ToxSci Advance Access originally published online on February 14, 2007
Toxicological Sciences 2007 97(1):163-180; doi:10.1093/toxsci/kfm018
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Assessing Toxicity of Fine and Nanoparticles: Comparing In Vitro Measurements to In Vivo Pulmonary Toxicity Profiles
DuPont Haskell Laboratory for Health and Environmental Sciences, Newark, Delaware 19714
1 To whom correspondence should be addressed. Fax: 302-366-5207. E-mail: david.b.warheit{at}usa.dupont.com.
Received November 14, 2006; accepted February 2, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous studies have reported little correlation between the relative toxicity of particle types when comparing lung toxicity rankings following in vivo instillation versus in vitro cell culture exposures. This study was designed to assess the capacity of in vitro screening studies to predict in vivo pulmonary toxicity of several fine or nanoscale particle types in rats. In the in vivo component of the study, rats were exposed by intratracheal instillation to 1 or 5 mg/kg of the following particle types: (1) carbonyl iron (CI), (2) crystalline silica (CS) (Min-U-Sil 5,
-quartz), (3) precipitated amorphous silica (AS), (4) nano-sized zinc oxide (NZO), or (5) fine-sized zinc oxide (FZO). Depending on particle type and solution state, these particles range in size from 90 to 500 nm in size. Following exposures, the lungs of exposed rats were lavaged and inflammation (neutrophil recruitment) and cytotoxicity end points (bronchoalveolar lavage [BAL] fluid lactate dehydrogenase [LDH] values) were measured at 24 h, 1 week, 1 and 3 months postexposure. For the in vitro component of the study, three different culture conditions were utilized. Cultures of (1) rat L2 lung epithelial cells, (2) primary alveolar macrophages (AMs) (collected via BAL from unexposed rats), as well as (3) AM—L2 lung epithelial cell cocultures were incubated with the particle types listed above, and the culture fluids were evaluated for cytotoxicity end points (LDH, 1-(4,5-dimethylthiazol-2-yl)-3,5-diphenylformazan [MTT]) as well as inflammatory cytokines (macrophage inflammatory 2 protein [MIP-2], tumor necrosis factor alpha [TNF-], and interleukin-6 [IL-6]) at one (i.e., cytokines) or several (cytotoxicity) time periods. Results of in vivo pulmonary toxicity studies demonstrated that instilled CI particles produced little toxicity. CS particles produced sustained inflammation and cytotoxicity. AS particles produced reversible and transient inflammatory responses. NZO or FZO particles produced potent but reversible inflammation which was resolved by 1 month postinstillation exposure. Results of in vitro pulmonary cytotoxicity studies demonstrated a variety of responses to the different particle types, primarily at high doses. With respect to the LDH results, L2 cells were the most sensitive and exposures to nano- or fine-sized ZnO for 4 or 24 h were more cytotoxic than exposures to CS or AS particles. Macrophages essentially were resistant and epithelial macrophage cocultures generally reflected the epithelial results at 4 and 24 h incubation, but not at 48 h incubation. MTT results were also interesting but, except for nano- and fine-sized ZnO, did not correlate well with LDH results. Results of in vitro pulmonary inflammation studies demonstrated that L2 cells did not produce MIP-2 cytokines, but CS- or AS-exposed AMs and, to a lesser degree, cocultures secreted these chemotactic factors into the culture media. Measurements of TNF- in the culture media by particle-exposed cells demonstrated little activity. In addition, IL-6 secretion was measured in CS, AS, and nano-sized ZnO-exposed cocultures. When considering the range of toxicity end points to five different particle types, the comparisons of in vivo and in vitro measurements demonstrated little correlation, particularly when considering many of the variables assessed in this study—such as cell types to be utilized, culture conditions and time course of exposure, as well as measured end points. It seems clear that in vitro cellular systems will need to be further developed, standardized, and validated (relative to in vivo effects) in order to provide useful screening data on the relative toxicity of inhaled particle types. Key Words: particle toxicity; nanoparticles; rat; in vivo; in vitro; pulmonary toxicity; cytotoxicity; ultrafine particles; fine particles.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The successful development of in vitro assays as predictive screens for assessing particle toxicity is an important goal during early product development prior to more substantive inhalation toxicity testing. If properly validated, the advantages of these early screening tests would result since they are simpler, faster, and less expensive than their in vivo counterparts. Moreover, successful development of in vitro toxicity systems should reduce animal use for hazard screening of potential pulmonary toxicants and to evaluate new particulate materials in small quantities. Although implementation of in vitro lung toxicity screening systems would be highly beneficial relative to current testing protocols, the accuracy and predictability of these tests with particulates have not been compared to in vivo results.
Very few systematic attempts have been made to validate the results from in vitro studies to in vivo toxicity effects on the same materials. Seagrave and coworkers assessed the pulmonary toxicity of diesel engine particulate extracts using in vivo assessments followed by in vitro screening studies on the same particulate test substances. These authors noted that the rank order of hazard potency from the in vitro assays in general did not correspond well to the ranking results from the in vivo comparisons of the same samples. In fact, most of the extracts that were determined to be in the least toxic range of tested particulates following assessment using in vitro assay methodology were, ironically, the most toxic following particle instillation in to the lungs of rats. The authors concluded that in vitro assays may provide cell type–specific mechanistic information, but do not accurately reflect in vivo comparative toxicity results (Seagrave et al., 2002
, 2005).
A recently published review/commentary focusing on the safe handling of nanotechnology has recommended strategic research strategies to support sustainable nanotechnologies by maximizing benefits and minimizing environmental and health risks (Maynard et al., 2006). This international panel of scientists recommended five grand challenges which were selected to stimulate research and to bring focus to a range of complex multidisciplinary issues. One of the five key grand challenges cited in the Maynard et al. (2006) commentary is to develop and validate alternatives to in vivo toxicity testing of engineered nanomaterials within the next 5–15 years. Conceptually, there would appear to be several limitations inherent to in vitro assay/cell culture systems in simulating the complex biological effects of inhaled or instilled particles in the lungs of experimental animals. These include, but are not limited to, (1) particle dose, (2) selection of cell types for simulating the lung microenvironment (concomitant with the issue of single-cell culture systems vs. coculture systems), (3) characterization and exposure interactions of lung cells with particle types in culture versus lung fluids, (4) time course of effects (acute vs. chronic) in vivo versus 1, 4, 24, or 48 h incubation in vitro, and (5) appropriate end points for hazard evaluation.
