Novel properties of statins: suppression of the systemic and bone marrow responses induced by exposure to ambient particulate matter (PM₁₀) air pollution

Abstract

Exposure to ambient particulate matter (PM₁₀) elicits systemic inflammatory responses that include the stimulation of bone marrow and progression of atherosclerosis. The present study was designed to assess the effect of repeated exposure of PM₁₀ on the turnover and release of polymorphonuclear leukocytes (PMNs) from the bone marrow into the circulation and the effect of lovastatin on the PM₁₀-induced bone marrow stimulation. Rabbits exposed to PM₁₀ three times a week for 3 wk, were given a bolus of 5′-bromo-2′-deoxyuridine to label dividing cells in the marrow to calculate the transit time of PMNs in the mitotic or postmitotic pool. PM₁₀ exposure accelerated the turnover of PMNs by shortening their transit time through the marrow (64.8 ± 1.9 h vs. 34.3 ± 7.4 h, P < 0.001, control vs. PM₁₀). This was predominantly due to a rapid transit of PMNs through the postmitotic pool (47.9 ± 0.7 h vs. 21.3 ± 4.3 h, P < 0.001, control vs. PM₁₀) but not through the mitotic pool. Lovastatin delayed the transit time of postmitotic PMNs (38.2 ± 0.5 h, P < 0.001 vs. PM₁₀) and shifted the postmitotic PMN release peak from 30 h to 48 h. PM₁₀ exposure induced the prolonged retention of newly released PMNs in the lung, which was reduced by lovastatin (P < 0.01). PM₁₀ exposure increased plasma interleukin-6 levels with significant reduction by lovastatin (P < 0.01). We conclude that lovastatin downregulates the PM₁₀-induced overactive bone marrow by attenuating PM₁₀-induced systemic inflammatory responses.

many well-documented studies have supported the concept that ambient particulate matter (particles less than 10 μm or PM₁₀) exposure induces systemic inflammatory responses, characterized by an increase in circulating proinflammatory mediators (10, 22) and blood elements such as leukocytes (8, 16) or platelets (18). These systemic responses to PM₁₀ exposure have been implicated in the risk for downstream cardiovascular events (4). The mechanisms underlying these systemic inflammatory responses remain unclear, but the “spill-over” of lung inflammatory mediators into the systemic circulation has the most experimental support (9, 10, 22, 28). Exposure to PM₁₀ elicits a brisk pulmonary inflammatory response (22) and local production of proinflammatory cytokines such as interleukin (IL)-6 that translocate into the systemic circulation (10), whereby stimulating the bone marrow to promote the release of leukocytes and platelets (6, 8, 15, 16) and activating the vascular endothelial cells to instigate adverse cardiovascular events (4).

We have previously shown that exposure to PM₁₀ stimulates the turnover of polymorphonuclear leukocytes (PMNs) in the bone marrow and accelerates their release into the circulation (6, 15, 16). The release of PMNs from the bone marrow can be stimulated by several mediators including granulocyte macrophage colony-stimulating factor (GM-CSF) (13), IL-6 (21), and IL-8 (24), all of which have been indicated as the mediators induced by PM₁₀ exposure (19, 28). These newly released PMNs have distinct behavioral and inflammatory properties, leading to their enhanced potential to sequester in the lung capillaries (25, 28, 31) and mediate vascular endothelial damage (29, 30). Therefore, the downregulation of an overactive bone marrow could be an important strategy to reduce adverse respiratory and vascular effects of PM₁₀ exposure.

Studies from our laboratory recently showed that the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins) suppress the local production of proinflammatory cytokines in the lung as well as the recruitment of PMNs into the lung tissues following PM₁₀ exposure (14). Statins are known to possess myelosuppressive effects (1); therefore, here we postulate that statins will dampen the bone marrow stimulation induced by PM₁₀ exposure. The present study was designed to determine the effects of statins on the bone marrow of rabbits exposed to PM₁₀ (EHC-93). We used a well-established rabbit model of PM₁₀ exposure for 3 wk (6, 16, 22) and determined the bone marrow response by using thymidine analog, 5′-bromo-2′-deoxyuridine (BrdU) to label the dividing myeloid cells in the bone marrow. BrdU labeling enables morphological differentiation to quantify newly released PMNs from the bone marrow. In this study, we investigated the kinetics of BrdU-labeled PMNs in the circulation to estimate the transit time of myeloid cells through the bone marrow as well as the size of the different pools of PMNs in the marrow. In addition, we determined the retention of BrdU-labeled PMNs (newly released PMNs) in the lung tissues following PM₁₀ exposure.

