Particle Toxicology - Chapter 19 ppt

19 Air Pollution and Human BrainPathology: A Role for Air Pollutants in the Pathogenesis of Alzheimer’s Disease Lilian Caldero´n-Garciduen˜as The Center for Structural and Functional Neu

19 Air Pollution and Human Brain

Pathology: A Role for Air

Pollutants in the Pathogenesis

of Alzheimer’s Disease

Lilian Caldero´n-Garciduen˜as

The Center for Structural and Functional Neurosciences,

University of Montana

William Reed

Department of Pediatrics and Center for Environmental Medicine,

University of North Carolina at Chapel Hill

CONTENTS

19.1 Introduction 331

19.2 Molecular Basis of Alzheimer’s Disease Pathogenesis 332

19.3 Alzheimer’s Disease Pathogenesis and COX2 334

19.4 The Mexico City Environment 334

19.5 COX2 and IL-1b Expression, Ab42 Accumulation, and Neuropathology in the Brains of Dogs and Humans Exposed to Severe Urban Air Pollution 337

19.6 Clinical Studies of Mexico City Children 341

19.7 Potential Mechanisms of Air Pollutant-Induced Inflammation and Neurodegeneration 343

19.7.1 Air Pollutant-Induced Systemic Inflammation 343

19.7.2 Transport of PM-Associated Metals to the Brain 343

19.7.3 LPS Toxicity 345

19.8 Summary 346

Acknowledgments 346

References 346

19.1 INTRODUCTION

Adverse health effects associated with chronic exposures to air pollutants (indoor, outdoor, and occupational settings) are an important issue for millions of people around the world As the world population becomes older, significant increases in neurodegenerative diseases such as Alzheimer’s have been projected over the next decades (Brookmeyer, Gray, and Kawas 1998; Hebert et al 2003) Alzheimer’s disease (AD) is an irreversible, fatal brain disorder that presently affects 4.5 million people in the United States and it is projected that it will affect between 13 and 16 million by

331

2050 (Brookmeyer, Gray, and Kawas 1998; Hebert et al 2003) Alzheimer’s patients have a major medical, social, and economic impact, thus any factors that could modify these projections need to

be pursued and integrated into multidisciplinary studies of AD The role played by the environment

in the pathogenesis of AD is unclear (Brown, Lockwood, and Sonawane 2005) Our findings suggest that exposures to significant levels of particulate matter and photo-oxidants may accelerate the appearance of precursors of Alzheimer’s disease in sentinel animals and in humans

In this chapter, we will review the pathophysiology of AD as it is currently understood and summarize comparative pathology, human neuropathology, and clinical studies of residents of cities with significant chronic concentrations of particulate matter, endotoxins, ozone, and a myriad of other air pollutants We discuss how air pollutants might promote AD indirectly by causing systemic inflammation or directly by causing brain injury following their entry into the brain via known pathways

19.2 MOLECULAR BASIS OF ALZHEIMER’S DISEASE PATHOGENESIS

Alzheimer’s brains exhibit two pathological hallmarks: (1) the accumulation of b-amyloid peptides (Ab) in the extracellular space in the form of neuritic plaques and (2) intraneuronal filamentous tangles (neurofibrillary tangles, NFTs) containing hyperphosphorylated tau protein (reviewed in Selkoe 2001) Ab peptides are 37–43 amino acid proteolytic fragments of b-amyloid precursor protein (APP), an alternatively spliced transmembrane protein expressed by all cells Normal neurons primarily express the 695 amino acid form of APP Ab peptides are generated by the proteolytic cleavage of APP by two proteases, b- and g-secretase (Selkoe 2004) Neuritic plaques are foci of extracellular Ab deposition that are associated with axonal and dendritic injury (Selkoe 2001) A large part of the fibrillar Ab found in plaques is the 42 amino acid-isoform (Ab42) that is more hydrophobic and prone to aggregation than other Ab isoforms Precursor lesions of neuritic plaques are referred to as “diffuse plaques” (Selkoe 2001)