In this study, we have attempted to assess the impact of many of these variables utilized for in vitro toxicity studies with the stated goal of developing an in vitro culture system that could provide a reasonably accurate, predictive early screen for assessing the pulmonary toxicity of particulate materials. Accordingly, we have utilized an in vivo pulmonary bioassay methodology to assess the toxicity of five different intratracheally instilled particles. Each of these particles has been shown to elicit characteristic pulmonary responses following inhalation or instillation exposures. Subsequently, we have conducted systematic studies wherein these same well-characterized particles have been incubated with lung cells in vitro, using three different culture conditions, including a range of doses and incubation time periods. For the evaluation of relevant biomarkers, we have selected inflammation and cytotoxicity indices as comparative end points for both in vitro and in vivo studies. The sustainability of these biomarkers represents important criteria for assessing the toxicity of inhaled particulates.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design
A systematic method has been devised to assess the efficacy and predictability of in vitro studies and to compare the results obtained with lung inflammation and cytotoxicity data obtained from in vivo pulmonary bioassay studies. Accordingly, we evaluated two pulmonary cell types (immortalized rat lung epithelial cells [L2 cells] and primary rat alveolar macrophages [AMs]) both separately and in an epithelial/macrophage coculture system for implementation of in vitro studies with five different particle types. The results were then used to compare with data generated from an in vivo pulmonary bioassay studies in exposed rats using the same fine-sized and nano-sized particles. The five particulates that were evaluated were (1) carbonyl iron (CI) particles, (2) crystalline silica (CS) (Min-U-Sil 5,
-quartz) particles, (3) precipitated amorphous silica (AS) particles, (4) nano-sized zinc oxide (NZO) particles, and (5) fine-sized zinc oxide (FZO) particles. Iron, -quartz, AS, and zinc oxide have unique pulmonary inflammatory profiles. The end points for the in vivo studies were bronchoalveolar lavage (BAL) biomarkers of neutrophilic inflammation and cytotoxicity (BAL fluid lactate dehyrogenase). For the in vitro toxicity study, cytotoxicity was assessed by measuring for lactate dehydrogenase (LDH) and 1-(4,5-dimethylthiazol-2-yl)-3,5-diphenylformazan (MTT) activity in the culture fluid. Inflammatory mediators included measurement of cytokines in the culture fluids—macrophage inflammatory 2 protein (MIP-2), tumor necrosis factor alpha (TNF-), and interleukin 6 (IL-6). In addition, we have assessed the predictive value of an in vitro hemolytic assay as a possible screening tool for gauging in vivo pulmonary toxicity effects of particles.
Preparation and Characterization of Particle Suspensions
Particle preparation.
CI powder (metallic iron) was purchased from BASF Chemical Company (Wayne, NJ) at > 99.5% purity. CS (Min-U-Sil 5, -quartz) was purchased from U.S. Silica Co. (Berkeley Springs, WV) at > 99% purity. AS (Zeofree 80) was purchased from J. M. Huber Corporation (Edison, NJ) at 98% purity. Nano-ZnO and fine-ZnO were purchased from Sigma Aldrich (St Louis, MO) at > 99.0% purity. The particle systems used here were produced via a "top down" approach (grinding or milling a bulk material into an ultrafine dust), therefore resulting in a wide distribution of particle sizes, > 20%. Stock suspensions (10 mg/ml) of each particle were prepared in ultrapure Milli-Q water, phosphate-buffered saline (PBS) solution (sterile filtered), and F-12K cell culture media. Each suspension was probe sonicated (Ultrasonic Processor, 50 W/60 Hz) for 30 min prior to exposure and characterization. For in vivo studies, each stock solution was subsequently diluted serially to yield concentrations of 1 and 5 mg/kg. For the in vitro studies, each stock solution was diluted serially to yield seven concentrations ranging from 0.001 µg/ml (ppm) to 30 mg/ml (0.052–520 µg/cm2). Note that we have made calculations that indicate that the 52 and 520 µg/cm2 doses are considered to be overload doses (Faux et al., 2003). This conclusion was based on a calculation to convert µg/cm2 dose units to cm2/cm2 units (i.e., µg/cm2 x surface area of the particle). The various surface areas of the particle types used in this study were the following: CI = 2.5 m2/g, CS = 5.1 m2/g, AS = 77.7 m2/g, NZO = 12.1 m2/g, and FZO = 9.6 m2/g.
Accordingly, using calculations, we converted the following doses at 52 and 520 µg/cm2, respectively, to cm2/cm2: CI = 1.3 and 13 cm2/cm2, CS = 2.65 and 26.5 cm2/cm2, AS = 40.4 and 404 cm2/cm2, NZO = 6.29 and 62.9 cm2/cm2, and FZO = 5.0 and 49.9 cm2/cm2.
Given that 1–3 cm2/cm2 is considered to be an overload dose for low toxicity dusts such as fine-sized TiO2, all the particle doses equivalent to 52 and 520 µg/cm2 were considered to be overload doses (Faux et al., 2003).
Particle characterization.
In both the dry powder form and the suspension, each particle system was analyzed for various physicochemical properties. Surface area was measured by the Brunauer-Emmett-Teller (BET) method; size, size distribution, and surface charge were examined by dynamic light–scattering (DLS) spectroscopy; and crystallinity was determined by X-ray diffraction (XRD).
BET measurements (Micrometrics ASAP 2405) were taken on each particle sample in its dry state under N2 adsorption to measure specific surface area (Brunauer et al., 1938).
DLS (Malvern Zetasizer Nano-S, Model Zen1600) was used to determine size, size distribution, and surface charge. DLS measurements were taken on each particle sample in each suspension: ultrapure water, PBS solutions, and F-12K cell culture media (Berne and Pecora, 1975).
