Respirable particles [aerodynamic diameter < 10 μm (PM₁₀)] and fine particles [< 2.5 μm (PM_2.5)] were collected at the Centro Nacional de Investigación y Capacitación Ambiental (National Center for Environmental Research and Training, CENICA). Fourteen (PM₁₀) and 13 (PM_2.5) samples were obtained simultaneously over a 24 hour period, form May, 2005 to February, 2006. The samples were obtained with Andersen-Graseby high volume samplers onto quartz fiber filters (Whatman). The CENICA site is situated in southeast Mexico City (Iztapalapa zone) at the Autonomous Metropolitan University campus. It is the most populated area of the city with some food industries and is less than 2 km from the most important food merchandise distribution center in the city. The samplers were located on the roof of a four-story building.

Before and after sample collection, the filters were conditioned at 22 ± 3°C and 40 ± 5% RH during a 24 hour period and weighed with an analytical balance (Sartorious, sensitivity 10^-4 grams). After weighing, a section of the PM₁₀ filter was subjected to chemical analysis following the standard procedures of USA EPA (1996 and 1998) by inductively coupled plasma atomic emission spectroscopy (Perkin Elmer, 3300 DV), and atomic absorption spectroscopy (Varian, Spectra A-2). A subsample of the PM₁₀ filters were analyzed by electron microscopy (JEOL, JSM-5900 LV) coupled with Energy Dispersive Spectrophotometer (Oxford) with X ray detector in order to know the size distribution and individual composition of the particles. The complete PM_2.5 filter was swept with a powder puff, collected in a polyethylene vial. The amount of particles recovered using this technique ranged from 18 to 80 mg. Once collected, the PM_2.5 were transferred to the Biochemistry and Environmental Medicine Department at the Instituto Nacional de Enfermedades Respiratorias (National Institute for Respiratory Diseases; INER).

The baseline characteristics of all subjects are shown in Table 1. The susceptibility of lipids to oxidation was used to calculate the sample size. According to the mean comparison formula 15 with a standard deviation of 157.53 and a difference of 616, Z_α of 95% and a Z_β of 80%, we obtained a sample size of 2. In total, 6 patients with mild to moderate asthma (AP) who came to the outpatient clinic for asthma management, were medicated with a β₂-agonist, and fulfilled the criteria of the Global Initiative for Asthma 1617 were recruited; 11 healthy volunteers (HV) were also enrolled. All of the subjects had lived in Mexico City for at least 5 years and were asymptomatic at the time of the experiment; none were smokers. On the morning of the experiment, patients and healthy volunteers underwent a spirometry test, which was performed by an experienced technician using a SensorMedics 2200 testing system (Yorba Linda, CA). The highest FVC and FEV₁ values were selected from a minimum of three FVC maneuvers. All subjects gave written informed consent, and the protocol was approved by the ethics committee of the institution (C-03-04).

Blood samples (10 ml) from both healthy volunteers and asthmatic patients were obtained by venepuncture, and neutrophils (N) were isolated with a density gradient using Polymorphprep™ solution (Axis-Shield PoC AS, Oslo, Norway) 18. Four layers were obtained (plasma, monocytes, neutrophils, isolation media and erythrocytes). We recovered the first and third layer in order to quantitate the oxidative damage. The neutrophils were washed twice with Krebs-Ringer phosphate buffer, pH 7.4, supplemented with 1 mg/ml glucose (KRPG). Between the washes, hypotonic shock was used to remove any remaining red blood cells from the white cell preparation. The cell pellet was resuspended in KRPG buffer at a final concentration of 1 × 10⁶ cells/ml.

Before the analysis of paraoxonase (PON) activity, plasma was preincubated with eserine at 0.66 mM for 10 min at room temperature to inhibit butyrylcholinesterase activity and prevent interference with the determination of PON activity, which was measured following the technique of Abbot et al. and expressed as nmol p-nitrophenol/mg APO-A 19.