Genetic studies of familial AD (FAD), an inherited, early onset form of AD, suggest that the generation of Ab42 plays a contributory role in AD pathogenesis Mutations in any of three genes, APP, presenillin-1, and presenillin-2 cause a specific increase in Ab42 generation that is associated presenillin with FAD (Scheuner et al 1996) All three genes might be expected to regulate Ab42 generation, as APP is the precursor of Ab peptides and the presenilins are essential components of g-secretase (reviewed in Selkoe and Kopan 2003) An increase in APP gene dose as occurs in Down’s syndrome (3 copies of chromosome 21 rather than 2) is associated with early onset AD as well (Lemere et al 1996) Finally, a major risk factor for the development of AD is the epsilon 4 polymorphism of the apolipoprotein E gene (Corder et al 1993), whose protein product enhances Ab42 stability and accumulation (Strittmatter et al 1993; reviewed in George-Hyslop 2000) These findings support the amyloid cascade hypothesis of AD pathogenesis

The amyloid cascade hypothesis postulates that increased production of the Ab42 results in its accumulation and oligomerization in limbic and association cortices leading to the gradual depo-sition of Ab42 oligomers as diffuse plaques and to subtle effects of the oligomers on synaptic efficacy These early changes are believed to cause microglial and astrocytic activation, prolifera-tive responses in microglia (gliosis), altered neuronal ionic homeostasis, and widespread synaptic dysfunction and neurodegeneration (Selkoe 2001; Cleary et al 2005) In this hypothesis, an increase in the generation of Ab42 is the essential and self-propagating pathogenic event in AD

An alternative pathogenic mechanism proposes that AD is a consequence of a failure of the control of neuronal differentiation (reviewed in Arendt 2003; Nagy 2005; and Webber et al 2005)

In general, brain neurons rest in a postmitotic, differentiated state However, adaptive behaviors such as learning and memory require that synapses are continually lost or formed and strengthened

or weakened in a remodeling process that relies upon cellular repair machinery It is hypothesized that normal synaptic turnover together with genetic factors that cause instability in the repair

Particle Toxicology 332

machinery or external factors that result in neuronal injury (e.g., oxidative stress and inflammation)

or both could result in excessive demand upon the repair machinery and increase the risk of dedifferentiation with subsequent cell cycle activation (Arendt 2003)

Alternatively, cell cycle activation in neurons could be mediated by inappropriate mitogenic stimuli caused by altered expression or mutation of molecular components of signal transduction pathways that regulate the cell cycle Although the function of APP is still poorly understood, limited evidence suggests that APP may regulate neuronal survival (Koo and Kopan 2004) by signaling to the nucleus, a process that probably involves its proteolytic cleavage by b and g secretases Thus, overexpression of APP and mutations in APP and presenilins might disturb neuronal cell cycle arrest initiating AD pathogenesis

Reentry into the cell cycle is believed to be a natural consequence of aging, however, in normal aging, brain neurons arrest in G1 phase In contrast, AD neurons appear to undergo DNA replication (S phase) (Yang, Geldmacher, and Herrup 2001) and become trapped G2 as suggested by the aberrant expression of G2-specific cell cycle regulating proteins in AD brain (Nagy 2005) Although the mechanism of arrest in G2 is unclear, it presumably occurs because neurons are incapable of completing the cell cycle by undergoing mitosis In this alternate hypothesis, the apparent failure

of the G1/S transition checkpoint is believed to be the essential pathogenic event in AD

G2 phase is characterized by the activation of specific cyclin-dependent kinases (cdk) that ensure the progression through G2 to mitosis The activation of G2-specific cdks is associated with a gradual destabilization of the microtubule (MT) cytoskeleton In neurons, MTs are essential for axonal transport Thus, cell cycle activation is a plausible cause of the axonopathy and transport deficits that are observed early in AD pathogenesis Moreover, transport deficits inhibit anterograde and retrograde transport of APP, causing the buildup of APP in the neuronal cell body and an increase in Ab42 generation (Stokin et al 2005)