XRD (Philips X'PERT automated powder diffractometer, Model 3040) was used to determine purity, crystal structure, and crystallite size of the NZO and FZO particle samples (Otwinowski and Minor, 1997).
Pulmonary Toxicity of Particles, In Vivo
Animals.
Male Crl:CD (SD)IGS BR rats (Charles River Laboratories, Inc, Raleigh, NC) were approximately 8 weeks old at study start (240–255 g). All procedures using animals were reviewed and approved by the Institutional Animal Care and Use Committee. The animal program is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care as described in Warheit et al., 1997.
Groups of rats (five rats per group per dose per time point) were intratracheally instilled with particle doses of 1 or 5 mg/kg in PBS. Groups of PBS-instilled rats served as controls. BAL fluid from the lungs of PBS and particle-exposed rats were analyzed at 24 h, 1 week, 1 month, and 3 months postinstillation exposure (pe). Specifically, groups of rats were exposed to (1) vehicle control—PBS, (2) negative control—CI, (3) positive control—CS (Min-U-Sil 5), (4) AS, (5) nano-sized ZnO, and (6) fine-sized ZnO particle types.
Pulmonary lavage.
The lungs of both control and particle-exposed rats were lavaged with PBS (PBS) solution at 37°C as described previously (Warheit et al., 1991). Methodologies for cell counts, cell differentials, and pulmonary cytotoxicity biomarkers in lavaged fluids were conducted as previously described (Warheit et al., 1991, 1997). Briefly, the first 12 ml of lavaged fluids recovered from the lungs were centrifuged at 700 x g, and 2 ml of the supernatant was removed for LDH and total protein.
LDH release.
LDH release was measured using Olympus Lactate Dehydrogenase reagents. Data were collected and recorded on an Olympus AU640 Chemistry Analyzer at 340/380 nm (Roe et al., 1972).
Statistical analyses.
Each experimental value was compared to the corresponding control value for each time point. One-way ANOVA and Bartlett's tests were calculated for each sampling time. When the F-test from ANOVA was significant, the Dunnett's or Dunn's test was used to compare means from the control group and each of the groups exposed to particles. Statistical significance versus PBS control was established as p < 0.05. Statistical tests were performed with SAS 9.1 software (SAS Institute Inc, Cary, NC).
Cytotoxicity of Particles, In Vitro
Cells in culture.
Three different cell culture systems were used in this study: (1) immortalized rat L2 lung epithelial cells (American Type Culture Collection, Manassas, VA), (2) rat lung AMs, and (3) cocultures of (1) and (2). All cell culture systems were cultured in F-12K medium (Kaighn's modification of Ham's F-12 medium) supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin (see Fig. 1) (Warheit et al., 1991). All experiments were conducted at a cellular density of 200,000 cells/cm2.
|
Tests for LDH release, mitochondrial activity (MTT), production of inflammatory mediators (MIP-2), cytokine production (TNF-
and IL-6), and hemolytic potential were performed on all three in vitro cell culture systems and done in triplicate.
LDH release.
Cells were seeded in 24-well plates, exposed to increasing concentrations of particle suspensions (0.0052–520 µg/cm2, where 52 and 520 µg/cm2 are particle-overload concentrations), and incubated for four different time points ranging 1–48 h. After incubation, the plate was centrifuged at 1900 rpm for 4 min. The media were transferred into a fresh 24-well plate and analyzed for LDH release as described earlier in "Materials and Methods."
Mitochondrial activity.
The MTT assay (Sigma Aldrich) was used to evaluate mitochondrial activity (Mossman, 1983). Cells were seeded in 24-well plates and exposed as mentioned before. After incubation, the cell culture media were aspirated; 150 µl of MTT (5 mg/ml) was added to each well and incubated for 4 h. Afterwards, 850 µl of the MTT solubilization solution (10% Triton X-100 in 0.1 N HCl in anhydrous isopropanol) was added to each well. The resulting formazan crystals were solubilized in acidic isopropanol and quantified by measuring absorbance at 570 nm. Data were calibrated to the appropriate calibration curve.
Production of MIP-2.
The Rat Macrophage Inflammatory Protein-2 Enzyme Immunometric Assay Kit (Alpco Diagnostics, Salem, NH) was used to evaluate the production of MIP-2). Cells grown in 24-well plates were exposed to varying particle concentrations (0.0052–520 µg/cm2) (Sherry et al., 1992; Widmer et al., 1993; Wolpe et al., 1989).
Production of TNF-.
The TNF- Enzyme Immunometric Assay Kit (Assay Designs, Ann Arbor, MI) was used to determine the level of TNF- in the biological samples. Cells grown in 24-well plates were exposed to the same particle concentrations (0.0052–520 µg/cm2) (Argilés et al., 1997; Beutler and Cerami, 1998).
Production of IL-6.
IL-6 Enzyme Immunometric Assay Kit (Assay Designs) was used to evaluate the production of IL-6 (Carmeliet, 1990; Hirano et al., 1990; Kishimoto, 1989). Cells grown in 24-well plates were exposed to the same particle concentrations as mentioned above (Carmeliet, 1990).
Hemolytic potential of particle suspension.
The hemolytic potential of the particles used in these studies was measured using the method of Harington et al. (1971). Erythrocytes were obtained from the fresh blood of healthy human donors. The RBCs were washed four times by suspending in PBS solution before centrifugation at 1200 x g. The final suspension consisted of 4% by volume RBC suspended in PBS. Each particle type was suspended in PBS, as previously described, and diluted in saline to give a stock solution of 30 mg/ml. The dilutions ranged from 15 to 0.059 mg/ml. Each suspension (150 µl) was mixed with 75 µl of erythrocyte suspension and incubated for 30 min at 37°C. After incubation, the samples were centrifuged at 717 x g for 10 min at room temperature. Each particle suspension was measured against a blank, PBS negative control, and 1% Triton X-100 positive control. The amount of hemoglobin released into the supernatant was determined spectrophotometrically at 540 nm (Razzaboni and Bolsaitis, 1990).