MATERIALS AND METHODS

Urban air particulate.

Urban air PM₁₀ (EHC-93) was obtained from Environmental Health Directorate, Health Canada (Ottawa, ON, Canada). The detailed characteristics of EHC-93, including particle preparation and chemical composition, have been described elsewhere (32). Briefly, EHC-93 particles have a median diameter of 0.8 ± 0.4 μm (median ± SD), consist of 99% (in number, not mass) of particles < 3.0 μm, and were collected over a major North American city in 1993. Like most ambient particulate matter, it contains a complex mixture of inorganics (e.g., sulfates, nitrates), a carbonaceous core, and 26 different metals with organic components. It has previously been shown to induce lung inflammation and endothelial dysfunction (22). The endotoxin content in EHC-93 is small (6.4 ± 1.8 EU/ml or < 3.0 ng/ml) and has been shown to cause neither a local nor a systemic inflammatory response per se when instilled into the rabbit lungs (33).

Experimental animals.

Female New Zealand White (NZW) rabbits (n = 24; weight, 2.6 ± 0.1 kg; Charles River Laboratories, Montreal, QC, Canada) were used in this study. All animals were 12 wk old at the start of the experimental protocol. They were housed in a clean-air and viral-free room with restricted access, given a standard rabbit diet and water ad libitum. The protocol was approved by the Animal Experimentation Committee of the University of British Columbia.

Experimental design.

The animals were assigned randomly into four experimental groups (each group, n = 6): 1) saline instillation, 2) saline instillation with lovastatin treatment, 3) PM₁₀ instillation, and 4) PM₁₀ instillation with lovastatin treatment. Either PM₁₀ or saline was administrated by a direct intratracheal instillation method, three times a week (the first, second, third are 48 h apart, and the fourth is 72 h after the third in the following week) for 4 wk. PM₁₀ particles (1.0 mg/kg) were suspended in 750 μl of normal (0.9%) saline solution and sonicated for 60 s just before each instillation. The animals were anesthetized with 20 mg/kg ketamine and 5 mg/kg xylazine by a single intramuscular injection assisted by 0.25% isofluorane when necessary. A 3.0-mm ID uncuffed endotracheal tube was placed into the trachea by a blind technique. Proper placement of the endotracheal tube was confirmed by a cough reflex, the respiratory sounds and to-and-fro condensation inside the tube. Once properly intubated, either PM₁₀ suspension (750 μl) or saline (750 μl) was administrated via a MicroSprayer (Dry Powder Insufflator, Model DP-4; Penn-Century, Wyndmoor, PA), cannulated through the endotracheal tube. This method ensures an aerosol and accurate dosing of PM₁₀. The relevance of the PM₁₀ dose (1.0 mg/kg) has been validated previously in the same rabbit model (16). Briefly, we have estimated that 40% of the delivered dose is aspirated into the lung by the direct intratracheal instillation, with <4% of which reaching the alveolar surface. Assuming a 5.8-m² alveolar surface area for a 2.6-kg rabbit, we calculated the alveolar exposure of 3.1 ng/cm² for each dose or 24.8 ng/cm² throughout the study period. This exposure is estimated to be equivalent with humans exposed to 150 μg/m³ of EHC-93 for 20 days, which is equivalent with exposure of humans during the Southeast Asia forest fires in 1997 (23).

Lovastatin (Toronto Research Chemicals, Toronto, ON, Canada) treatment was given orally at a dosage of 5 mg/kg per day for 8 wk, beginning 4 wk before the PM₁₀/saline exposure. Lovastatin dose (5 mg/kg per day) was selected as the lowest effective dose reported in an experimental model using lovastatin for NZW rabbits (20). Lovastatin treatment was given in small volume (2–3 g) of food mixture (cookie dough) that contained the required amount of lovastatin. Complete consumption of the drug-containing food mixture was confirmed on a daily basis. The length of statin treatment (4 wk before PM₁₀ exposure and an additional 4 wk during PM₁₀ exposure) was decided based on our preliminary study, in which statins were given for different periods (1, 2, 3, and 4 wk) and the effects of statins became increasing more evident over the time course (data not shown). Four hours after the final (12th) exposure, rabbits were euthanized with an overdose of pentobarbital sodium.

Blood samples.

