Carbon nanotube dosimetry: from workplace exposure assessment to inhalation toxicology
© Erdely et al.; licensee BioMed Central Ltd. 2013
Received: 2 August 2013
Accepted: 20 September 2013
Published: 21 October 2013
Dosimetry for toxicology studies involving carbon nanotubes (CNT) is challenging because of a lack of detailed occupational exposure assessments. Therefore, exposure assessment findings, measuring the mass concentration of elemental carbon from personal breathing zone (PBZ) samples, from 8 U.S.-based multi-walled CNT (MWCNT) manufacturers and users were extrapolated to results of an inhalation study in mice.
Upon analysis, an inhalable elemental carbon mass concentration arithmetic mean of 10.6 μg/m3 (geometric mean 4.21 μg/m3) was found among workers exposed to MWCNT. The concentration equates to a deposited dose of approximately 4.07 μg/d in a human, equivalent to 2 ng/d in the mouse. For MWCNT inhalation, mice were exposed for 19 d with daily depositions of 1970 ng (equivalent to 1000 d of a human exposure; cumulative 76 yr), 197 ng (100 d; 7.6 yr), and 19.7 ng (10 d; 0.76 yr) and harvested at 0, 3, 28, and 84 d post-exposure to assess pulmonary toxicity. The high dose showed cytotoxicity and inflammation that persisted through 84 d after exposure. The middle dose had no polymorphonuclear cell influx with transient cytotoxicity. The low dose was associated with a low grade inflammatory response measured by changes in mRNA expression. Increased inflammatory proteins were present in the lavage fluid at the high and middle dose through 28 d post-exposure. Pathology, including epithelial hyperplasia and peribronchiolar inflammation, was only noted at the high dose.
These findings showed a limited pulmonary inflammatory potential of MWCNT at levels corresponding to the average inhalable elemental carbon concentrations observed in U.S.-based CNT facilities and estimates suggest considerable years of exposure are necessary for significant pathology to occur at that level.
KeywordsWorkplace exposure assessment Inhalation exposure Mouse model MWCNT Dose response and time dependence Protein Gene expression
The pulmonary toxicity of carbon nanotubes (CNT) has been well described. Findings from CNT inhalation exposures included cytotoxicity, inflammatory cell influx, and interstitial fibrosis in the lung [1–6]. Some more recent studies also suggest the potential of CNT to promote lung tumorigenesis . Several studies also have shown systemic effects such as immunosuppression, systemic inflammation, and changes in molecular signaling in extrapulmonary tissues [8–12]. Reduced vascular responsiveness and increased suceptibility to ischemia / reperfusion injury in cardiac tissue were also a product of CNT exposure [13, 14].
Studies with detectable personal breathing zone mass concentrations
Type of samples collected
Personal breathing zone mass concentrations (μg/m3)
Estimated inhalable mass
0.7 - 53
Maynard et al. 2004 
Total gravimetric mass
N.D. - 331.7
Han et al. 2008 
Total gravimetric mass
7.8 - 320.8
Lee et al. 2010 
Total carbon-inhalable size fraction
Methner et al. 2010 
Elemental carbon- inhalable size fraction
N.D. - 38
Methner et al. 2012 
Elemental carbon- inhalable size fraction
N.D. - 7.86
Dahm et al. 2012 
Elemental carbon- respirable size fraction
45 - 80
Birch et al. 2011 
Maynard et al. collected the total mass of a metal catalyst used to produce single-walled (SW) CNT and then estimated the mass of SWCNT exposures to be between 0.7 - 53 μg/m3. Another study conducted by Han et al. collected full-shift samples for the total mass of all particulates in a multi-walled (MW) CNT research facility and found PBZ concentrations ranging from non-detectable concentrations - 331.7 μg/m3. The exposures at the high end were measured during blending of CNT with no control measures. Once control measures were in place, exposure was greatly diminished. Similarly, Lee et al. visited seven facilities ranging from industrial settings to lab scale settings and found full-shift total gravimetric mass concentrations from PBZ measurements to range from 7.8 - 320.8 μg/m3.
