ToxSci Advance Access originally published online on May 21, 2008
Toxicological Sciences 2008 105(2):368-383; doi:10.1093/toxsci/kfn097
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Genomic Signatures and Dose-Dependent Transitions in Nasal Epithelial Responses to Inhaled Formaldehyde in the Rat
The Hamner Institutes for Health Sciences, Research Triangle Park, North Carolina 27709-2137
1 To whom correspondence should be addressed at Computational Biology Division, The Hamner Institutes for Health Sciences, Six Davis Drive, PO Box 12137, Research Triangle Park, NC 27709-2137. Fax: (919) 558-1300. E-mail: MAndersen{at}Thehamner.org.
Received January 4, 2008; accepted May 10, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Repeated and acute exposure studies assessed time and concentration-dependencies of nasal responses to formaldehyde. Exposures were to 0, 0.7, 2, and 6 ppm for 6 h/day, 5 days/week for up to 3 weeks. Neither cell proliferation nor histopathology was observed at 0.7 ppm. At 6 ppm, cell proliferation increased at the end of the first week (day 5), but not at the end of week 3 (day 15). Squamous metaplasia occurred at day 5; epithelial hyperplasia occurred at both day 5 and day 15. In microarray studies, no genes were altered at 0.7 ppm. At 2 ppm, 15 genes were changed on day 5; only half of them were changed at 6 ppm. No genes were changed significantly at 2 ppm at day 15. The pattern of gene changes at 2 and 6 ppm, with transient squamous metaplasia at day 5, indicated tissue adaptation and reduced tissue sensitivity by day 15. The acute study included an additional concentration (15 ppm) and an instillation group (40 µl, 400mM per nostril). Three times more genes were affected by instillation than inhalation. U-shaped dose responses were noted in the acute study for many genes that were also altered at 2 ppm on day 5. On the basis of cellular component gene ontology benchmark dose analysis, the most sensitive changes were for genes were associated with extracellular components and plasma membrane. With formaldehyde, there are temporal and concentration-dependent transitions in epithelial responses and genomic signatures between 0.7 and 6 ppm. Low concentrations primarily affect extracellular matrix or external plasma membrane portions of the epithelium.
Key Words: genomics; mode of action; formaldehyde; epithelial cell responses; progression; phenotypic anchoring; dose-dependent transitions.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Formaldehyde is a key, high volume industrial chemical due to its versatility as an intermediate in the manufacture of numerous products including urea-formaldehyde, phenol-formaldehyde, and melamine-formaldehyde resins that are used as adhesives in the production of particle board, medium density fiberboard, an plywood. Because of its, chemical reactivity it has also been used for preservation and disinfection (Schulte et al., 2006
). However, formaldehyde is a normal product of intermediary metabolism in mammals, formed endogenously from serine, methionine, choline, and glycine by demethylation of N-, O-, and S-methyl compounds. It is present at concentrations near 0.1–0.2mM in blood and tissues (Heck et al., 1982, 1985). Airborne formaldehyde is a potent eye and respiratory tract irritant at concentrations of several ppm. For example, the mouse RD50, that is, the concentration causing a 50% decrease in respiratory rate, was 3.1 ppm (Buckley et al., 1984). In 2-year inhalation studies, formaldehyde increased squamous cell cancer in the nose of rats (Kerns et al., 1983; Monticello et al., 1996) at concentrations of 6 ppm or higher. Due to its high reactivity with water (it forms a reversible hydrate), formaldehyde is taken up readily into epithelial tissues as it passes through the nose and has a significant anterior to posterior concentration gradient along the nasal epithelium (Kimbell et al., 1993). Histological lesions in acute, subchronic, and chronic inhalation studies confirm the expectation of a front to back gradient in severity of tissue pathology in the nasal passages of rodents.
In addition to increases in cell proliferation at 6 ppm and higher, formaldehyde also increased DNA-protein cross-links (Casanova et al., 1994). However, the carcinogenicity of inhaled formaldehyde in rats appears to arise primarily from enhanced cell proliferation due to cytotoxic responses to formaldehyde rather than from a mutagenic potential of the cross-links (Conolly et al., 2003). Formaldehyde responses of tissues show dose-dependent transitions in mode of action (Slikker et al., 2004a, b). The nasal tissues already have a level of endogenous formaldehyde and low concentration exposures are not expected to cause any appreciable increase above background. At some intermediate inhaled concentration, the exposure should lead to increases in tissue levels and initiate the first signs of cellular interactions associated with formaldehyde reactivity with cellular constituents. Finally, at sufficiently high inhaled concentrations of 6 ppm and above, tissue concentrations are expected to increase significantly and produce cytotoxicity, inflammatory responses, and carcinogenicity noted in the long-term repeated exposures. Key questions for risk assessments for formaldehyde are to ascertain at what concentrations and exposure durations are perturbations caused by formaldehyde (1) sufficient to lead to initial biological responses and (2) when do pertubations become sufficiently great to cause frank cytotoxicity, extensive cross-linking, and carcinogenicity. A second question in assessing dose-dependent transitions is in assessing the qualitative relationship between the initial cellular responses compared with the later cytotoxic, proliferative tissue responses to formaldehyde.