To label dividing cells in the bone marrow, 100 mg/kg of BrdU (Sigma-Aldrich, St. Louis, MO) was administered 24 h before the ninth instillation by infusion through the marginal ear vein at a concentration of 10 mg/ml in normal sterile saline over a period of 15 min (7, 16, 26). One ear was thoroughly shaved, and a 20G Teflon catheter (Angiocath; BD, Franklin Lakes, NJ) was inserted into the central ear artery for the following serial blood samplings. The catheter was flushed with 50–100 μl of heparin (1,000 U/ml), closed with a plastic luer-lok cap and held in place with Beiersdorf Leukoplast adhesive tape. Blood samples (1.5 ml at a time) were obtained just before the ninth instillation (baseline) and at 4, 8, 12, 16, 20, 24, 30, 36, 48, 72, 96, 120, 144, and 168 h after the ninth instillation. No sedative agents were required to collect blood from the indwelling 20G catheter, and towel restraint provided warmth and comfort to the rabbit. Blood (0.5 ml) was collected in tubes containing ethylenediaminetetraacetic acid (Vacutainer no. 366643; BD Biosciences) for leukocyte counts and cytokine analyses, and 1 ml was collected in tubes containing acid-citrate dextrose (ACD) (Vacutainer no. 364606; BD Biosciences) for the detection of BrdU-labeled PMNs (PMN^BrdU).

White blood cell counts were determined on Abbott Cell Dyn 3700 (Abbott Laboratories, Abbott Park, IL). Differential counts were obtained by Wright's stained blood smears, and 100 PMNs were evaluated in randomly selected fields of view to determine the changes in the number of band cells. Blood collected in ACD tubes was processed to obtain leukocyte-rich plasma (LRP). Erythrocytes were allowed to sediment for 25–30 min after the addition of an equal volume of 4% dextran (average molecular weight, 162,000; Sigma-Aldrich) in PMN buffer. The resulting LRP was cytospun onto 3-aminopropryl-triethoxysilane-coated slides by cytocentrifugation (150 μl, 500 rpm, 5 min; Cytospin 4; Thermo Shandon, Pittsburgh, PA) at room temperature. The cytospun specimens were air dried, fixed in acetone for 10 min, and stored at −80°C for subsequent immunohistochemical analyses.

Immunohistochemical detection of PMN^BrdU cells in the circulation.

The cytospun specimens were stained for the presence of nuclear BrdU by using the dextran polymer conjugate two-step visualization system (K5355, Envision G 2 System/AP; DakoCytomation, Glostrup, Denmark) according to the manufacturer's instructions. Briefly, the slides were digested at room temperature for 5 min in a 0.08% pepsin solution acidified to pH 2.5. DNA in the samples was denatured in 2N HCl at room temperature for 10 min. The 2N HCl was neutralized by washing the slides three times with 0.1 M borate buffer, pH 8.5, each for 5 min. After nonspecific antigen blocking (Protein Block Serum-Free; Dako, ON, Canada) for 10 min, the specimens were incubated with 2.62 mg/l (1:100 dilution) monoclonal mouse anti-BrdU antibody (Clone Bu20a; Dako) for 30 min in a humidity chamber at room temperature. Nonspecific mouse immunoglobulin G (5 mg/ml) was used as a negative control. The slides were washed with Tris-buffered saline with Tween-20 and then incubated with anti-mouse secondary antibody conjugated with a Dextran backbone of a large number of alkaline phosphatase (AP) molecules for 30 min. AP-labeled amplification polymers were added for 30 min to further enhance the AP complex, and the reactions were visualized by red chromogen. The slides were counterstained with Gill hematoxylin.

Evaluation of PMN^BrdU in the circulation.

PMN^BrdU were evaluated as previously described in detail (26). Briefly, PMNs with any nuclear stain were counted as BrdU-labeled PMNs. PMN^BrdU were divided into three groups according to the intensity of nuclear staining, using an arbitrarily designated grading system, weakly positive (staining of less than 5% of the nucleus: G1), moderate positive (staining of 5 to 80% of the nucleus: G2), and highly positive (staining of more than 80% of the nucleus: G3). This grading system was designed to evaluate the transit time of the myeloid cells that were in their last division in the mitotic pool when exposed to BrdU (G3), those that were in the middle (G2), and those that were in their first division (G1). PMNs exist in the marrow in two divisions: 1) the proliferative or mitotic pool (myeloblasts, promyelocytes, myelocytes) and 2) the maturation storage or postmitotic pool (metamyelocytes, bands, mature neutrophils, polymorphonuclear leukocytes); therefore, G1 and G3 cells represent the mitotic and postmitotic pool, respectively. All the slides were scanned, and their images were captured with the Aperio ScanScope XT (Aperio Technologies, Vista, CA). The coded images were analyzed using Image Pro Plus (Media Cybernetics, Bethesda, MD) without knowledge of the group or sampling time. Fields were selected in a systematic randomized fashion, and 200 cells were evaluated per specimen.