Statistical analyses.
Statistical analysis was performed as described earlier, i.e., statistical significance versus a control of µg/cm2 dose was established as p < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Particle Characterization
We tested three different in vitro systems and compared the results to in vivo toxicity profiles of test systems exposed to five different particle samples (Fig. 2). The particle types, including CI, CS (Min-U-Sil), AS, NZO, and FZO, were characterized for physicochemical properties in their dry native state, in a water suspension, in PBS solution, and in cell culture media.
|
Different physicochemical properties were observed among each of the particles used in this study (Figs. 3 and 4). Further, certain properties of each particle varied when suspended in the different solvents. In general, as compared to the particle's dry native state, the mean particle size and percent size distribution were increased when placed in a water suspension. The mean particle size and size distribution were increased further in both PBS and F-12K cell culture media suspensions. Aggregation was consistently most pronounced in the F-12K cell culture media. In all particle systems, surface charges were decreased when dispersed in the cell culture media, except for CS (Min-U-Sil) particle types.
|
|
In all cases, the mean particle sizes reported by the manufacturer were significantly different than the actual measured size using BET and XRD in the native dry state and DLS in three different suspensions. From the formula D = 6/(d x A), where D is particle diameter, d is particle density, and A is measured surface area, the particle's average diameter can be calculated, assuming that the particle is spherical. Similarly, the full width at half maximum peak height of the ZnO signature can be used to estimate crystallite size. Both of the calculations are estimates and neither includes aggregation parameters. For example, in the NZO system, the supplier reported a particle size of 50–70 nm. However, three independent sizing techniques indicated that the NZO particles ranged from 90 to 283 nm, which was similar to the size of the FZO particles. The XRD patterns of both NZO and FZO particle samples were identical, and DLS sizing results show the little differences in size of the two zinc oxide particle systems suspended in solvents. Three techniques, independent from the manufacture's reported particle size, indicate that the NZO particles ranged from 90 to 283 nm depending on the particle's state.
In Vivo Effects of Instilled Particle Types as Measured by BAL Fluid Total Cell Numbers, LDH, and Inflammatory Responses
In Vivo BAL Fluid LDH Responses
Exposures to 1 and 5 mg/kg CS particles produced elevated and sustained cytotoxic responses at all measured pe time points (24 h, 1 week, 1 month, and 3 months) (Fig. 5). AS exposures resulted in increased BAL fluid LDH responses at 24 h pe without subsequent significant effects. Exposures to the NZO or FZO particle suspensions produced enhanced cytotoxic responses at the 24-h and 1-week pe time periods but were not different from PBS-instilled controls thereafter. CI exposures did not produce any significant effects compared to controls. Further, longer term BAL fluid analyses as well as histopathological evaluations at later postexposure time points (i.e., 1 and 3 months) indicated that the effects associated with CI, AS, NZO, and FZO exposures were transient.
|
In Vivo BAL Fluid Total Cell Numbers
The numbers of cells recovered by BAL from the lungs of high-dose CS-exposed (5 mg/kg) groups were significantly higher than any of the other groups for all pe time periods (Fig. 6). The number of cells at the 24-h time point for rats exposed to AS, NZO, and FZO were significantly higher than the other time points. This increase at the 24-h time point is a transient effect due to the instillation procedure and was diminished at the 1-week time point.
|
In Vivo Pulmonary Inflammation
The numbers of cells recovered by the lavage from the lungs of rats exposed to CS (Min-U-Sil, 5 mg/kg) were substantially higher than any of the other groups for all pe time periods. Figure 7 demonstrates pulmonary inflammatory responses as indicated by the percentages of neutrophils present in BAL fluid of groups of PBS or particle-exposed rats. Intratracheal instillation exposures of all particle types produced short-term, pulmonary inflammatory responses, as evidenced by increases in the percentages of BAL-recovered neutrophils relative to PBS-instilled controls, measured at 24 h pe.
|
Exposures to 5 mg/kg CI or to 1 or 5 mg/kg AS particles produced transient and reversible neutrophilic lung inflammatory responses at 24 h pe. The responses to AS were much more potent when compared to CI particles. Intratracheal instillation exposures to high-dose NZO (5 mg/kg) or FZO particles produced substantial lung inflammatory responses measured at 24 h pe followed by a minimal, recruitment of neutrophils through 1 week pe (i.e., 15–20% polymorphonuclear leukocytes. These effects were not measured at the 1- and 3-month pe time points, indicating resolution of the inflammatory responses. In contrast, exposures to CS particles (1 and 5 mg/kg) produced persistent pulmonary inflammatory responses, with both doses producing > 25% neutrophils through 3 month postexposure.
In Vitro Cytotoxicity Studies
LDH Studies (1, 4, 24, and 48 h)
LDH values from various doses of five different particle types at four different time points were compared with corresponding culture media (control) values.
L2 (rat lung epithelial cells) LDH values.
After a 1-h incubation period with lung epithelial cells, all the particles produced increased LDH levels in the culture media only at the highest dose: 520 µg/cm2 (Fig. 8).
|
After a 4-h incubation period, CI produced significant increases in LDH levels only at 520 µg/cm2. CS and AS particles produced significant LDH increases at 52 and 520 µg/cm2. In contrast, NZO or FZO particulates produced increases in LDH levels at five concentrations: 0.052, 0.52, 5.2, 52, and 520 µg/cm2.
After a 24-h incubation period, CI produced no significant increases in LDH values at any doses. CS and AS particles produced significant increases in LDH levels at 5.2, 52, and 520 µg/cm2. In contrast, NZO and FZO particles produced increases in LDH levels at 0.52, 5.2, 52, and 520 µg/cm2.
At 48-h incubation period, exposures to CI, NZO, or FZO particles produced no significant increases in LDH values at any doses. CS and AS particles produced significant increases in LDH levels at 5.2, 52, and 520 µg/cm2.
AMs (rat primary AMs) LDH values.