Neurons in AD brain upregulate the expression of protein inhibitors of the cell cycle, such as glycogen synthase kinase-3 (reviewed in Bhat, Budd Haeberlein, and Avila 2004) and cyclin-dependent kinase inhibitors (Arendt et al 1996; reviewed in Nagy 2005) It is suggested that this phenomenon is caused by the cellular stress incurred by the inability of neurons in G2 to undergo mitosis and is presumably intended to inhibit progression to mitosis However, there are likely side effects of these changes in gene expression For example, glycogen synthase kinase-3 could phosphorylate the soluble tau protein, which has been released from depolymerizing MTs, thus promoting development of NFTs Indeed, a number of stress-activated kinases are capable of phosphorylating tau protein (Lovestone and Reynolds 1997) and may also promote the development NFTs

In the alternative hypothesis, the pathogenic event is irreversible, because the neuron is left with

no mechanism for progressing to mitosis or returning to G0 phase Rather it progressively degen-erates as axonopathy and transport deficits worsen and Ab42 and hyperphosphorylation of tau protein build up eventually forming neuritic plaques and NFTs

The alternative hypothesis of AD pathogenesis introduces a mechanism by which external events, such as exposure to air pollutants, could affect AD pathogenesis Perturbations of the neuronal microenvironment, such as toxicant-induced oxidative stress or inflammation, could cause enough damage to elicit reentry into the cell cycle In other words, damage to a neuronal network that retains synaptic remodeling ability could present a mitogenic stimulus and trigger the development of AD

Alternatively, both toxicant-induced brain oxidative stress and inflammation could accelerate the consequences of cell cycle reentry by neurons The “two hit hypothesis” of AD pathogenesis postulates that both cell cycle reentry and oxidative stress are necessary for the development of Alzheimer’s disease (Zhu et al 2004) This suggestion seems reasonable given that neurons trapped

in G2 are probably more vulnerable insults Thus external factors that cause oxidative stress or inflammation could significantly accelerate AD pathogenesis

19.3 ALZHEIMER’S DISEASE PATHOGENESIS AND COX2

Vane (1971) showed that the anti-inflammatory action of nonsteroidal anti-inflammatory drugs (NSAIDs) depends on their ability to inhibit cyclooxygenases, which in turn results in a diminished synthesis of prostaglandins In the early 1990s, cyclooxygenases were shown to exist as at least two distinct isoforms: COX1 and COX2 COX2 has emerged as the isoform that is primarily responsible for the synthesis of the prostanoids involved in acute and chronic inflammatory states (Hinz and Brune 2002)

Under basal conditions, COX2 has limited constitutive neuronal distribution in the CNS and contributes to synaptic activity and memory consolidation (Breder, Dewitt, and Kraig 1995; Breder and Saper 1996; Minghetti 2004) COX2 expression is induced by various proinflammatory stimuli, such as cytokines, growth factors, and tumor promoters (Hinz and Brune 2002; Minghetti 2004) Age is the most important risk factor for development of AD The number of people with the disease doubles every 5 years beyond age 65 Oxidative stress is one of the crucial factors parti-cipating in the aging process Cyclooxygenase-derived reactive oxygen species (ROS) generation increases with age, and cyclooxygenase-mediated prostanoid synthesis is one of the major sources

