can provide data for such consideration remains to be determined

P h y s i o l o g i c a l S t a t u s

Exercise can directly increase three major cardiopulmonary parameters: pulmonary ven-

tilation (tidal volume and respiratory rate), cardiac output and blood flow (Brain et al.,

1988c) It can also profoundly affect both total pollutant deposition and pollutant distribution at sites in the lung Most humans, and especially children, are likely to be more heavily exposed to toxic materials than has been appreciated because of varied activity profiles or work-related exercise Thus, exercise is typically used in human exposure

studies (e.g acute 0 3) to attain maximal sensitivity for detecting effects (Devlin et al., 1991; Koren et al., 1989) Exercise imposition has been occasionally applied to rodent

studies of air pollutants, with the bulk of the data coming from studies evaluating changes in the deposition pattern of particles and gases associated with altered breathing

patterns (Brain et al., 1988c; Hatch et al., 1994; Mautz et al., 1988; Tepper et al., 1990,

1991)

It has been shown that the deposition of inhaled 35SO2 in the nose and along the airways of dogs can be profoundly affected when flow rate is increased (reviewed in Brain et

the amount of 35SO2 presented to the trachea When the increased flow rate was directed through the mouth, as would be expected in exercising humans, the penetration increases

to 660-fold (Frank et al., 1969; Brain 1970) Syrian hamsters similarly have been shown

to deposit 2-3 fold more 99mTc aerosols when minute ventilation (i.e oxygen consumption) is doubled on an exercise wheel (Harbison and Brain, 1983) Other experimental manipulations have been attempted in animal studies to mimic the increased ventilation of exercising humans (e.g increased CO 2 concentrations in the breathing air)

(Hatch et al., 1994; Tepper et al., 1990) These and other studies support the contention

that exercise or elevated ventilation can be a significant contributor to susceptibility to air pollutants This effect needs to be evaluated further to ascertain the potential risk that physical activity may pose for humans It should be understood, however, that putative susceptible human subgroups (e.g those with chronic cardiopulmonary ailments) are less likely to have high exercise profiles

Trang 12

176 u.P Kodavanti and D.L Costa Respiratory disease models

Epidemiological evidence linking air pollution to increased mortality and morbidity in chronic obstructive pulmonary disease (COPD) patients and increased hospitalization among asthmatics supports the belief that pre-existent disease imposes an increased risk

of air pollutant health effects (Dockery et al., 1993; Pope et al., 1995; Schwartz, 1994) The collective epidemiological evidence for effects in both healthy and susceptible groups has led the Environmental Protection Agency to revise the current National Ambient Air Quality Standard (NAAQS) for PM air pollution, since effects have been observed at or below the levels previously thought to be safe (McClellan and Miller, 1997) Critics of this decision suggest that the observations lack biological plausibility and that the animal toxicity database is neither sufficient nor confirmatory enough to support the revised standards (McClellan and Miller, 1997; Vendal, 1997) While the health effects of air pollutants are noted specifically in susceptible individuals (such as those with pre-existent cardiopulmonary diseases), a question has been raised about the validity of using toxicity data derived from healthy animals to make risk estimations Although the susceptibility

of individuals with pre-existent disease to the adverse effects of pollutants is generally acknowledged, less attention has been paid to the use of animal disease models in addressing this issue because of the complexity associated with variability in responses and extrapolation of data to humans (Kodavanti et al., 1998) In this section, we provide an overview/description of those rodent respiratory disease models that have been employed

or that have the potential for use in studies of air pollutant susceptibility We then look

at their actual application in air pollution toxicity studies Finally, we address several issues associated with disease models use, and consider the future of research in this field Disease models can be developed by chemical, surgical or genetic manipulation of the animal species Many animal models of respiratory disease are well established and relevant reviews and papers have been published dealing with criteria, pathogenesis and their limited use in air pollution studies (Gilmour and Koren, 1999; Kodavanti et al., 1998; Kumar, 1995; Reid, 1980; Slauson and Hahn, 1980; Sweeney et al., 1988) A recently published review paper focuses on a broad spectrum of rodent models of cardiopulmonary diseases (Kodavanti et al., 1998) Below is a brief description of several well-established and characterized respiratory disease models that may be useful for the study of air pollution susceptibility Selected examples of air pollution health effect studies using animal models of cardiopulmonary disease are given in Table 10.2