After a 1-h incubation period with rat AMs, none of the particles produced increased LDH levels in the culture media (Fig. 9).
|
After a 4-h incubation period, CS produced a significant increase in LDH levels versus controls at 520 µg/cm2. No significant increases in LDH values were measured in any of the other particles at any time periods.
After a 24-h incubation period, CI and CS produced significant increases in LDH levels versus controls at 520 µg/cm2. No significant increases in LDH values were measured in any of the other particles at any time periods.
After a 48-h incubation period, none of the particles produced increased LDH levels in the culture media at any time periods.
Coculture (epithelial/macrophage coculture) LDH values.
After a 1-h incubation period with rat epithelial/macrophage cocultures, none of the particles produced increased LDH levels in the culture media (Fig. 10).
|
After a 4-h incubation period, CS particles produced increased LDH values only at 520 µg/cm2. In contrast, NZO and FZO particles produced increases in LDH levels at doses of 5.2, 52, and 520 µg/cm2. No significant increases in LDH values were measured in CI or AS particles at any time periods.
After a 24-h incubation period, exposures to CI or AS particles produced increased LDH values only at 520 µg/cm2. CS, NZO, and FZO particles produced significant increases in LDH levels at 5.2, 52, and 520 µg/cm2.
At 48 h incubation, exposures to CI, NZO, or FZO produced no significant increases in LDH values. CS and AS particles produced significant increases in LDH levels at 5.2, 52, and 520 µg/cm2.
To summarize the in vitro LDH findings with L2 lung epithelial cells, primary rat AMs or epithelial/macrophage cocultures, the L2, and cocultures appeared to have the greatest sensitivity to CS particles as well as to NZO and FZO particles. The significant effects associated with these particles appeared to be dose related. It is noteworthy that the macrophage cultures appeared to be resistant to the cytotoxic effects of particles in culture, as evidenced by the lack of LDH secretion by these cells.
MTT Studies (4 and 24 h)
MTT values from various doses of the five different particle types at two different time points were compared with corresponding culture media (control) values. Data were assessed at the 4- and 24-h time points. An independent experiment confirmed that ZnO has no effect directly on MTT reduction.
L2 (rat lung epithelial cells) MTT values.
After a 4-h incubation period, both NZO and FZO particles produced decreases in MTT levels at 5.2 and 52 µg/cm2 (Fig. 11). No significant decreases in MTT values were measured in CI, CS, or AS particles at this time period.
|
After a 24-h incubation period, CI and AS particles produced significant decreases in MTT levels at 5.2 and 52 µg/cm2. CS and NZO particles produced significant MTT decreases at 52 µg/cm2. FZO particulates produced decreases in MTT levels at doses of 5.2 and 52 µg/cm2.
AMs (rat primary AMs) MTT values.
After a 4-h incubation period, AS, NZO, and FZO particles produced decreases in cellular MTT levels at doses of 5.2 and 52 µg/cm2 (Fig. 11). No significant decreases in MTT values were measured in CI, CS, or AS particles at this time period.
After a 24-h incubation period, AS and NZO particulates produced decreases in cellular MTT levels at 5.2 and 52 µg/cm2. FZO particles produced decreased MTT cellular levels at 52 µg/cm2. No significant decreases in MTT values were measured in CI or CS particles at this time period.
Coculture (epithelial/macrophage coculture) MTT values.
After a 4-h incubation period, FZO particles produced decreases in cellular MTT levels at doses of 5.2 and 52 µg/cm2 (Fig. 11). NZO particles produced decreased MTT levels at 52 µg/cm2. No significant decreases in MTT values were measured in CI, CS, or AS particles at this time period.
After a 24-h incubation period, AS, CS, and FZO particulates produced decreases in cellular MTT levels at 5.2 and 52 µg/cm2. No significant decreases in MTT values were measured in CI- or NZO-exposed cocultures at this time period.
In Vitro—Inflammatory Mediator Studies
MIP-2 production by cells exposed in vitro for 24 h to AS, CS, NZO, or FZO particles is shown in Figure 12. Little or no MIP-2 production was measured in the culture fluids of lung epithelial cells exposed to any of the particle types. Rat AMs exposed to AS or CS particulates, but not to NZO or FZO particles generated MIP-2 in the culture fluid. Similarly, epithelial/macrophage cocultures exposed to AS or CS generated MIP-2 in the culture fluid, although the macrophage cultures appeared to be more efficient when compared to the cocultures.
|
TNF-
production by cells exposed in vitro for 24 h to AS, CS, NZO, or FZO particles is shown in Figure 13. It was interesting to note that in vitro exposures of lung epithelial cells, macrophages, or cocultures to CS particles did not produce any secretions of TNF- in the cell culture fluid. Cocultures exposed to AS particle types produced TNF- at doses of 0.52 and 5.2, but the effect did not conform to a dose-response relationship.
|
IL-6 production by cells exposed in vitro for 24 h to AS, CS, NZO, or FZO particles is shown in Figure 14. The results of studies were rather mixed and the relevance of the results was difficult to interpret. In this regard, macrophages and cocultures produced IL-6 following exposures to CS, but only cocultures produced IL-6 in response to AS particulates. NZO and FZO particle exposures produced very little IL-6 and were not dose related.
|
To summarize the results of in vitro cytokine generation studies as a surrogate for in vivo inflammation, it is difficult to draw any meaningful conclusions regarding the relevance of measured biomarkers as predictors of in vivo inflammatory responses. This is due, in large part, to the finding that each test produced different results. The data suggest that MIP-2 generation and corresponding recruitment of neutrophils by macrophages may be operative in the lungs of AS- or CS-exposed rats. However, the TNF-
assay was not useful in discriminating differences between the three test systems studied. In addition, the IL-6 assay showed that the L2 lung epithelial cell line does not produce IL-6 when exposed to these particles.