The interpretation of gravimetric sampling for high aspect ratio carbon-based nanomaterials, including CNT and carbon nanofibers (CNF), has proven difficult due to the unique characteristics of the material which includes low bulk densities and entangled / bundled structures rather than discrete fibers. Recently, several studies have utilized methodologies to measure the chemical specific mass of elemental carbon, using NMAM 5040, as a marker for CNT or CNF exposure (Table 1) [22–24]. The use of this marker has provided a more refined mass-based workplace exposure estimate, as opposed to total carbon or gravimetric dust measurements. The study by Methner et al. collected PBZ samples at a CNF end user worksite, among many other nanomaterial facilities, and measured the inhalable mass of total carbon (elemental + organic carbon), a less specific marker for CNT exposure . These samples were collected during short duration exposures, aimed to identify worst-case scenarios, and found concentrations between 64 μg/m3 - 1094 μg/m3.
Subsequent studies have reported elemental carbon (EC) measurements, a more specific marker for CNT/CNF exposure. A more recent study by Methner et al. found few detectable exposures during short duration tasks at four SWCNT or CNF downstream user facilities . PBZ samples were collected and analyzed for EC. One facility yielded detectable quantities of EC, which were 33 μg/m3 and 38 μg/m3. Although few of the PBZ samples had measurable EC concentrations, most of the collected samples showed evidence of SWCNT or CNF exposure by electron microscopy . A study conducted by Birch et al. collected full-shift PBZ samples at a large volume CNF production facility and found concentrations at the respirable size fraction of 45 μg/m3 and 80 μg/m3. Dahm et al. found patterns of exposures between primary producers and downstreams users of MWCNT, SWCNT, and CNF materials with EC exposures ranging from non-detectable concentrations to 7.86 μg/m3 for the inhalalable size fraction . Overall, the measurement of EC is a more specific and sensitive marker of exposure which provides a more realistic workplace exposure concentration when compared to gravimetric sampling.
The occupational workforce handling CNT and CNF (beyond research and development scale) currently employs at least 500 workers at 61 companies in the U.S. with an expected growth of about 22% annually [15, 26]. Currently there are no known end-point effects in humans following CNT exposure. This leads to extrapolation from rodent studies. Given the toxicity observed in rodents, epidemiologic studies have been initiated world-wide . One key question becomes obvious: how do realistic U.S. workplace exposures relate to the toxicity found in animal studies? The present study was designed to address this question. The goal was to expose animals to a high dose that would cause significant inflammation with histological findings and then a low dose to serve as a no observable effect level. This design will serve as a reference for detailed molecular analysis, pulmonary pathology, systemic inflammation, and evaluation of cardiovascular dysfunction at human relevant exposures.
The inhalation study utilized the MWCNT produced by Hodogaya, commonly referred to as the Mitsui MWCNT or MWNT-7. This particular product was utilized for several reasons: 1.) A majority of the U.S. workforce handling carbonaceous nanomaterials primarily produces or utilizes MWCNT . 2.) Economically, the global market showed that CNT represents 28% of the total engineered nanomaterial market share with MWCNT being 94% of the total CNT production value (http://www.nanowerk.com/spotlight/spotid=23118.php). 3.) The Hodogaya MWCNT has been extensively characterized and pulmonary effects are known for certain lung burdens [5, 27]. 4.) Ongoing studies are obtaining and evaluating specific CNT materials that are being used primarily in U.S. facilities. Results for the Hodogaya-produced MWCNT will permit comparison with products utilized by U.S. facilities.
Results and discussion
PBZ elemental carbon measurements in MWCNT facilities
The average EC concentrations at the inhalable size fraction from the eight total MWCNT sites was found to have an arithmetic mean of 10.6 μg/m3 with a standard deviation of 17.2 (geometric mean- 4.21 μg/m3 and geometric standard deviation of 4.15). In these 8 MWCNT facilities, exposures ranged from non-detectable samples to 79.6 μg/m3 and the exposure levels were log normally distributed (Shapiro-Wilk p=0.97). Nearly all of the samples that were found to be > 10 μg/m3 came from a single facility; however, the number of PBZ samples collected at this site was representative of the facility’s percentage of the total MWCNT-exposed workforce. Thus, the reported distribution of exposure levels should reflect the target population of workers.