PMN transit time from bone marrow to circulation.

Transit time of PMN^BrdU through the bone marrow was calculated as previously described (26). Briefly, the number of PMN^BrdU was corrected for the disappearance (t_1/2) of cells in the circulation. In previous studies, we have reported that the half-life (t_1/2) of PMN^BrdU in rabbits is 270 min or 4.5 h, using a whole blood transfusion method (3). We have applied this rate of exponential loss of PMN^BrdU from the circulation to calculate the number of PMN^BrdU released from the bone marrow and the transit time through the different pools of the bone marrow in the following manner: ΔN (Δt) = Nt_j - Nt_i exp^−(kΔt), where N is the relative number of labeled cells; t_i and t_j are the initial and successive time; Δt = t_j - t_i; and k = ln2/t_1/2.

These calculations were made for each blood collection interval, and a histogram was drawn showing the distribution of the PMN^BrdU released from the bone marrow during each interval. The mean transit time for all the PMN^BrdU and the different populations of PMN^BrdU (G1, G2, and G3 cells) were calculated individually for each rabbit.

Cytokine analyses.

Plasma IL-6 levels were analyzed by enzyme-linked immunosorbent assay (ELISA) using commercially available kits (CSB-E06903RB; Cusabio Biotech, Beijing, China) according to the manufacturer's instructions. Each sample was measured in duplicate.

Immunohistochemical detection of PMN^BrdU cells in the lung tissue.

After euthanasia, the chest was opened rapidly, the base of the heart was ligated, and the right lung was inflated at 25 cmH₂O by intratracheal instillation of 4% paraformaldehyde. After full inflation, the right main bronchus was ligated and resected from the trachea. The right lung was separated into the four lobes (cranial, middle, caudal, and accessory) and fixed in 4% paraformaldehyde for 24 h. The sliced lung tissues were embedded in paraffin, sectioned 4 μm thick, and mounted on glass slides. After deparaffinization and rehydration, the presence of nuclear BrdU was examined by the same immunostaining method as described above except for the dilution factor of primary antibody, in which 1:300 dilution (0.87 mg/l) was used for the lung tissues.

Evaluation of PMN^BrdU in the lung.

At least five serial cuts from each lobe were stained, scanned with the Aperio ScanScope XT, evaluated for quantitative analysis, and then averaged per animal. To evaluate BrdU-positive staining nuclei of PMNs, the area of positive staining was accurately measured by a color image analyzer (Image Pro-Plus 5.0; Media Cybernetics). Each lung crosssection was captured at ×320 magnification, and color segmentation was employed to determine the area (in μm²) that was stained red (positive staining) in PMNs. The area of analysis was manually traced with Image Pro-Plus 5.0, which allowed the areas of positive staining to be calculated as a percentage of the overall nuclear size of PMN^BrdU compared with the whole area of a given field of view. Ten randomly selected fields from each lobe were analyzed so that 40 fields per animal were analyzed. In all the quantitative analyses, the images were processed by the investigators (R.M. and N.B.) unaware of treatment group assignments. Intra- and interobserver variability were less than 5% for all measurements.

Statistics.

Values are expressed as means ± SE. The repeated-measure variables were analyzed using a mixed-model repeated-measures ANOVA with two between-subject factors (PM₁₀ exposure and lovastatin treatment) and one within-subject factor (time after PM₁₀ exposure) for each dependent variable. To confirm the contribution of PM₁₀ or lovastatin in the repeated-measures variables, a mixed-model repeated-measures ANOVA on a selected set of analysis intervals was performed as appropriate. The transit time of PMN^BrdU, the volume fractions of PMN^BrdU in the lung, and average cytokine levels were compared by two-way factorial ANOVA with PM₁₀ exposure and lovastatin treatment as independent variables, followed by Bonferroni's post hoc test. Transcripts with a significant PM₁₀ exposure × statin treatment interaction (P < 0.05) were additionally analyzed with a one-way ANOVA followed by Newman-Keuls post hoc test for pairwise comparisons. Statistical significance was considered at P < 0.05 (two-tailed).

RESULTS

Leukocytes in the circulation.