First, 10 μl of plasma from HV or AP patients were placed in separate polyethylene tubes in 800 μl of 0.05 M acetate buffer, pH 5.4, supplemented with 0.3 M sucrose, 10 μl of 1.4 mM tetramethylbenzidine dissolved in dimethyl sulfoxide and 100 μl of 3.0 mM hydrogen peroxide. After incubation at 37°C for 10 min, 10 μl of catalase (1300 U/ml) and 100 μl of 0.2 M acetic acid were added. The samples were stirred and then centrifuged at 3000 ×g for 5 min and the absorbance at 655 nm was measured 20. The results are expressed as MPO units. One unit (U) was defined as the quantity of enzyme necessary to catalyze an increase of 0.1 in the absorbance at 655 nm and 25°C. The specific activity was expressed as U MPO/mg protein.

Circulating plasma phospholipids, which are rich in unsaturated fatty acids, were examined for their resistance to a specific oxidative aggressor that generates thiobarbituric acid reactive substances (TBARS) 21. In this case, we performed an in vitro evaluation of TBARS formation using Fenton's reaction as a hydroxyl radical (HO^.) generator and evaluated how much TBARS could be formed acutely in the plasma of each subject. The procedure was as follows: 5 μl of plasma from asthmatic patients or healthy volunteers was placed in a glass-covered tube with 7.2 mM Tris buffer (pH 8.2) and the mixture was incubated at 37°C for 15 min in the presence of 5 μM H₂O₂ and 5 μM FeCl₂. At the end of the incubation, 1 mL of thiobarbituric acid 0.375% in 0.2 N HCl was added to the incubation mixture, which was stirred and boiled for 15 min. When the sample reached ambient temperature, 0.5 ml of 0.2 M HCl was added, and the absorbance at 532 nm was measured. The values obtained were expressed as μM of TBARS. The 1,1,3,3-tetramethoxypropane 0.1 mM in sulfuric acid 1% was used as standard.

To measure the amount of free radicals generated, a chemiluminescence (CL) assay was performed as described by Trush 22 using a luminescence counter (20/20 n Luminometer, Turner BioSystems, Sunnyvale, CA). Luminol (5-amino-2,3-dihydro-1,4-phthalazinedione) was initially dissolved in DMSO to a concentration of 25 mM. This solution was stored in the dark at 4°C. On the morning of the experiment, 2 μl of this solution were added to the sample to give a final concentration of 100 μM. The CL response was measured in a polyethylene vial in a reaction volume of 0.5 ml, with 25 μl of the 1 × 10⁶ cells/ml suspension containing neutrophils from healthy volunteers (NHV) or asthmatic patients (NAP). We first recorded the neutrophil CL signal over 10 minutes. After this time, we made a new sample the same way but this time we added 10 μl (1 mg/0.5 ml KRP) of PM_2.5 suspension and recorded the CL response over 10 minutes.

Data are expressed as means ± standard deviation. Paired t-tests were run to compare two groups, and ANOVA with post hoc Bonferroni multiple comparison tests were used for intergroup comparisons. Differences were considered significant when p was < 0.05. Data analyses were performed using the GraphPad Prism software (version 5.0 for Windows; GraphPad Software Inc., La Jolla, CA).

The general and clinical characteristics of the healthy volunteers and asthmatic patients are shown in Tables 1 and 2. All patients were in stable condition at the time of the study. An important point is that some clinical laboratory analyses showed significant differences between asthmatics and healthy volunteers; nevertheless, the measured parameters were not outside the limits established by institutional laboratory standard values.