The NIOSH Current Intelligence Bulletin on CNT and CNF has recently published a recommended exposure limit (REL) of 1 μg/m3 for an 8 h time-weighted average of EC . The REL is for the respirable fraction of CNT. The data discussed in the above results is for the inhalable fraction. A recently study on the dustiness of various types of nanomaterials provided the respirable to total or inhalable size fraction ratio . Only one MWCNT and one CNF were reported. Ratios of 0.17 and 0.28 respirable to total were found, respectively. It is anticipated that different types of CNT powders may behave differently and the ratio of respirable to inhalable or total may vary depending on the material handled, the dust generation process, the aerosol concentration, and the sampling location. Continuing studies within nanomaterial facilities are routinely collecting PBZ respirable fractions for MWCNTs. Preliminary results indicate similar findings to those predicted by dustiness studies. Therefore, assuming a 25% respirable fraction, the respirable fraction as an average of all 8 facilites would be approximately 2.65 μg/m3. This is in sharp contrast to 45 and 80 μg/m3 of respirable EC found at a CNF facility . However, this facility produced large quantities of material and was using inadequate exposure control strategies. Given that 75% of MWCNT exposure measurements for inhalable EC were < 10 μg/m3, it would suggest that exposures to CNT can be contained with proper engineering controls [19, 23]. However, it should be noted that the possible health effects of exposure to inhalable (but non-respirable) MWCNT are unknown.
Images of workplace exposures
Extrapolation of lifetime cumulative worker exposures to rodent dosimetry
Therefore, in a mouse this worker exposure is equivalent to 2 ng/d.
Predicted lung burden and relationship to human equivalence
Total lung burden per day (ng)
Alveolar lung burden per day (ng)
Total alveolar burden (μg)
Alveolar lung burden compared to human exposed at 10 μg/m3for 8 h
5 mg/m3 for 5 h/d for 19 d
1000 days (76 years)
0.5 mg/m3 for 5 h/d for 19 d
100 days (7.6 years)
0.5 mg/m3 for 0.5 h/d for 19 d*
10 days (0.76 years)
Caveats of the above extrapolation were the MMAD of 5.5 μm and utilization of mass as the dose metric. As shown in Figure 2, the potential exists for a range of dispersions in the facility which would affect the MMAD and, therefore, the alveolar deposition. The inhalation system developed at NIOSH for in vivo testing dispersed the MWCNT used in this study to a MMAD of 1.5 μm . If a MMAD of 1.5 μm was observed in a MWCNT facility, then the estimated alveolar deposition would increase from 4% to 11% [33, 34]. Extrapolating 11% alveolar deposition, the inhalation design of this study would represent 0.27, 2.7, and 27 years of a human exposed to an inhalable EC concentration of 10.6 μg/m3. Lastly, extrapolation of effects in this study assumes complete exposure and does not take into account respirator protection. The correct and properly fitted respirator would protect according to the assigned protection factor.
The exposure assessment data and extrapolations from this study were done using mass-based analysis as opposed to fiber number (e.g. nanotubes/cm3). Currently, NIOSH is attempting to develop methodologies utilizing controlled inhalation exposures to MWCNT to accurately assess the fiber number in the aerosol  and such fiber count methodologies are being explored from the collected human PBZ from this study. Even when counting methodologies are utilized, it will be difficult to link directly to end-point pathology as degree of dispersion can affect pulmonary fibrosis and recent evidence suggests the potential for larger agglomerates to dissociate into smaller structures or singlets over time . Difficulties will also arise based on the type of CNT material being produced and handling procedures which will alter the agglomerate characteristics and potentially the pathology. Given the current REL for CNT of 1 μg/m3 is a mass-based measurement , extrapolations from this study utilized mass.