In the past, dose-response studies with formaldehyde have evaluated histopathology, DNA-protein adducts, and cell proliferation. High content data methodologies such as gene transcript profiling and proteomics can look at a broader suite of alterations in epithelial biology at different inhaled concentrations. Hester et al. (Hester et al., 2002, 2003, 2005) assessed gene expression in rat nasal epithelium after instillation of a high concentration formaldehyde solution using a custom array with a restricted number of genes. In the present study we have coupled evaluation of histopathological changes in nasal tissues of formaldehyde-exposed rats with changes in gene expression using a full-genome rat microarray. The tissue pathology served as a phenotypic anchor for interpretation of the microarray results. Our goal was to determine if gene expression profiling at different exposure concentrations would confirm dose-dependent transitions in epithelial cell responses to formaldehyde.
Our study design first assessed the time course of tissue changes over a 3-week exposure period with exposures similar to the three lowest concentrations used in the cancer bioassays—0, 0.7, 2, and 6 ppm, 5 days/week for up to 3 weeks. In addition, a more complete dose-response study was accomplished for single 6-h exposures by adding groups for 15 ppm inhalation and for instillation at a dosage similar to that used by Hester et al. (2003). This experimental design allowed for (1) observation of time dependence of histopathology and gene changes for a 3-week period over a concentration range from 0.7 to 6 ppm; (2) evaluation of the dose response following a single inhalation exposure of up to 15 ppm; and (3) comparison of instillation with high concentration inhalation. Our results clearly demonstrate temporal and dose-dependent transitions in nasal epithelial responses to formaldehyde. The most sensitive targets appear to be associated with the external cell membrane and cell-matrix interactions.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Rats.
Male (F344)/CrlBR rats were purchased from Charles River Breeding Laboratories, Inc. (Wilmington, MA). This strain was chosen due to previous work with formaldehyde exposures. The animals were 6–8 weeks old on receipt weighing 114–221 g. Animals were acclimated for at least 14 days. During the acclimation period the animals were fed the same diet used during the study and observed for clinical signs of disease. Each animal on study was uniquely identified by a numbered steel ear tag and individually housed.
Animal husbandry.
Rats were housed in the inhalation chambers in suspended stainless steel caging with wire mesh floors during both the acclimation and exposure phases of the study. Environmental parameters were maintained within specified ranges: temperatures from 18 to 26°C, relative humidity from 30 to 70%, and air flows from 12 to 15 air changes per hour (National Res. Council, 1996). Automatic light controls provided fluorescent lighting for a 12-h photoperiod (approximately 0700–1900 h light phase). NIH-07 certified diet (Zeigler Brothers, Inc., Gardners, PA) and deionized water (Hydro Service and Supplies, Durham, NC) were available ad libitum, except food was withheld during exposures.
Test substance.
Paraformaldehyde (CAS# 30525-89-4) (Lot# 035K1609) was purchased from Sigma-Aldrich (St Louis, MO). Certified lot-specific purity was listed by the supplier as 102%, indicating that the purity was essentially 100%. In-house gas chromatography/mass spectrometry (GC/MS) methods were used to check for purity and identity. No impurities were detected by GC/MS aside from that expected from atmospheric gases (oxygen and nitrogen) and the spectra were consistent with reference spectra for formaldehyde.
Formaldehyde exposures.
Rats were exposed to target concentrations of 0.7, 2, 6, or 15 ppm formaldehyde in 8-m3 stainless steel and glass chambers. Controls were exposed to filtered air. Chambers had air at a flow rate of approximately 12–15 air changes each hour and the airflow rate was monitored and recorded. Total airflow through the chambers was monitored by using the pressure drop across an orifice at the inlet of the chamber and adjustments were made with a previously calibrated exhaust fan control. All chambers were maintained at a slightly negative pressure. The temperature and relative humidity of the exposure chambers were monitored continuously during exposure and nonexposure periods. The average temperature, relative humidity, airflow, and static pressure over a 30-min period were recorded and a report of the environmental parameter data was printed every day for each exposure period.
Exposure atmospheres were generated by thermal depolymerization of paraformaldehyde and delivery of the vapor to the chamber air supply. Atmosphere concentrations of formaldehyde were monitored during the 6-h exposure using a calibrated infrared analyzer (Miran 1A, Foxboro, MS). In addition to the infrared analyzer, the control chamber atmosphere was monitored for formaldehyde using high pressure liquid chromatography based on the reaction of formaldehyde with dinitrophenyl hydrazine and subsequent analysis of the reaction product. This analytical technique has sensitivity in the ppb range.