Hemolysis Assay Results
Results from the hemolytic potential assay demonstrated the differential surface activity of each particle sample when incubated with RBCs (Fig. 15). Incubation of amorphous or CS particle types produced hemolysis of erythrocytes. However, treatment of NZO or FZO or CI particles had little or no effect on red cell hemolysis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study was designed to assess the capacity of in vitro screening studies to predict in vivo pulmonary toxicity of several fine or nanoscale particles in rats. Previous studies have reported little correlation between the relative toxicity of particles when comparing lung toxicity rankings following in vivo instillation versus in vitro cell culture exposures. In the in vivo component of the study, rats were exposed by intratracheal instillation to 1 or 5 mg/kg of the following particles: (1) CI, (2) CS (Min-U-Sil 5,
-quartz), (3) precipitated AS, (4) NZO, or (5) FZO. Following exposures, the lungs of exposed rats were lavaged and inflammation and cytotoxicity end points were measured at 24 h, 1 week, 1 month, and 3 months postexposure.
For the in vitro component of the study, three different cell culture conditions were utilized. Cultures of (1) rat L2 lung epithelial cells, (2) primary AMs (collected via BAL from unexposed rats), as well as (3) AM-L2 lung epithelial cell cocultures were incubated with the particle types listed above. The culture fluids were evaluated for cytotoxicity end points (LDH, MTT) as well as inflammatory cytokines (MIP-2, TNF-, and IL-6). Initially, all three particle-exposed culture types were assessed for cytotoxicity biomarkers at four different time periods—1, 4, 24, or 48 h. Subsequently, MTT end points were measured following incubation for 4 or 24 h, and the cytokine secretion into the culture fluids were measured after treatment for 24 h.
Results of in vivo pulmonary toxicity studies demonstrated that instilled CI particles produced little toxicity. CS exposures produced sustained inflammation and cytotoxicity. Exposures to AS particles produced reversible and transient inflammatory responses. NZO and FZO particle instillations produced potent but reversible inflammation which was resolved at 1 month pe.
Results of in vitro pulmonary cytotoxicity studies demonstrated a variety of responses to the different particle types. With respect to the LDH results, L2 cells were the most sensitive, and exposure to NZO or FZO for 4 or 24 h was more potent than exposures to crystalline or AS particles. Macrophages were essentially resistant and epithelial/macrophage cocultures generally reflected the epithelial results at 4 and 24 h incubation, but not at 48 h incubation. MTT results were also interesting except for NZO and FZO which did not correlate well with LDH results. Results of in vitro pulmonary inflammation studies demonstrated that L2 lung epithelial cells did not produce MIP-2 cytokines. Crystalline- or AS-exposed AMs and cocultures, to a lesser degree, secreted these chemotactic factors into the culture media. Measurements of TNF- in the culture media by particle-exposed cells demonstrated little activity. In addition, IL-6 secretion was measured in CS, AS, and NZO-exposed cocultures.
When considering the range of toxicity to five different particles, the comparisons of in vivo and in vitro measurements demonstrated little correlation, particularly when considering many of the variables assessed in this study, such as cell types, culture conditions, time course of exposure, and measured end points. It seems clear that in vitro cellular systems will need to be further developed, standardized, and validated (relative to in vivo effects) in order to provide useful screening data on the relative toxicity of inhaled particles.
Seagrave and coworkers have attempted to systematically compare the results of in vitro and in vivo pulmonary toxicity studies in ranking the toxicity of gasoline and diesel extract particulate samples. Initially they assessed the mutagenicity and in vivo toxicity of combined particulate and semivolatile organic compound (SVOC) fractions of gasoline and diesel engine emissions. In general, responses to intratracheal instillation in rats demonstrated parallel rankings of the samples by multiple end points reflecting cytotoxic, inflammatory, and lung parenchymal changes. According to the authors, the end points selected to assess oxidative stress and macrophage function yielded little useful information. It was concluded that a subset of this panel of assays was useful in providing rapid, cost-effective feedback on the biological impact of modified technology, in the form of engine emissions. In a subsequent, companion in vitro study, these investigators evaluated the same samples to determine whether the in vitro screening studies could serve as an accurate preliminary screen to rank the toxicities of particulate samples. To facilitate this comparison, Seagrave and coworkers analyzed selected responses of cultured human lung epithelial cells and rat AMs to determine whether the tests could discriminate among combined particulate matter and SVOC fractions of emissions collected from normal- and high-emitter, in-use gasoline and diesel vehicles. The investigators reported that macrophages were more susceptible to cytotoxicity than to epithelial cells. Samples from gasoline vehicles generally caused enhanced toxicological effects compared to the diesel engine samples. However, it is important to note that the rank order of potency from the in vitro assays in general did not correspond with the earlier rankings from in vivo comparisons of the same samples. Accordingly, the authors concluded that, while the in vitro assays may provide useful mechanistic information, revealing cell type–specific responses; however, in their current form, they did not accurately reflect in vivo comparative toxicity end points (Seagrave et al., 2002, 2003, 2005).
In another comparative study, Becher et al. (2001) exposed in vitro rat lung AMs and alveolar epithelial type II cells for 20 h to various stone particles with differing contents of metals and minerals (a type of mylonite, gabbro, feldspar, and quartz). As evaluative criteria, the capacity to induce the release of the inflammatory cytokines IL-6, TNF-, and MIP-2 was investigated. The investigators reported marked differences in potency between the various particles, with mylonite being most potent overall, followed by gabbro, and with feldspar and quartz having an approximately similar order of lower potency. The results also demonstrated differences in cytokine release patterns between the macrophages and alveolar epithelial cells. For all particle types including quartz, alveolar type II epithelial cells, demonstrated the most marked increase in MIP-2 and IL-6 secretion, whereas the largest increase in TNF- release was observed in macrophages. To investigate possible correlations between in vitro and in vivo inflammatory responses, rats were instilled with the same types of particles, and BAL fluid was collected after 20 h. The results demonstrated a correlation between the in vitro cytokine responses and the number of neutrophilic cells in the BAL fluid. The investigators further assessed whether a similar graded inflammatory response would be continued in alveolar epithelial type II cells and AMs isolated from the exposed animals. Again a differential cytokine release pattern was observed between epithelial type II cells and macrophages, although the order of potency between particle types was altered. It was concluded that various stone particles caused differential inflammatory responses after both in vitro and in vivo exposure, with mylonite being the most potent stone particle. The authors claimed that the alveolar type II epithelial cells were important participants in the inflammatory response following exposure to particles. The concordance between in vitro and in vivo effects was measured at only a single time point—20 h. It is noteworthy that in our studies presented herein, the major sources of MIP-2 and IL-6 secretions were the primary AM lavaged from the lungs of rats or the epithelial/macrophage coculture systems, but not the L2 (lung epithelial) cells, which are derived from alveolar epithelial type I cells. In our studies, TNF- was rarely expressed in lung epithelial cells, macrophages, or epithelial/macrophage cocultures.