In vivo inhalation parameters
Summary of inhalation exposure parameters
Daily exposure concentration and duration
Mean concentration (mg/m3)
Daily mean exposure (min)
C x T (mg/m3x h)
1 – High dose
2 – Middle dose
3 – Low dose
4 – High dose
5 – Middle dose
6 – Low dose
Pulmonary response to MWCNT inhalation exposure
The BAL inflammatory findings detailed in this study qualitatively agree with those of other MWCNT inhalation studies (Figure 4). Porter et al. reported cytotoxicity and increased PMN at the lowest cumulative C x T exposure dose of 100 mg/m3 x h. . Studies using lower cumulative exposures showed no cytotoxicity or PMN influx [4, 10]. In this study, there was a transient increase in cytotoxicity at 0 d in the middle dose. This indicates that 19 d inhalation of 0.5 mg/m3 for 5 h/d is cytotoxic but the deposition level of 197 ng/d did not sustain the response for the particular MWCNT studied. Changes in BAL total protein or albumin appears to have a lower threshold. The middle dose in this study as well as the low dose in the Pauluhn study indicated increased BAL albumin and total protein, respectively . Increased BAL albumin levels without cytotoxcity have been observed following other particulate exposures . As should be expected, not all studies are in perfect agreement with the arbitrary cutoffs in Figure 4. In particular, a recent study by Ma-Hock et al. showed significant cytotoxicity at lower C x T than the transient response of the middle dose from this study without increased total protein in the BAL . Increased inflammatory proteins were measured in the BAL at the middle dose in the panel analyzed; this is in agreement with previous studies [5, 11] as well as the middle dose by Ma-Hock et al. . These effects also tended to be below the threshold for marked cytotoxicity or PMN influx. Lastly, increased transcription of inflammatory markers was evident at the lowest cumulative C x T (Figure 4). Although increased protein levels were not evident, these data show a low level of underlying inflammatory signaling at lung burdens which do not cause any measurable toxicity. The increased mRNA expression may be in response to the process of pulmonary handling of the daily nanogram levels of deposited MWCNT. The genes measured in Figure 8 are either produced by or enhance the accumulation of macrophages. The MWCNT used in this study were primarily found in alveolar macrophages following exposure supporting macrophage-related signaling . Ongoing studies are using global mRNA expression profiling by microarray paired with subsequent pathway analysis to compare and contrast response mechanisms associated with a high dose exposure and a lung burden more relevant to the workplace. Given that transcriptional changes were evident at a 10 times lower deposition than necessary to induce more traditional markers of pulmonary toxicity, the approach seems logical. Many of the human analogs for these gene expression products are being analyzed in a cross-sectional study of workers exposed to CNT or CNF .
Pulmonary pathology of MWCNT
One effect that was not considered or measured in this study, or those evaluated in Figure 4, was the potential for carcinogenicity. Recent work by Sargent et al. showed that inhalation of MWCNT promoted mouse lung tumor formation following initiation by 3-methylcholanthrene . The C x T for that study was 375 mg/m3 x h (5 mg/m3 x 15 d x 5 h/d) putting the cumulative exposure in the top tier of studies in Figure 4. While the data did not indicate initiation by MWCNT, they also did not show a no effect level for promotion. This was similar to in vitro studies that showed chromosomal effects were still measured at the lowest administered dose of 0.024 μg/cm2. Therefore, a cutoff for extrapolated lung burden that would not result in tumor promotion cannot be currently determined, further indicating caution when handling CNT.
Several limitations were inherent to this study. One limitation was mass-based EC concentration cannot distinguish between other carbon species (e.g. diesel exhaust, burning of biomass, or other forms of nanomaterials such as graphene or fullerenes) which may over-estimate exposure to MWCNT. However, all efforts were attempted to minimize this by subtracting anthropogenic background concentrations of elemental carbon for each sample. For the human dosimetry calculation, we had to make several assumptions including using a fixed respirability to inhalability ratio to estimate the MMAD that was not specifically measured in the field. We used all of the the empirical data available to estimate an average deposition although it should be understood that variations in those parameters will affect alveolar deposition. Another limitation is the relatability of the animal exposure to the MWCNT found in workplace facilities. The MWCNT used in the animal inhalation study is a material that disperses very well as compared to many of the images observed from the PBZ samples. The inhalation material was also able to penetrate into the pleural space which is less likely for more agglomerated or thinner more flexible CNT . Therefore, we believe that the toxicity for this particular exposure resides more in the worst case scenario category. To provide further insight, ongoing studies are evaluating toxicological endpoints of materials utilized by U.S. facilities.
The cutoffs in Figure 4 were made in an arbitrary fashion from consensus findings across all studies. Discrepancies include the beginnings of histological aberrations at a lower dose , no PMN influx at any dose , and the transient LDH response soon after cessation of exposure at the middle dose of this study. In general, the recent findings from Ma-Hock et al. have increased pulmonary toxicity at lower cumulative C x T compared to other MWCNT inhalation studies depicted in Figure 4. The differences between studies can be the result of different MWCNT structures produced by different inhalation system generations, differences in the types of MWCNT tested, timing of measurements, and the age, strain, and species used in each study. Irrespective, Figure 4 shows a relatively uniform prediction of effects for traditional measures of pulmonary toxicity when normalized to cumulative C x T with the assumption that a range exists for each specific cutoff depending on the type of MWCNT and inhalation exposure design. Recent studies have shown that specific functionalization of CNT can alter the toxicity [44–46]; it would be interesting to see how those materials would differ under similar inhalation exposure designs as previous studies.