The kinetics of leukocytes after PM₁₀/saline intratracheal instillation was characterized by an initial drop, followed by a compensatory increase. For the total circulating white cell counts, the compensatory increase was observed for ∼20 h, followed by a smaller second drop and the second compensatory increase (a biphasic response; Fig. 1A). PM₁₀ exposure contributed to consistent reduction in circulating leukocytes (PM₁₀ main effect, P < 0.01), and lovastatin gradually improved PM₁₀-induced leukopenia at later time points with significant lovastatin main effect after 24 h (P < 0.05). The kinetics of PMNs was also biphasic in nature (Fig. 1B). PM₁₀ exposure suppressed circulating PMNs (PM₁₀ main effect, P < 0.01), and lovastatin improved this neutropenia. The effect of lovastatin was more evident at later time points with a significant lovastatin main effect after 12 h (P < 0.05). In contrast, PM₁₀ exposure strongly promoted the band cell influx into the systemic circulation (PM₁₀ main effect, P < 0.01; Fig. 1C). Band cell counts peaked at 12 h after PM₁₀ exposure, whereas lovastatin delayed the peak to 24 h. The effect of lovastatin was observed at earlier time points (20 h and before, lovastatin main effect, P < 0.01), but not at later time points.

Fig. 1.

Circulating white blood cell (WBC) counts (A), polymorphonuclear leukocyte (PMN) counts (B), and band cell counts (C). Rabbits were exposed to ambient particulate matter (PM₁₀) (gray circles; n = 6), PM₁₀ with lovastatin (closed circles; n = 6), or saline (control; open circles; n = 6) for 3 wk. The kinetics of WBC/PMN was characterized as biphasic waveforms, possibly indicating repeated cycles of marginalization into the lung and compensatory bone marrow response. PM₁₀ exposure caused persistent leukopenia and neutropenia, both of which were improved by lovastatin, especially at later time points (after 24 h for WBC or 12 h for PMN). By contrast, PM₁₀ exposure strongly stimulated the release of band cells with significant reduction by lovastatin at earlier time points (20 h and before). **P < 0.01, PM₁₀ main effect; ##P < 0.01, lovastatin main effect; #P < 0.05, lovastatin main effect.

Effect of statins on plasma IL-6 levels.

The mean IL-6 level (averaged from blood samples collected at intervals of 4 h to 96 h) was increased after PM₁₀ exposure (P < 0.05; Fig. 2A) and significantly downregulated to saline control levels by lovastatin (P < 0.01). The IL-6 levels after PM₁₀ exposure quickly reached their peak at 4 h and stayed high for more than 20 h. By contrast, the IL-6 levels of the saline group reached their peak at 20 h and attenuated rapidly. As was in band cells, the effect of lovastatin was observed at earlier time points (20 h and before, lovastatin main effect, P < 0.01) but not at later time points (Fig. 2B).

Fig. 2.

A: average plasma IL-6 levels within 96 h following direct intratracheal instillation of either PM₁₀ or saline ± lovastatin treatment. PM₁₀ exposure significantly increased plasma IL-6 levels with significant downregulation by lovastatin. B: changes in plasma IL-6 levels after the 9th PM₁₀/saline instillation following exposure to PM₁₀ (gray circles; n = 6), PM₁₀ with lovastatin (closed circles; n = 6), or saline (control; open circles; n = 6) for 3 wk. PM₁₀ exposure showed no main effect, whereas lovastatin showed a significant main effect at early time points (20 h and before, ##P < 0.01).

Effect of statins on the release of PMNs from the bone marrow.

The release of PMN^BrdU after PM₁₀ exposure showed two peaks, the first peak at 24 h and the second at 72 h, whereas the saline group showed only one peak at 72 h (Fig. 3A), indicating an early-phase response of the bone marrow to PM₁₀ exposure (PM₁₀ main effect, P < 0.01). This acute response was attenuated by lovastatin (lovastatin main effect, P < 0.05). The peak at 72 h was blunted by lovastatin, and the decline after the peak was slower, showing a moderate amount of PMN^BrdU even at 144 h. The release of BrdU-labeled G1 cells (the mitotic pool; G1^BrdU) reached their peak at 96 h for the saline group, whereas PM₁₀ exposure accelerated the release to shift the peak to 72 h (PM₁₀ main effect, P <0.01; Fig. 3B). The peak for the PM₁₀ + lovastatin group was blunted, peaking between 72 and 120 h (lovastatin main effect, P < 0.05). The kinetics of BrdU-labeled G3 cells (the postmitotic pool; G3^BrdU; Fig. 3C) showed a significant early increase in the release of G3^BrdU cells induced by PM₁₀ (PM₁₀ main effect, P < 0.001). This early G3 response contributed substantially to the first PMN^BrdU peak shown at 24 h. The effect of lovastatin on the early G3 response was also significant (lovastatin main effect, P < 0.05).