In this study, we have attempted to address several of the fundamental issues relevant to the development of a validated in vitro assay/cell culture system in order to best simulate the anatomical lung microenvironment concomitant with the complex biological effects (physiology) in the lungs of experimental animals exposed to inhaled particulates. These include, but are not limited to the following parameters: (1) particle dose, (2) selection of cell types for simulating the lung microenvironment (concomitant with the issue of single-cell culture systems vs. coculture systems), (3) characterization and exposure interactions of lung cells with particle types in culture versus lung fluids; (4) time course of effects (acute vs. chronic) in vivo versus 1, 4, 24, or 48 h incubation in vitro, and (5) appropriate end points for hazard evaluations (Table 1).
|
Particle dose is a critical issue because most in vitro as well as in vivo studies employ high exposure doses of putative toxicants in order to determine the potential (worst case scenario) hazard effect related to that particle type. Unfortunately, overload doses of particulates can stimulate an effect that may have little or no physiological relevance. In the in vitro component of our study, we utilized a wide range of doses to assess the dose-response relationship. In many cases, the effects were measured only at doses of 52 or 520 µg/cm2, which should be viewed as a very high, particle-overload dose. The relative resistance of the CS–exposed L2 cells, macrophages, and cocultures to secreting LDH in the culture fluid is a significant cause for concern. Furthermore, when these cell types responded to the silica exposure, the greatest effect was usually measured only at 24 or 48 h with L2 cells and epithelial/macrophage cocultures. It seems likely that particle clearance or translocation mechanisms in the dynamic physiological state of the lung would be operative prior to 24 or 48 h. This suggests that potential hot spots of particle deposition would yield fluid fluxes or particle clearance via macrophages.
Inhaled particles are known to deposit on alveolar epithelial cells at alveolar duct bifurcations (Brody and Roe, 1983; Warheit and Hartsky, 1990). AMs are recruited to sites of particle deposition wherein they phagocytize particles (Warheit et al., 1988). Alternatively, particles may translocate through alveolar epithelial cells from alveolar regions to interstitial sites. Therefore, the selection of epithelial/macrophage cocultures represents a more relevant simulation of the in vivo lung microenvironment. The macrophage component can phagocytize particles in culture but cannot simulate particle clearance from the lung. Perhaps more problematic, the in vitro culture system cannot simulate recruitment of inflammatory cells into the lung, which is an important component of the lung pathogenic process and has been utilized as an important toxicity biomarker for in vivo studies. Cytokine generation by cultured lung cells has been utilized as an indirect biomarker of inflammation. The efficacy of this approach has not been tested or validated.
To summarize our findings, we systematically evaluated a number of variables which strongly impact the ability of in vitro screening studies to accurately reflect in vivo pulmonary toxicity of several particle types in rats. The variables tested included particle dose, time course, duration of treatment exposure, and pulmonary cell types. In vivo effects in the lungs of rats were utilized as the barometer for comparison and validation. Under the conditions of this study, the results of in vivo and in vitro cytotoxicity and inflammatory cell measurements demonstrated little correlation. Our conclusion is that there is a need for better culture systems. In addition, more research is needed to designate the most relevant (chemically, biologically, and physiologically best) end points to test for in vitro toxicity. Accordingly, more sophisticated in vitro cell culture systems need further development and validation (relative to in vivo effects) to facilitate cost-effective screening systems to gauge the relative toxicity of inhaled particles.
|
ACKNOWLEDGMENTS |
|---|
This study was supported by a DuPont Haskell Laboratory Research Proposal Grant. Denise Hoban and Elizabeth Wilkinson conducted the BAL fluid biomarker assessments. Don Hildabrandt provided animal resource care. We thank Dr Lloyd Abrams, Louis Rosen, and Nathaniel Barney for discussion of data and Dr Scott Loveless for helpful comments on this manuscript. The authors also thank Drs Ken Donaldson and Lang Tran for help with the dosimetric calculations.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Argilés JM, Lopez-Soriano J, Busquets S, Lopez-Soriano FJ. Journey from cachexia to obesity by TNF. FASEB J. (1997) 11:743–751.[Abstract]
Baggiolini M, Clemens Sorg. Interleukin-8 (NAP-1) and related chemotactic cytokines. Cytokines (1992) Vol. 4. Basel: Karger. 117–130.
Becher R, Hetland RB, Refsnes M, Dahl J, Dahlman HJ, Schwarze PE. Rat lung inflammatory responses after in vivo and in vitro exposure to various stone particles. Inhal. Toxicol. (2001) 13:789–805.[Web of Science][Medline]
Berne BJ, Pecora R. Dynamic Light Scattering. (1975) New York: John Wiley.
Beutler B, Cerami A. Cachectin (tumor necrosis factor): A macrophage hormone governing cellular metabolism and inflammatory response. Endocr. Rev. (1988) 9:57–66.
Brody AR, Roe MW. Deposition pattern of inorganic particles at the alveolar level in the lungs of rats and mice. Am. Rev. Respir. Dis. (1983) 128:724–729.[Web of Science][Medline]
Brunauer S, Emmett PH, Teller E. Adsorption of gases in multimolecular layers. J. Am. Chem. Soc. (1938) 60:309–319.[CrossRef][Web of Science]
Carmeliet P. Interleukin-6: A putative mediator of anterior pituitary hormone secretion during the stress of infection. Trends Endocrinol. Metab. (1990) May/June:237.