The primary goal of this study was to provide context to relate exposure assessment studies in facilities manufacturing and handling MWCNT to dosimetry for in vivo rodent exposures. Exposure assessment data indicated an average inhalable EC concentration of 10.6 μg/m3 from PBZ measurements. Assuming a 25% inhalable to respirable ratio, the approximate respirable concentration is 2.65 μg/m3. It is clear from toxicological evaluations that MWCNT have a relatively high hazard when compared to other materials. These hazards may include fibrosis, promotion of lung tumors, cardiovascular dysfunction, and pulmonary and systemic inflammation. The present findings show that limiting cumulative exposures is imperative to reducing adverse effects.
A literature search was performed for exposure assessment studies at facilities handling MWCNT, SWCNT, or CNF in which personal breathing zone mass-based samples were collected. A total of 7 studies, over the past 10 yr, met the above criteria and are listed in Table 1. These studies represent a wide range of mass-based analysis methods, industries, and exposure scenarios all producing CNT/CNF or using the material downstream as a secondary manufacturer.
Exposure assessments were conducted at 8 different facilities producing or using MWCNT. Five of these facilities were previously reported  and an additional 3 MWCNT facilities were added to create an inhalable EC average from 8 facilities which were adapted to generate Figure 1. A detailed analysis of the findings from all facilities will be published separately (Dahm et al. , in preparation).
In brief, personal breathing zone, mass-based samples were collected to estimate the inhalable size fraction for EC. These samples were collected on open-faced, 25-mm-diameter quartz fiber filters (SKC Inc., Eighty Four, PA, USA) using Leland Legacy pumps (SKC Inc.) operating at 6–7 l min-1. The samples were subsequently analyzed according to NMAM Method 5040  which is currently recommended by NIOSH to assess exposures to CNTs/CNFs . Background samples for EC were collected in the same manner due to the potential interference from anthropogenic sources of EC. The background samples were collected as indoor or outdoor area samples based on the characteristics of each facility.
Concurrent, side-by-side personal breathing zone samples were also collected for transmission electron microscopy (TEM) to confirm the presence of MWCNT. Samples were collected on 25-mm mixed cellulose ester filters (0.8-lm pore size; SKC Inc.) using Leland Legacy pumps (SKC Inc.) operating at 5 l min-1. The TEM samples were then analyzed on a JEOL2100F TEM (JEOL USA, Inc., Peabody, MA, USA) using a modified NMAM 7402, asbestos by TEM . Additional details on the sampling methodologies can be found in Dahm et al. .
Study design and inhalation exposure in mice
Specific pathogen-free, male C57BL/6J mice from Jackson Laboratory (Bar Harbor, ME) were used in this study. All mice housed in the AAALAC-approved NIOSH Animal Facility were provided food and tap water ad libitum in ventilated cages in a controlled humidity and temperature environment with a 12 hr light/dark cycle. Animal care and use procedures were conducted in accordance with the “PHS Policy on Humane Care and Use of Laboratory Animals” and the “Guide for the Care and Use of Laboratory Animals” (NIH publication 86–23, 1996). These procedures were approved by the National Institute for Occupational Safety and Health Institutional Animal Care and Use Committee.
C57BL/6J mice, eight weeks of age, were exposed by inhalation to MWCNT (produced by Hodogaya, Japan) at various parameters to achieve two orders of magnitude range of lung burdens as shown in Table 2. Animals were exposed using a computer controlled whole body inhalation exposure system designed and constructed at NIOSH . In brief, the inhalation exposure system combines air flow controllers, aerosol particle monitors, data acquisition devices, and custom software with automatic feedback control to achieve constant and repeatable exposure chamber temperature, relative humidity, pressure, aerosol concentration, and particle size distributions. The MWCNT used in this study have been extensively characterized previously [5, 49]. The average diameter was 49 nm with a length of 3.86 μm (GSD 1.94). Purity was >99% carbon. Mice were exposed for 19 d over a total of 26 d (mice were not exposed for the three weekends and the second Monday of each four week exposure). The doses chosen were 5 mg/m3 for 5 h/d, 0.5 mg/m3 for 5 h/d, and 0.5 mg/m3 for 0.5 h/d. The lowest dose method was chosen because consistency was questioned in tests of exposure levels at 0.05 mg/m3. In this group, mice were exposed to air for 4.5 h then MWCNT for 0.5 h to maintain the same timeframe as the two other groups. Mice were euthanized at 0 d (immediately following the last exposure), 3 d, 28 d, and 84 d. The left lung lobe was ligated and frozen in liquid nitrogen and the right lung lobes were lavaged. Two separate groups were exposed for each dose for a total of n=12 air and n=12 MWCNT at each time point unless otherwise indicated. Mice in set one were used for histopathology (n=6 air; n=6 MWCNT) at 84 d post-exposure and mice in set two were used to include 3 d post-exposure (n=6 air; n=6 MWCNT).