Fig. 3.

The release of 5′-bromo-2′-deoxyuridine (BrdU)-labeled PMNs (PMN^BrdU; A), G1 cells (the mitotic pool, G1^BrdU; B) and G3 cells (the postmitotic pool, G3^BrdU; C) in percentages into the circulation. Rabbits were exposed to PM₁₀ (gray circles; n = 6), PM₁₀ with lovastatin (closed circles; n = 6), or saline (control; open circles; n = 6) for 3 wk. A: PM₁₀ exposure stimulated the release of PMNs showing 2 peaks, the first at 24 h and the second at 72 h. Lovastatin attenuated the early response at 24 h, and the peak at 72 h was blunted (lovastatin main effect, P < 0.05). B: PM₁₀ exposure shifted the peak of G1^BrdU release from 96 h to 72 h. Lovastatin blunted the peak showing a modest increase kept between 72 h and 144 h (lovastatin main effect, P < 0.05). C: PM₁₀ exposure accelerated G3^BrdU release to peak at 24 h, and lovastatin decelerated the release to peak at 48 h (lovastatin main effect, P < 0.05). ***P < 0.001, PM₁₀ main effect; **P < 0.01, PM₁₀ main effect; #P < 0.05, lovastatin main effect.

To evaluate the effect of PM₁₀ and lovastatin on the size of the bone marrow pools, the cumulative number of PMN^BrdU in the circulation was calculated as described previously (27). Figure 4 shows the cumulative frequency distribution of all BrdU-labeled PMNs, G1 cells, and G3 cells. PM₁₀ exposure accelerated the accumulation of PMN^BrdU, and lovastatin attenuated the accumulation (PM₁₀ main effect, P < 0.01; lovastatin main effect, P < 0.05; Fig. 4A). PM₁₀ exposure also accelerated the accumulation of G1^BrdU (the mitotic pool; PM₁₀ main effect, P <0.01; Fig. 4B) that was suppressed by lovastatin with the main effect between 48 h and 120 h (lovastatin main effect, P < 0.01). The effect of PM₁₀ exposure on the accumulation of G3^BrdU (the postmitotic pool) was more rapid and observed before 48 h (PM₁₀ main effect, P < 0.001), and this accumulation was strongly suppressed by lovastatin (lovastatin main effect, P < 0.001 before 48 h; Fig. 4C).

Fig. 4.

Cumulative number of BrdU-labeled PMNs (A), G1 cells (B), and G3 cells (C) in the circulation. Rabbits were exposed to PM₁₀ (gray circles; n = 6), PM₁₀ with lovastatin (closed circles; n = 6), or saline (control; open circles; n = 6) for 3 wk. PM₁₀ exposure accelerated the accumulation of PMN^BrdU, G1^BrdU, and G3^BrdU, and lovastatin attenuated the accumulation of G1^BrdU at later time points, and G3^BrdU at earlier time points. ***P < 0.001, PM₁₀ main effect; **P < 0.01, PM₁₀ main effect; ###P < 0.001, PM₁₀ lovastatin main effect; ##P < 0.01, lovastatin main effect; #P < 0.05, lovastatin main effect.

Transit time of PMNs through the bone marrow.

Table 1 shows the calculated transit time of all PMN^BrdU and the different subpopulations of PMN^BrdU (G3 and G1 cells). PM₁₀ exposure shortened the transit time of all PMN^BrdU and PMNs in the postmitotic pool (G3^BrdU, P < 0.001) but not significant in the mitotic pool (G1^BrdU). Lovastatin delayed the transit time for the postmitotic pool (P < 0.001) but not for all PMN^BrdU or the mitotic pool.

Table 1.

Transit time of PMNs through bone marrow

Group	G3, h	G1, h	All PMNs, h
Saline	47.9 ± 0.7	78.8 ± 3.4	64.8 ± 1.9
Saline + Lovastatin	42.7 ± 3.7	91.7 ± 2.9	70.0 ± 1.5
PM10	21.3 ± 4.3*	62.5 ± 2.8	34.3 ± 7.4*
PM10 + Lovastatin	38.2 ± 0.5†	70.7 ± 8.1	42.0 ± 1.5

All values are expressed as means ± SE. PMN, polymorphonuclear leukocytes.

^*P < 0.001 vs. saline;

^†P < 0.001 vs. ambient particulate matter (PM₁₀).

Sequestration of PMN^BrdU in the lung.