of ROS in the aging process (Kim et al 2000) Transgenic mice overexpressing human COX2 in hippocampal neurons develop neuronal apoptosis and cognitive deficits in an age-dependent manner, and are more susceptible to b-amyloid toxicity with potentiation of redox impairment (Ho et al 1999; Bazan 2001; Bazan and Lukiw 2002) COX2 expression in hippocampal neurons may be a predictor of early AD (Ho et al 2001) and chronic increased COX2 production in brain may have a number of consequences, including free radical mediated cellular damage, vascular dysfunction, alterations in cellular metabolism and neuronal cell cycle, and increases in total Ab content (Naslund et al 2000; Strauss et al 2000; Ho et al 2001; Xiang et al 2002a; Scali et al 2003) COX2 influences processing of APP and promotes amyloid plaque deposition in a mouse model of AD (Xiang et al 2002b) That COX2 plays a role in AD pathogenesis is also supported by epidemiological studies showing an association between long-term use of NSAIDs and a reduced risk of developing AD (Aisen 2002), by the protective effects of COX2 inhibitors in models of AD (Giovannini et al 2003), and by gene expression profiling showing significantly upregulated stress induced proteins, including COX2, in human brain (Lukiw 2004)

19.4 THE MEXICO CITY ENVIRONMENT

Mexico City represents an extreme of urban growth and environmental pollution (Chow et al

discontinuous mountain ranges that limit the natural ventilation of the basin The basin has more than 30,000 industrial facilities and 3.5 million vehicles, with an estimated annual emission of 2.6

at three monitoring stations The higher concentrations of ozone are registered in the SW,

SW and downtown Mexico City

Pollutant levels in Mexico City vary within a relatively narrow range throughout the year, so its residents are exposed all year long to significant burden air pollutants The pollution levels have been sustained or worsened in the last 20 years (Bravo and Torres 2002), so exposures of current children and teenagers are truly lifelong, having begun in utero Moreover, there is a relatively low mobility of Mexico City residents, so individuals tend to be exposed to the same environment for a prolonged period Thus Mexico City presents an opportunity to study chronic health effects associ-ated with prolonged year-round exposures to severe air pollution

Particle Toxicology 334

Xalostoc 0.150

8-h standard

1 2 3 4 5 6 7

Local standard time

0.120 0.090

0.060 0.030 0.000

Pedregal

0.150

8-h standard

1 2 3 4 5 6 7

Local standard time

0.120 0.090 0.060 0.030 0.000

0.150

8-h standard

1 2 3 4 5 6 7

Local standard time

0 10 20 km 23

0.120

0.090

0.060

0.030

0.000

FIGURE 19.1 Spatial and temporal profile of O3air pollution in metropolitan Mexico City The P50(D), and the monthly diurnal average of 8 h O3concentrations (,) observed during January 2005 are illustrated at three representative monitoring stations: Xalostoc, located in a northeast industrial area; Merced, located downtown; and Pedregal, a residential area located in the southwest The higher concentrations of ozone are registered in Pedregal, downwind of the areas where the ozone precursors are produced

0

50

3 )

100

150

200

250

300

350

Merced

San agustin

2 3 4 5 6 7

Local standard time

8 9 10 11 12 13 14 15 16 17 1819 20 21 22 24

0 10 20 km 23

1 2 3 4 5 6 7

Local standard time

1 2 3 4 5 6 7

Local standard time

0 50

3 )

100 150 200 250 300 350 400

0 50

3 )

100 150 200 250 300 350 400

FIGURE 19.2 Spatial and temporal profile of PM2.5air pollution in metropolitan Mexico City Typically, the highest PM2.5concentrations are observed in San Agustin in the northeast and Merced downtown, with the lowest levels in Pedregal in the southwest PM2.5annual concentrations are above the annual standard for all three stations

The potentially toxic components of PM air pollution include acids, polyaromatic hydrocar-bons, metals, biological products such as lipopolysaccharide (LPS), and inorganic carbon particles

particles have a common mechanism of toxicity—the depletion of cellular anti-oxidant defenses (cellular oxidative stress) through the generation of ROS (i.e., hydrogen peroxide, superoxide, hydroxyl radical, and singlet oxygen) However, the mechanism by which ROS are elicited varies significantly with the pollutant The pollutant-induced oxidative stress probably activates stress-response signal transduction pathways by the activation or inactivation of oxidant-sensitive com-ponents of these pathways (e.g., transcription factors and phosphatases) Some pollutants, such as the transition metal vanadium (a potent phosphatase inhibitor itself), may activate these pathways