Lung burden of MWCNT
Previous studies have determined that male C57BL/6J mice exposed to 5 mg/m3 for 12 d have a total lung burden of 28.1 μg, or 2.34 μg/d . In that same study, alveolar deposition was calculated at 84% of the total lung burden (or 1.97 μg/d). Assuming linearity, these values were extrapolated to achieve the lung burdens shown in Table 2. While these values are estimates because of the assumption of linearity, the values were generated from previously quantitated lung burdens in aged matched male mice of the same strain, C57Bl6/J [5, 30].
Bronchoalveolar lavage (BAL)
BAL collected from right lobes was assessed for lactate dehydrogenase activity, albumin concentration, and inflammatory protein concentrations. Differentials from the cellular fraction were made from counts of 300 cells per slide stained with Wright-Giemsa stain.
Pulmonary gene expression
RNA was isolated from frozen lung using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA). Evaluation of gene expression was determined by standard 96-well technology using the StepOne™ (Applied Biosystems, Carlsbad, CA, USA) with pre-designed Assays-on-Demand™ TaqMan® probes and primers including Il6 (Mn00446190_m1), Ccl2 (Mn00441242_m1), Cxcl2 (Mn00436450_m1), and Ccl22 (Mn00436439_m1) (Applied Biosystems). Using 96 well plates, one μg of total RNA was reverse transcribed using random hexamers (Applied Biosystems) and Superscript III (Invitrogen, Carlsbad, CA). Nine μl of cDNA (1/10) was then used for gene expression determination. Hypoxanthine-guanine phosphoribosyltransferase was used as an internal reference. Relative gene expression was calculated using the comparative threshold method (2-ΔΔCt) with vehicle-treated mice serving as the reference group .
BAL protein analysis
Collected acellular first fraction BAL was sent to Myriad / RBM (Austin, TX) for protein profiling by multiplex immunoassay RodentMAP v3.0. Only select proteins were used for illustration in the manuscript and the complete analysis will be published in a subsequent manuscript.
Right lung sections were cut, stained, and sent for histopathology assessment by Charles River Research Animal Diagnostic Services. Histopathology that was assessed included peribronchiolar inflammation, bronchiolar epithelial hyperplasia, and fibrosis. Sections were scored on a 1–5 scale where 1=minimal and 5=severe.
Statistical analysis of all outcome variables was performed using SAS version 9.3 for Windows. The exposure assessment PBZ EC concentrations were calculated, specific to each facility, by subtracting the area EC background data from the PBZ data. This was done to separate any anthropogenic EC sources from potential CNT/CNF EC exposures. Any samples that were below the limit of detection (LOD) for method 5040 were calculated by taking the LOD of the method and dividing it by two and then calculating the air concentration using the volume of that specific sample. These data were used to calculate the arithmetic means and SD. In order to calculate the geometric means and GSDs, the data were log transformed. The normality of the log-transformed data were assessed using the Shapiro-Wilk test.
Data from rodent studies were log transformed prior to analysis to meet the assumptions of the statistical tests. roc Mix was used to run a three-way factorial analysis of variance. Significant three-way interactions were examined by utilizing two-way ANOVA’s stratified by time. Pairwise comparisons were performed using Fishers Least Significant Difference test. Differences between experimental groups were considered significant with p-values less than 0.05.
Special thanks to Dr. Linda Sargent and Dr. Paul Nicolaysen for their review of the manuscript. In addition, thanks to the NIOSH animal quarter staff, including Amy and Jared Cumpston, for facilitating all the inhalation exposures. Thanks also go out to Dr. Charles B. Clifford, Charles River, for the pathological analysis and images shown in Figure 11.
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.