To determine whether these newly released PMNs (PMN^BrdU) are retained in the lung tissues, we harvested lung tissues 7 days after BrdU injection and immunostained with anti-BrdU antibodies to detect PMN^BrdU in the lung (Fig. 5A). Most PMN^BrdU were band cells with highly positive nuclear stain or the G3 pattern (i.e., PMNs from the postmitotic pool). A greater number of PMN^BrdU was observed in the lung tissues exposed to PM₁₀ (P < 0.01; Fig. 5B). Lovastatin abolished this increase (P < 0.01) and promoted the clearance of these immature PMNs from the lung.

Fig. 5.

A: representative microphotographs of BrdU-positive staining nuclei (in red) of PMNs in the alveolar walls 7 days after BrdU injection (100 mg/kg) followed by PM₁₀ instillation. The double arrows show BrdU-labeled segmented PMNs; the single arrows, BrdU-labeled band cells; and the arrowheads, BrdU-labeled mononuclear cells. Most BrdU-positive cells were mononuclear cells in airspaces with a small portion of BrdU-labeled PMNs. Most BrdU-positive PMNs were band cells showing G3-pattern nuclear stain (i.e., postmitotic PMNs). The bars represent 10 μm. B: volume fractions of sequestrated PMN^BrdU in the lung. The results were expressed as a percentage of the total area of BrdU-positive staining nuclei compared with the area of analysis. PM₁₀ exposure promoted the prolonged sequestration of PMNs in the lung that was significantly reduced by lovastatin.

DISCUSSION

Consistent with the observations from our previous studies, the present study showed that exposure to ambient PM₁₀ particles induces systemic inflammatory responses characterized by an increase in systemic proinflammatory mediators such as IL-6 (22). PM₁₀ exposure also stimulates the bone marrow to accelerate the turnover (differentiation) and release of PMNs (16). This study extends these observations by showing that statins (lovastatin) dampen these systemic responses by decreasing the levels of PM₁₀-induced circulating mediators (IL-6), thereby suppressing the bone marrow stimulation. The effect of statins was predominantly on PMNs in the postmitotic pool (G3^BrdU) as evidenced by marked delay in the transit time of G3^BrdU and a decrease in the release of G3^BrdU. Interestingly, statins also reduce the retention of these newly released PMNs in the lung tissues. These novel pleiotropic effects of statins provide some new insight in their anti-inflammatory properties, indicating that this class of drugs could be beneficial to reduce the adverse cardiovascular effects by air pollution exposure. In addition, this study indicates that statin use should be considered a variable that might affect results of epidemiological studies of particle health effects. Lastly, our pilot data showed that a shorter period of statin pretreatment has less impact on the systemic inflammatory response, suggesting that acute statin treatment during high-pollution episodes may not be beneficial, although this awaits direct experimental demonstration.

In this study, the thymidine analog BrdU was used to label dividing PMNs in the marrow that allows us to monitor the intravascular kinetics and tissue distribution of the PMNs newly released from the bone marrow. This approach takes full advantage of animal models because BrdU administration should be avoided for humans due to its acute toxicity, teratogenicity, and carcinogenicity. Using this technique, we have previously shown that repeated PM₁₀ exposure promotes the local (lung) production of proinflammatory mediators such as IL-6 and GM-CSF, which translocate into the systemic circulation to cause downstream bone marrow stimulation that accelerates the turnover of both monocytes and PMNs in the postmitotic pool of the marrow (6, 15, 16). In these previous studies using the same rabbit model, we observed that repeated PM₁₀ exposure shortens the transit time of postmitotic PMNs, but not that of all PMNs (15), which contrasts our findings herein showing the effects of PM₁₀ exposure on transit time for both all and postmitotic PMNs (Table 1). Our PM₁₀ exposure protocol was slightly modified in the present experiments from biweekly (6) to triweekly but keeping the dose of PM₁₀ for each instillation at the same amount. As a result, we assessed the kinetics of PMNs after the ninth instillation, instead of the sixth (as was in the previous studies), which might be implicated in the dose-dependent nature of PM₁₀-induced bone marrow stimulation. Moreover, the instillation method was also improved by a microsprayer that allows even peripheral distribution of PM₁₀ particles. We suspect that our present exposure protocol reflects more real-life exposure, and our data suggest that this protocol is more stimulative for the bone marrow.