PM values measured at the CENICA site were 73 and 32 μg/m³ for PM₁₀ and PM_2.5, respectively. The 24 hours average concentration measured in this study were below the Mexican air standars for PM₁₀ (120 μg/m³) and PM_2.5 (65 μg/m³), however the measured concentrations exceeded the Mexican annual standards of 50 μg/m³ for PM₁₀ and 15 μg/m³ for PM_2.5 campaign, showed seasonal variation, PM_2.5 fraction accounted for 49 to 47% of the PM₁₀ fraction during the rain season (May-June) and from 31 to 38% during the dry season (January-February) due to the effects of soil resuspension and land erosion which contributes to an increase on the PM₁₀ fraction (Figure 1). Metals including Cu, Fe and Zn were evaluated in PM₁₀ filter; the average concentrations found were 0.193, 0.838 and 0.127 μg/m³. A mass variability was found respecting those elements probably influenced by whether conditions and seasonal variation, eg. Fe mass as soil indicator, showed a two-fold increase during the dry season and correlated with PM₁₀ concentration (p < 0.05); Zn and Cu were not clearly associated with each other, however on May 14^th, an apparent Cu-Zn episode occurred. Zn showed a light increment during the dry season contrary to Cu concentration, Figure 2. In order to know the composition of PM_2.5, samples of PM₁₀ filters were analyzed by means of Scanning Electron Microscopy, 216 individual selected particles were manually evaluated using energy dispersive X-ray microanalysis (EDX). Individual shape and size particle characterization and semiquantitative percent composition of carbon, oxygen, S, Fe, and Cu were recorded in a database. Conformed information is presented in Table 3. The particles possessed diverse forms including spheres (1, 3 and 8), clusters (2, 4 and 7), plates (5 and 6) and reticular forms (9) corresponding to PM₁₀ particles (indicated by numbers 1–5) and the fine fraction (6–9), (Figure 3). These analyses show that carbon and oxygen were the principal components, derived from incomplete combustion of fossil fuels and mineral contents; S only was observed in cluster (<4.1%) and irregular (<12%) forms in PM₁₀ and in irregular forms in the fine fraction with less of 2% of its content. Moreover, the presence of metallic elements such as iron and copper was detected, the former reached the higher percent in cluster and irregular, both in the fine and PM₁₀ fractions; the latter with exception of cluster shape in the fine fraction was found in all categories and accounted for less than 3% and 1.5% in the coarse and fine fractions, respectively. The presence of Fe and Cu content into spherical and soot aggregates of the fine fraction indicates a combination of natural and anthropogenic sources influenced by smelter and incineration emissions in the study area.

The in vitro generation of ROS was measured by luminol-enhanced chemiluminescence (CL) and expressed as the area under the curve (AUC). The CL AUC from NHV and NAP samples under basal conditions (background) were 3.425 × 10⁶ ± 2.018 × 10⁶ and 2.044 × 10⁶ ± 1.462 × 10⁶, respectively, as a consequence of normal metabolism. The addition of PM_2.5 did not stimulate CL in NHV (3.425 × 10⁶ ± 2.018 × 10⁶ vs. 2.889 × 10⁶ ± 2.894 × 10⁶). In the NAP group, there was nearly a three-fold increase in the CL response; however, this increase failed to reach statistical significance, p = 0.07 (2.044 × 10⁶ ± 1.462 × 10⁶ vs. 5.623 × 10⁶ ± 4.678 × 10⁶) (Figure 4A). When considering individual responses, the NHV group showed a decreased response after addition of PM_2.5 when compared to the basal response (for example, one individual response was 1.148 × 10⁶ vs. 0.157 × 10⁶) before and after particle addition, while the response in the NAP group after PM_2.5 addition was higher (2.63 × 10⁶ vs. 3.74 × 10⁶) (Figure 4B).

Table 2 shows MPO activity expressed as units/mg protein (1 U = ΔA 0.01/min at 655 nm). Enzyme activity increased by 2.18-fold in the AP group when compared to the HV group (p < 0.05). In order to normalize the data, we took the ratio of MPO activity in the plasma to the chemiluminescence response since MPO is found in neutrophils; thus, we could account for the attenuation of the activation of neutrophils in the exposed and control groups (Figure 4).

The plasma paraoxonase activity was expressed as nmol of p-nitrophenol phosphate formed per milligram of apolipoprotein A (Table 2). The paraoxonase activity was reduced by 3.5-fold when compared to the control group (p < 0.001). We normalized paraoxonase activity as described above for MPO activity (Figure 5).

Table 2 also shows the in vitro formation of TBARS as a result of plasma lipoperoxidation by Fenton's reaction. TBARS formation was expressed as μmol per L of plasma (μM) and was 3-fold higher in the AP group than in the HV group (p < 0.001). Because the NAP response increased, we decided to compare it with the oxidative stress parameters in order to determine a general response. In Figure 5, the AUC/MPO ratio shows a pattern similar to that of the chemiluminescence signal. Reduced PON activity indicated inflammation generated by the loss of NAP modulation of ROS (Figure 6). This response is reflected as higher susceptibility to lipoperoxidation in those patients (Figure 7).