Faux SP, Tran CL, Miller BG, Jones AD, Monteiller C, Donaldson K. In vitro Determinants of Particulate Toxicity: The Dose-Metric for Poorly Soluble Dusts. (2003) Research Report 154. Norwich, Great Britain: Institute of Occupational Medicine for the Health and Safety Executive.
Harington JS, Macnab GM, Miller K, King PC. Enhancement of haemolytic activity of asbestos by heat-labile factors in fresh serum. Med. Lav. (1971) 62:171–176.[Medline]
Hirano T, Akira S, Taga T, Kishimoto T. Biological and clinical aspects of interleukin 6. Immunolol. Today (1990) 11:443–449.[CrossRef]
Kishimoto T. The biology of Interleukin-6. Blood (1989) 74:1.
Maynard AD, Aitken RJ, Butz T, Colvin V, Donaldson K, Oberdörster G, Philbert MA, Ryan J, Seaton A, Stone V, et al. Safe handling of nanotechnology. Nature (2006) 444:267–269.[CrossRef][Medline]
Mossman T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J. Immunol. Methods (1983) 65:55–63.[CrossRef][Web of Science][Medline]
Otwinowski Z, Minor W. Macromolecular Crystallography, Part A (1997) Vol. 276. San Diego, CA: Academic Press Inc. 307–326.[CrossRef]
Razzaboni BL, Bolsaitis P. Evidence of an oxidative mechanism for the hemolytic activity of silica particles. Environ. Health Perspect. (1990) 87:337–341.[Web of Science][Medline]
Roe CR, Limbird LE, Wagner GS, Nerenberg ST. Combined isoenzyme analysis in the diagnosis of myocardial injury: Application of electrophoretic methods for the detection and quantitation of the creatine phosphokinase MB isoenzyme. J. Lab. Clin. Med. (1972) 80:577–590.[Web of Science][Medline]
Seagrave JC, Mauderly JL, Seilkop SK. In vitro relative toxicity screening of combined particulate and semivolatile organic fractions of gasoline and diesel engine emissions. J. Toxicol. Environ. Health Part A (2003) 66:1113–1132.[Web of Science][Medline]
Seagrave JC, McDonald JD, Gigliotti AP, Nikula KJ, Seilkop SK, Gurevich M, Mauderly JL. Mutagenicity and in vivo toxicity of combined particulate and semivolatile organic fractions of gasoline and diesel engine emissions. Toxicol. Sci. (2002) 70:212–226.
Seagrave JC, McDonald JD, Mauderly JL. In vitro versus in vivo exposure to combustion emissions. Exp. Toxicol. Pathol. (2005) 57:233–238.[CrossRef][Web of Science][Medline]
Warheit DB, Carakostas MC, Hartsky MA, Hansen JF. Development of a short-term inhalation bioassay to assess pulmonary toxicity of inhaled particles: Comparisons of pulmonary responses to carbonyl iron and silica. Toxicol. Appl. Pharmacol. (1991) 107:350–368.[CrossRef][Web of Science][Medline]
Warheit DB, Hansen JF, Yuen IS, Kelly DP, Snajdr S, Hartsky MA. Inhalation of high concentrations of low toxicity dusts in rats results in pulmonary and macrophage clearance impairments. Toxicol. Appl. Pharmacol. (1997) 145:10–22.[CrossRef][Web of Science][Medline]
Warheit DB, Hartsky MA. Species comparisons of proximal alveolar deposition patterns of inhaled particulates. Exp. Lung Res. (1990) 16:83–99.[Web of Science][Medline]
Warheit DB, Overby LH, George G, Brody AR. Pulmonary macrophages are attracted to inhaled particles through complement activation. Exp. Lung Res. (1988) 4:51–66.
Widmer U, Manogue KR, Cerami A, Sherry B. Genomic cloning and promoter analysis of macrophage inflammatory protein (MIP)-2, MIP-1 alpha, and MIP-1 beta, members of the chemokine superfamily of proinflammatory cytokines. J. Immunol. (1993) 150:4996–5012.[Abstract]
Wolpe SD, Sherry B, Juers D, Davatelis G, Yurt RW, Cerami A. Identification and Characterization of Macrophage Inflammatory Protein 2. Proc. Natl. Acad. Sci. U.S.A. (1989) 86:612–616.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. M. Waters, L. M. Masiello, R. C. Zangar, B. J. Tarasevich, N. J. Karin, R. D. Quesenberry, S. Bandyopadhyay, J. G. Teeguarden, J. G. Pounds, and B. D. Thrall Macrophage Responses to Silica Nanoparticles are Highly Conserved Across Particle Sizes Toxicol. Sci., February 1, 2009; 107(2): 553 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Card, D. C. Zeldin, J. C. Bonner, and E. R. Nestmann Pulmonary applications and toxicity of engineered nanoparticles Am J Physiol Lung Cell Mol Physiol, September 1, 2008; 295(3): L400 - L411. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Rouse, G. Murphy, M. J. Boudreaux, D. B. Paulsen, and A. L. Penn Soot Nanoparticles Promote Biotransformation, Oxidative Stress, and Inflammation in Murine Lungs Am. J. Respir. Cell Mol. Biol., August 1, 2008; 39(2): 198 - 207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yokohira, T. Kuno, K. Yamakawa, K. Hosokawa, Y. Matsuda, N. Hashimoto, S. Suzuki, K. Saoo, and K. Imaida Lung Toxicity of 16 Fine Particles on Intratracheal Instillation in a Bioassay Model Using F344 Male Rats Toxicol Pathol, June 1, 2008; 36(4): 620 - 631. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Sato, T. Shimosato, W. G. Alvord, and D. M. Klinman Suppressive Oligodeoxynucleotides Inhibit Silica-Induced Pulmonary Inflammation J. Immunol., June 1, 2008; 180(11): 7648 - 7654. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Murphy Jr., R. L. Rouse, W. W. Polk, W. G. Henk, S. A. Barker, M. J. Boudreaux, Z. E. Floyd, and A. L. Penn Combustion-Derived Hydrocarbons Localize to Lipid Droplets in Respiratory Cells Am. J. Respir. Cell Mol. Biol., May 1, 2008; 38(5): 532 - 540. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||