The mechanisms underlying the myelosuppressive effects of statins are unclear. It is known that statins possess well-established anti-inflammatory properties including an ability to reduce proinflammatory cytokines such as IL-1, IL-6, IL-8, and GM-CSF from several cell types (14, 19), all of which are known to stimulate the differentiation of neutrophilic cell line. Among others, IL-6 is a multifunctional cytokine that plays a key role in neutropoiesis (21). We showed herein that statins suppress plasma IL-6 levels following PM₁₀ exposure at early time points (∼20 h), and their effect tapered off with time (Fig. 2B). Statins also delay the transit time of PMNs in the postmitotic pool, with little or no effect on the mitotic pool (Table 1), a differential effect that might reflect their downregulation of IL-6 predominantly at early time points. Likewise, PM₁₀ exposure increased the pool size of all PMN^BrdU at early time points (Fig. 4A), which was predominantly due to a significant increase in the release of PMNs from the postmitotic pool (Fig. 4C). Again, statins showed their strong inhibitory effects against this early increase. We have previously shown that IL-6 produced in the lung following PM₁₀ exposure translocates into the circulation within 4 h (10), and we suspect that this early release of the cytokine into the bloodstream is responsible for activating the marrow, thereby promoting the release of PMNs. Collectively, these data suggest that the regulation of IL-6 by statins is likely to contribute to indirect deactivation of the bone marrow, thereby downregulating neutropoiesis as well as neutrophil kinetics. In addition, there is a growing number of reports that have indicated the antitumor effects of statins (2, 5, 17), and it is possible that statins have a similar direct suppressive effect on hematopoietic cells via the same mechanisms as other antitumor agents. However, the molecular mechanism of the antitumor properties of statins is an area of active research and still yet to be identified.

Exposure to PM₁₀ also promotes PMN retention in the lung tissues (Fig. 5B). Studies from our laboratory have shown that PM₁₀ exposure accelerates the release of immature PMNs from the bone marrow (6, 8, 15, 16), and these immature PMNs sequester preferentially in pulmonary capillaries (11, 29) and migrate less efficiently into the tissues (29, 30), both indicating their greater potential to induce lung injury. This study demonstrated that these newly released immature PMNs remain in the lung tissues for a prolonged period of time (Fig. 5B). We have recently showed that statins reduce lung inflammation induced by PM₁₀ exposure (14), and here we showed that statins promote the clearance of these retained PMNs from the lung tissues that might contribute to, at least in part, their anti-inflammatory property against PM₁₀ exposure in the lung.

Possible factors that could impact the bone marrow transit times are the multiple blood samplings taken over the 7-day experimental period. Although the blood volume at each collection is small (1.5 ml = 1% of the total blood volume), the total volume over the 7 days is more substantial (22.5 ml = 15% of the total blood volume). Hematopoietic growth factors that stimulate the bone marrow are known to be not cell type specific (12). Indeed, with this relatively small amount of blood removed over a long period, it is unlikely that the blood loss significantly impacts PMN transit times. In this study, we assume that the effect of blood loss on the bone marrow is similar in all groups. However, we cannot exclude the possibility that statins uniquely impact the bone marrow response to blood loss. Further studies are needed to clarify the effect of statins on the bone marrow response to blood loss.

In summary, chronic exposure to ambient PM₁₀ in rabbits promotes systemic inflammatory responses characterized by an increase in circulating IL-6. It also stimulates the bone marrow, especially the turnover and release of postmitotic PMNs. Treatment with statins attenuates these responses. Exposure to PM₁₀ also promotes prolonged retention of PMNs in the lung tissues, reflecting the potential persistent harm of chronic PM₁₀ exposure. Statins abolish this retention and promote the clearance of retained PMNs from the lung. These inhibitory effects of statins on the bone marrow and lung retention add a new aspect of their anti-inflammatory properties. In this context, statins have attractive qualities for promising applications as therapeutic agents to reduce the adverse health effects of air pollution exposure.

GRANTS

This work was supported by the Heart and Stroke Foundation of Canada, Michael Smith Foundation for Health Research, The Canadian Institute for Health Research, CIHR IMPACT Strategic Training Postdoctoral Fellowship. Stephan F. van Eeden is an American Lung Association Career Investigator and the recipient of the William Thurlbeck Distinguish Research Award and is currently the GSK/CIHR professor in Chronic Obstructive Pulmonary Disease. Ryohei Miyata is a recipient of a CIHR IMPACT Strategic Training Postdoctoral Fellowship.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: R.M. performed experiments; R.M., N.B., and R.V. analyzed data; R.M., N.B., R.V., and S.F.V.E. interpreted results of experiments; R.M. prepared figures; R.M. drafted manuscript; D.D.S. and S.F.V.E. conception and design of research; D.D.S. and S.F.V.E. edited and revised manuscript; S.F.V.E. approved final version of manuscript.