A s t h m a / A l l e r g y

Asthma is a complex obstructive lung disease characterized in humans by reversible bronchoconstriction, mucus production, airways inflammation, and hyperreactivity to pharmacologic bronchoconstrictor agonists (McFadden and Hejal, 1995) Frequently there is an allergic component to the disease (atopy), which may be reflected in increased serum IgE antibody titer (Arm and Lee, 1992; Frew, 1996) The incidence of asthma in humans has been rising in developed countries and is a major health concern worldwide (Cookson and Moffatt, 1997) Asthma typically is episodic but can lead to chronic lung disease and, when triggered acutely, it may even be life-threatening It is believed that environmental and genetic interactions are involved in expression (initiation and elicitation)

Trang 14

178 U.R Kodavanti and D.L Costa

of the disease Several genes have been identified in humans that are linked to asthma susceptibility (Kim et al., 1998; Ober, 1998)

Animal models of asthma have been widely characterized and used in research (Cassee and van Bree, 1996; Gilmour, 1995; Karol et al., 1985; Selgrade et al., 1997; Selgrade and Gilmour, 1994; van Loveren et al., 1996) The most popular models of laboratory rodents include ovalbumin sensitized and subsequently challenged mice, Brown Norway rats and guinea pigs (Elwood et al., 1991; Gilmour, 1995; Selgrade et al., 1997; Selgrade and Gilmour, 1994) The model development protocols in general include initial intraperi- toneal or subcutaneous injection of an allergen (ovalbumin in most cases) in the host (sensitization) followed by (2-3 weeks later) large inhalation challenge (elicitation) with the allergen (Gilmour, 1995; Selgrade and Gilmour, 1994) A house dust mite allergen has also been used successfully to develop allergic rat and mouse models (Cheng et al., 1998;

Gilmour and Selgrade, 1996) One or several challenge doses of allergen increases the expression of pulmonary T helper cell 2 (TH2) cytokines and eosinophilic as well as neutrophilic airways inflammation, injury and airway hyperresponsiveness (Selgrade et al., 1997) Serum and bronchoalveolar lavage fluid levels of IgE are also increased (Selgrade and Gilmour, 1994) These acute changes are analogous to human asthma and may per- sist for one to several weeks, depending on the sensitization and challenge protocols used Large animal models have also been developed (such as the dog and the sheep) using biological antigens, e.g Ascaris protein (Matsumoto and Ashida, 1993; Reynolds et al., 1997) However, these large animal models are generally limited to mechanistic studies because of the size of the animals and hence the limitation in sample numbers

Each asthma model in air pollution studies has its own advantages and disadvantages The mouse models are amenable for immunological and molecular investigations because

of the variety of probes that are available, but unlike humans, their bronchoconstriction response is weak (Corry et al., 1996) Guinea pigs, in contrast, have a strong bronchoconstrictive response to pharmacologic agonists, but the limited availability of guinea pig-specific immunological and molecular probes is problematic (Savoie et al., 1995) Use of the Brown Norway rat is growing since it develops eosinophilia and airway hyperresponsiveness after sensitization and challenge; both are analogous to the reactions in atopic human asthmatics (Bellofiore and Martin, 1988) Access to probes is not

as common as for the mouse, however The use of the mouse and other asthma models is limited to acute studies, because none has yet developed the 'chronicity' of airways inflammation, fibrosis, and recurrent congestion that exist in humans A number of studies have been done of the interactions of environmental pollutants with allergen sensitization and the relative susceptibilities of these rodent asthma models (Gilmour, 1995; Selgrade and Gilmour, 1994) However, much remains to be investigated in terms of biological mechanisms

et al., 1987, 1995), although guinea pigs and hamsters have also seen limited use

Trang 15

(Goldring et al., 1970; Ito et al., 1995) Exposure of rats and dogs t o SO 2 at concentrations of 200-600 ppm, usually 4-6 h/day, five days a week for a total of 4-8 weeks, has been shown to induce bronchitic lesions that resemble the human pathophysiology

(Drazen et al., 1982; Shore et al., 1995) Although the dog respiratory physiology and

structure perhaps more closely resembles that of humans, SO 2 causes a decrease in airway responsiveness to bronchoconstrictor agonists, while in humans and in rats airways

responsiveness is increased with the disease (Shore et al., 1995)