by ROS-independent mechanisms as well The stress-response pathways regulate the expression of genes encoding antioxidant synthesis and reducing enzymes, detoxification enzymes, and they also regulate immune, inflammatory, cell survival, and apoptotic responses LPS activates many of the same pathways by acting through a specific receptor complex composed of the LPS binding protein CD14 and a heterodimeric plasma membrane protein complex composed of Toll-like receptor-4 (TLR4) and MD2 Chronic activation of the stress response pathways culminates in

a chronic inflammatory response in targeted tissues

We have focused on toxicity due to metals and LPS The most abundant metals in Mexico City

PM are: Ca, Fe, K, Zn, and Pb, while metals typically present in motor vehicle exhaust and fuel oil combustion products, Cr, Ni, V and S, are present in lower concentrations (Chow et al 2002) LPS content in PM samples show a range of 15.3 to 20.6 ng/mg of PM and SE samples show the highest endotoxin concentrations 59 EU/mg PM (Alfaro-Moreno et al 2002; Osornio-Vargas et al 2003)

19.5 COX2 AND IL-1b EXPRESSION, Ab42 ACCUMULATION, AND

NEUROPATHOLOGY IN THE BRAINS OF DOGS AND HUMANS

EXPOSED TO SEVERE URBAN AIR POLLUTION

Healthy Mexico City dogs experience chronic upper and lower respiratory tract inflammation and breakdown of both the respiratory and olfactory epithelial barriers The brains of these dogs exhibited endothelial and astrocytic upregulation of cycloxygenase-2 (COX2) expression in olfac-tory bulb, frontal cortex, and hippocampus, three critical targets in Alzheimer’s disease (Caldero´n-Garciduen˜as et al 2003a) There was also activation of neuronal NF-kB, increased induced nitric oxide synthase (iNOS) expression in cortical endothelial cells as early as 4 weeks

of age, and breakdown of the blood brain barrier (BBB) (Caldero´n-Garciduen˜as et al 2002) iNOS derived nitric oxide (NO) contributes to the generation of peroxynitrite and BBB breakdown (Winkler 2001), leading to vasogenic edema and secondary brain damage (Thiel and Audus 2001; Chao et al 1992) Thus the increase in iNOS expression may be related to the breakdown

of the BBB Furthermore, NO may also contribute to the maintenance, self-perpetuation, and progression of neurodegenerative processes (Grammas et al 1997)

Olfactory bulb and hippocampal apurinic/apyrimidinic sites in genomic DNA were signi-ficantly higher in exposed Mexico City dogs versus controls (Caldero´n-Garciduen˜as et al 2003a), suggestive of increased oxidative stress There was significant DNA damage in the olfac-tory bulb of young Mexico City dogs several months before cortical Ab42 diffuse plaques were detected (Caldero´n-Garciduen˜as et al 2003a) In Mexico City dogs olfactory mucosa pathology appeared by 4 months of age and thus appears to be a precursor of the olfactory bulb pathology (Caldero´n-Garciduen˜as et al 2003a)

Even young dogs (!1 year) showed accumulation of Ab42 in neurons, glial cells and blood vessels, and the presence of Ab42-positive diffuse plaques Dogs from Mexico City exhibited white matter perivascular gliosis as early as 3 mo of age and astrogliosis increased significantly with age (Figure 19.3)

These observations were consistent with an acceleration of an Alzheimer’s-like pathology in apparently healthy dogs

The foregoing studies using dogs as sentinels suggested that chronic exposure to severe air pollution might have adverse effects on human brain To examine this possibility, we conducted a study using autopsy brain samples from Mexican subjects, all lifelong residents of two large cities with severe air pollution, Mexico City and Monterrey, and five small cities with low levels of air pollution Evidence of chronic respiratory tract inflammation was present in all residents of cities with severe air pollution