Tracheobronchial submucosal glands and epithelial goblet cells contribute to the mucus layer in humans, in contrast to the rat in which Clara cells and those limited goblet cells

that exist produce the mucus (Goco et al., 1963) With initial challenge, the goblet cells proliferate and contribute the bulk of the excess (Basbaum et al., 1990) Unlike that in the

human, the lesion in the rat readily reverses when SO 2 is removed (Snider, 1992a; US Department of Health, Education and Welfare, 1969) Although some improvement in bronchitic patients can be achieved by removal of the causative agents (e.g smoking), the time required for reversal can extend to several months or years, and in humans can be confounded by infection as well as by frequent use of medication (Wilson and Reyner,

regard, the reversibility aspect should be carefully considered when selecting these models

to study air pollution susceptibility, especially with longer-term pollution exposures Repeated exposure to lipopolysaccharide (endotoxin) in rats results in bronchitis analogous to human disease in terms of mucus hypersecretion, airways inflammation, fibrosis and epithelial cell hyperplasia (Harkema and Hotchkiss, 1992) The essential difference between SO 2 and endotoxin models is the degree Of airway inflammation (greater in endotoxin model) and mucus hypersecretion (greater in sulfur dioxide model) The guinea pig model of polymyxin B-induced bronchitis has been characterized as primarily

eosinophilic inflammation but the goblet cell pathology is not well defined (Ogawa et al.,

Emphysema

In humans, emphysema is most often associated with cigarette smoking and the genetic deficiency of 0~-1 antiprotease (Surgeon General, 1984) Exposures to cadmium, NO 2 and hyperoxia as well as starvation and copper deficiency, can also induce emphysema-like disease, but there are pathological differences based on the type of injury (Snider, 1992b) The most common form of emphysema in humans is centriaciner, with the disease being centered in bronchio-alveolar duct junctions consistent with a pattern of inhaled toxicant deposition (Snider, 1992b) In contrast, emphysema associated with 0~-1 antiprotease deficiency and elastase overload appears as permanent enlargement of airspaces distal to the

terminal bronchioles and is described as panlobular emphysema (Snider et al., 1986)

Destruction of alveolar walls and airspace enlargement constitute the morphological hall- marks of emphysema regardless of its distribution In laboratory animals, emphysema can

be readily produced by a single intratracheal exposure to elastase It most closely resem-

bles the panlobular disease (Snider et al., 1986)

Purified preparations of bovine pancreatic elastase have been commonly used to produce emphysemic animals, although reports utilizing human neutrophil or pancreatic

elastase, or even papain can be found (Snider et al., 1986) Hamsters have been used

widely to establish the elastase-induced emphysema model because of their sensitivity to

Trang 16

180 U.P Kodavanti and D.L Costa

elastolytic enzymes However, rats, mice and rabbits have also been frequently employed

in this regard (Deschamps et al., 1995; Karlinsky and Snider, 1978; Snider et al., 1986)

Intratracheal instillation of elastase results in acute alveolitis characterized by degradation

of elastin, rupture of alveolar epithelium, pulmonary edema, hemorrhage, and inflammation followed by airspace enlargement The initial inflammation resolves within a few weeks, followed by synthesis of new elastin and collagen; however, the distortion and derangement of alveolar structures are permanent (Snider, 1992b)

Spontaneously occurring genetic models of emphysema are also available The emphysema of the tight-skin and pallid mice results from a genetic deficiency in 0~-1 antiprotease (de Santi et al., 1995; Szapiel et al., 1981) Progressive overload of elastolytic activity can

be observed from the first months of life Unlike the exogenous elastase model, total lung elastin appears to decrease over time and inflammation is not evident (as determined by lavage fluid analysis) This model may be a choice in determining emphysema-associated susceptibility without coexistent inflammation Another strain, the blotchy mouse, also exhibits panlobular emphysema but its genetic impairment is in copper absorption which

is critical for lysyl oxidase, an essential connective tissue cross-linking enzyme (Ranga et al., 1993) This model of emphysema may be unrelated to that in humans, since it does not involve 0~-1 antiprotease dysfunction which is thought to be critical to pathogenesis

in humans As with the bronchitis model, a few air pollution studies have been conducted using emphysema animal models Those studies involving emphysema models generally show little difference from healthy controls The lack of underlying inflammation in those models may partly explain their lack of sensitivity

Chronic Obstructive Pulmonary Disease (COPD)

Chronic obstructive pulmonary disease represents a group of related respiratory diseases presenting with chronic airways obstruction, mucus hypersecretion, chronic inflammation and injury to the pulmonary airways and parenchyma (O'Connor et al., 1989; Repine et al., 1997) Right ventricular hypertrophy and cot p u l m o n a l e often develop as result of abnormalities in the blood flow through damaged or remodeled and diseased pulmonary vasculature (MacNee, 1994) Typically the term COPD is used to refer collectively to chronic bronchitis and emphysema (Higgins and Thom, 1989) These patients suffer frequent and persistent infections because of their inability to clear infectious agents from the respiratory tract (Jansen et al., 1995; White, 1995) In humans, COPD has been associated with chronic cigarette smoking, chronic asthma or even recurrent infection (Cugell, 1988) Although animal models of bronchitis, emphysema, asthma and airways infections are well characterized as separate entities, there is no one model available that mimics this pathophysiology Occasionally, rodent models of pulmonary injury/vasculitis and bronchitis are presented as COPD models; however, neither adequately represents the typical COPD characteristics in terms of both pathology and physiology

Pulmonary Fibrosis

Idiopathic pulmonary fibrosis is relatively rare in humans; however, occupational exposure

to various metals, mineral and organic dusts as well as certain drugs have been associated with the disease (Crouch, 1990) The pathogenesis of the disease is characterized by initial pulmonary injury and/or inflammation, and by concomitant repair and subsequently

Trang 17

fibrogenesis The pattern of fibrotic lesions depends upon the causative agent, the type of insult, and the route of exposure (Hepleston, 1991) A number of animal models involving drug-, chemical-, or mineral dust-induced fibrosis have also been described, however, usually at high doses (Crouch, 1990; Hay et al., 1991; Hepleston, 1991) Chronic endotoxin exposure has also been reported to result in airways fibrosis in animals (Harkema and Hotchkiss, 1992) Depending upon the study objectives, models can be selected and produced simply by inhalation, intratracheal or even oral exposure of animals to drugs or mineral dusts

The most extensively used pulmonary fibrosis model is that induced by bleomycin Bleomycin, an antineoplastic agent that can be administered systemically or by intratracheal instillation, results in development of alveolar fibrosis in 2-3 weeks in most laboratory animals; the rat and hamster are the most widely used models (Crouch, 1990; Hay et al., 1991; Raghow et al., 1985) The pathogenesis in these models includes initial lung injury, inflammation, upregulation of number of cytokines, fibroblast proliferation and hypertrophy (2-10 days) (Raghow et al., 1985) The initial inflammatory response begins to regress after this period, with a resolution of edema and increased collagen deposition in the lung parenchyma (Hernnas et al., 1992) We and others have used this model in determining particulate-induced lung injury (Adamson and Hedgecock, 1995; Adamson and Prieditis, 1995; Kodavanti et al., 1996b)

Respiratory Infection

Respiratory infection is a common human ailment and is frequently a complicating, if not

a direct contributor to most respiratory-associated deaths (Babiuk et al., 1988; Murray and Lopez, 1997; Wilson, 1997) Most infections are self-limiting and are cleared up quickly in healthy individuals (Wells et al., 1981) The outcome or severity of the infection depends on the virulence of the infectious microorganism and the susceptibility of a host For example, in patients with AIDS, cystic fibrosis or COPD, respiratory infections clear only slowly and are recurrent (Jansen et al., 1995; Waxman et al., 1997;

White, 1995) Although animal models of bacterial and viral infection have been used in

a variety of studies, their major limitation is that the organisms that infect humans have variable species-dependent virulence in the animal models And since these infections are often cleared very quickly in animals, their relevance to compromised humans where infections are long lasting is questioned (O'Reilly, 1995)

Acute airways viral infections cause epithelial necrosis, increased bronchial epithelial and endothelial permeability, and inflammatory cell influx T helper cell 1 (TH1), and in some infections TH 2, are thought to be involved, as the production of relevant cytokines

is increased (Holtzman et al., 1996; Cookson and Moffatt, 1997) Mice and rats are the rodents most frequently used to develop viral infection models through inoculation of infectious viruses in the airway (Green, 1984; Lebrec and Burleson, 1994) A highly virulent and lethal influenza strain (influenza A/Hong K o n g / 8 / 6 8 , H3N2 virus) adapted

to B6C3F1 and CD mice, a less virulent strain (A/Port Chalmers/1/73, H3N2) adapted

to CD mice, and a similar non-lethal strain adapted to rats have all been used in the development of influenza infection models in mice and rats (Lebrec and Burleson, 1994) The lethal mouse model terminates in extensive pneumonia and lung consolidation, thereby limiting its relevance to detailed study of infectious disease; however, the non- lethal rat and mouse models exhibit airways epithelial damage, inflammation and

Định dạng
Số trang	35
Dung lượng	7,55 MB