COX2 mRNA abundance was measured by real-time RT-PCR analysis of total RNA isolated from brain tissues There was a significant elevation of COX2 mRNA levels in frontal cortex and

immunoreactivity in frontal cortex confirmed by quantitative image analysis of COX2 immuno-reactivity (IR) (Figure 19.4b) In subjects from the low exposure group, COX2 IR was confined

to neuronal cell bodies, whereas subjects from the high exposure group exhibited COX2 staining in neuronal cell bodies and dendrites, as well as strong COX2 staining of the endothelium in the frontal cortex (Figure 19.4c) COX2 IR was largely confined to neurons in the hippocampus (Figure 19.4f)

There was a strong positive association between COX2 mRNA levels and oxidative DNA damage as measured by apurinic/apyrimidinic (AP) sites (rZ0.89, pZ0.001) in frontal cortex (Caldero´n-Garciduen˜as et al 2004) The positive correlation between COX2 mRNA and AP sites could be a consequence of COX2-mediated prostanoid synthesis, a major source of ROS that are capable of damaging DNA The DNA damage in frontal cortex suggests that oxidative stress could be a relevant and early event Oxidative damage is an early event in AD, and it is greatest early in the disease and decreases with disease progression (Nunomura et al 2001; Perry et al 2002)

In normal brain expression, the 695 amino acid form of APP (APP695) is much greater than the expression of the 751 amino acid form (APP751) AD brain is characterized by a reversal in the relative expression of APP isoforms as APP751 expression increases dramatically We measured the APP751/APP695 mRNA ratio by real-time RT-PCR in frontal cortex and hippocampus There

FIGURE 19.3 (Seecolor insert) Reactive gliosis and astrocytic proliferation in the frontal cortex white matter

of healthy Mexico City dogs Glial cells and proliferating astrocytes were localized in paraffin sections of frontal cortex of Mexico City dogs by immunohistochemistry using fluorescein-labeled anti-glial fibrillary acidic protein (GFAP, green) and phycoerythrin-labeled anti-bromodeoxyuridine (BrdU, red), respectively, and examined by confocal microscopy: (a) 3-year-old male, (b) 5-year-old female, and (c) 14-year-old female Gliosis worsens with age Images represent maximum intensity projections, showing the maximum intensity of all layers along the viewing direction The inserts represent 3D reconstructions from the same data sets (Pictures were taken by Dr Barbara Rothen-Rutishauser Ph.D Institute of Anatomy, University of Bern, Bern, Switzerland.)

Particle Toxicology 338

was no statistically significant difference between the APP751/APP695 ratios in the high and low exposure groups However, there was a significant positive correlation between COX2 mRNA in frontal cortex and the APP751/APP695 ratio in the frontal cortex of the high exposure group only (Caldero´n-Garciduen˜as et al 2004)

Ab42 accumulation in frontal cortex and hippocampus was measured by quantitative immu-nohistochemistry Ab42 was detected in the perikaryon of pyramidal frontal cortex neurons and in

(Figure 19.5b) in subjects from the high exposure group Ab42 accumulation in the frontal cortex (Figure 19.5c) and hippocampus (Figure 19.5d) of the high exposure group was significantly elevated compared to the low exposure group Three subjects in the high exposure group had rare diffuse Ab42 plaque-like staining in the frontal cortex of (32, 38, and 43 years old) The diffuse Ab42 plaques were associated with reactive astrocytes (e.g., Figure 19.5e) or apoptotic nuclei (not shown) None of the three subjects carried the apolipoprotein E 34 allele (Caldero´n-Garciduen˜as et al 2004), a risk factor for the development of Alzheimer’s disease

In a follow-up autopsy study, the olfactory bulb was examined for evidence of COX2 expression and Ab42 accumulation In accordance with findings in dogs, there was a significant up-regulation of COX2 and IL-1b mRNA expression and Ab42 accumulation in the olfactory bulb

bulb was also seen in arterial smooth muscle cells starting as early as the second decade of life (Figure 19.7)

(f)

(c) 0

1 2 3 4

Low High

2 molec.

Low High 0.0

0.1 0.2

0 1 2 3

Low High

2 molec.

Low High 0

1 2 3 4 5

FIGURE 19.4 (Seecolor insert) COX2 expression in frontal cortex (a–c) and hippocampus (d–f) COX2 mRNA abundance was measured by RT-PCR and normalized for 18s rRNA levels COX2 protein was localized in sections of paraffin-embedded tissues by IHC and its abundance was measured by quantitative image analysis COX2 mRNA was significantly elevated in the high exposure group in both frontal cortex (a, pZ0.009), and hippocampus (d, pZ0.04) COX2 immunoreactivity (IR) was significantly elevated in the high exposure group in frontal cortex (b, pZ0.01), but not in hippocampus (e) Means G SEMs are shown in

A, B, D, and F (c) Representative COX IHC in frontal cortex from a subject in the high exposure group showing strong staining of endothelial cells in the capillaries (*), and pyramidal neurons (arrow), while other neurons were negative (arrowheads) ScaleZ20 mm (f) Representative COX IHC in dentate gyrus from a subject in the high exposure group showing COX2 positive neurons (arrowheads) and capillaries (short arrow) ScaleZ15 mm (Caldero´n-Garciduen˜as, L et al., Brain inflammation and Alzheimer’s-like pathology in individuals exposed to severe air pollution, Toxicol Pathol., 32, 650–658, 2004 With permission.)

The major neuropathological findings in exposed humans included: (1) breakdown of the BBB,

as indicated by the presence extravascular red blood cells, hemosiderin-laden macrophages,

reactive gliosis in the supratentorial white matter (GFAP-positive cells) (Figure 19.8b), and (3)

0

Low Frontal cortex (c) (d)Hippocampus

High Low High 5

10

15 10

* * 7 5 2 0

(a) (b)

p

p

*

*

*

(e)

FIGURE 19.5 (Seecolor insert) Ab42 accumulation in frontal cortex and hippocampus Ab42 was localized

in sections of paraffin-embedded tissues by IHC (a) Anti-Ab42 stained pyramidal neurons (p), astrocytes (arrows) and astrocytic processes (arrowheads) around blood vessels (*) (b) In addition to accumulation in pyramidal neurons, (p) Ab42 was deposited in smooth muscle cells (arrows) in cortical arterioles (*) A dead neuron surrounded by glial cells is indicated (arrowhead) (c and d) Quantitative image analysis of Ab42 IHC showed a significant increase in Ab42 immunoreactivity (Ab42 IR) in both frontal cortex (c, * pZ0.04) and hippocampus (d, * pZ0.001) in the high exposure group MeansGSEMs are shown (e) Ab42 IHC of frontal cortex from a 38-year-old subject from Mexico City showing diffuse plaque-like staining with surrounding reactive astrocytes (arrows) ScaleZ20 mm (Caldero´n-Garciduen˜as, L et al., Brain inflammation and Alzheimer’s-like pathology in individuals exposed to severe air pollution, Toxicol Pathol., 32, 650–658,

2004 With permission.)

Low High 0.01

0.1

1

10

6 molec

Low High 0.1

1

10

5 Eq

Low High 1

2 4 8 16 32

FIGURE 19.6 COX2, IL-1b and Ab42 expression in olfactory bulb of low versus high exposed subjects COX2 and IL1b mRNA abundances were measured by RT-PCR and normalized for 18 s rRNA Ab42 protein was localized in paraffin-embedded tissues by IHC and its abundance was measured by quantitative image analysis COX2 and IL-1b mRNA were significantly elevated in the high exposure group (pZ0.002 and 0.024, respectively) Highly exposed subjects showed a significant increase in Ab42 immunoreactivity as well

Particle Toxicology 340