Instillation exposure.
Formaldehyde solution was prepared in deionized water at a concentration of 400mM. Rats were anesthetized with isofluorane and 40 µl of formaldehyde solution were instilled into the nostril at a point just inside the nares. Control animals received deionized water. The animals were sacrificed 6 h after the instillation exposure.
Study design.
The results presented cover two characteristics of gene expression associated with formaldehyde exposure. The first portion was a repeated exposure study, with exposures to 0, 0.7, 2, and 6 ppm for up to 3 weeks. The second portion was a broader dose response for a single acute exposure. The dose response after a single 6-h exposure included a group exposed to 15 ppm formaldehyde and the instillation group. For the inhalation studies, animals were exposed for 6 h/day, 5 days/week, for a maximum of 3 weeks. In the time course study, animals were divided into three experimental groups (0.7, 2, and 6 ppm) and one control group. A weight randomization procedure in Provantis (Instem; 161 Washington Street, Suite 1550, Conshohocken, PA, 19428) assigned animals to each dose group (Table 1). The selection of the extra animals for the acute, dose-response study were based on postacclimation body weights and were separated into four experimental groups including a 15 ppm inhalation exposure with a corresponding air control and an instillation exposure group and a deionized water control.
|
Six sacrifices were scheduled: at the end of 6 h of exposure (day 1), 18 h after 6 h of exposure before the start of the second exposure (day 1 recovery), at the end of 5 consecutive days of exposure (day 5), at the end of 6 days exposure (day 6), 18 h after the end of the sixth exposure (day 6 recovery) and at the end of 15 days of exposure (day 15). Euthanasia was by exsanguination of animals in deep anesthesia induced by intraperitoneal injection of sodium pentobarbital (50 mg/kg). The 15 ppm and instillation animals were euthanized, respectively, at the end of the 6-h exposure or 6 h after the instillation of formaldehyde.
Nasal cell proliferation.
Animals were implanted with osmotic pumps (pump model 2ML1, delivering10 µl/h) (Alza, Palo Alto, CA) containing 5 mg/ml bromodeoxyuridine (BrdU) 3 days before euthanasia. The nasal cavities of euthanized animals were retrogradely flushed with 10% neutral buffered formalin. The whole nose was fixed in 10% neutral buffered formalin for 72 h. Following fixation the noses were rinsed and decalcified (Immunocal, Decal Chemical, Tallman, NY) for 3–4 days. Decalcified noses were trimmed and embedded in paraffin. A piece of duodenum (BrdU incorporation in the duodenum on the same tissue section) served as a positive control for immunohistochemical staining and delivery of BrdU to the animal).
Multiple standardized sections were cut and mounted on coated glass slides and stained for the presence of incorporated BrdU. Immunostaining was performed with an automated immunostainer (Biotek Instruments, Inc., Winooski, VT, 05404) using a commercially available monoclonal antibody to BrdU (Caltag Labs, Burlingame, CA). Tissue sections were deparaffinized and rehydrated. DNA was denatured and the sections trypsin-digested prior to staining. The unit length labeling indices (ULLI) were determined for select regions by counting the number of labeled cells at each site and dividing by the length of basement membrane of the region where the cells were counted (Monticello et al., 1990). Cell proliferation at the sites described by Monticello et al. (1991), for both sides of the nose (Supplemental materials F1), was determined for each of eight animals in the day 5 and day 15 groups. The sites counted were: level I, level II—anterior lateral meatus (lateral wall), anterior septum, and the medial aspect of the maxilloturbinate and level III—posterior lateral meatus and posterior septum. In some instances, labeling index (LI) based on the number of labeled cells expressed as a percentage of the total of labeled and unlabeled cells (LI) was also collected.
Histopathology.
Decalcified noses were trimmed and embedded in paraffin. Tissues were subjected to routine microtomy at 5 µm and were hematoxylin-eosin (H&E) stained. The H&E-stained nasal sections from all exposure groups and all time points were evaluated for formaldehyde-induced histopathologic changes at levels I, II, and III.
Microarray analysis.
The noses from animals designated for genomic analyses were dissected to isolate the regions of high formaldehyde flux (Kimbell et al., 1993) corresponding to sites with a high incidence of nasal tumors in chronic exposure studies (Kerns et al., 1983; Monticello et al., 1991, 1996). These regions (Supplemental materials F2) include the lateral meatus and nasoturbinate. Dissected nasal portions were rinsed with cold phosphate-buffered saline (PBS) to remove blood. A mixture of proteases (5 ml), consisting of collagenase (80 units/ml, Sigma-Aldrich) and pronase (10 mg/ml, Sigma-Aldrich) in Hams F12 medium (GIBCO, Carlsbad, CA) buffered with 0.1M N-2-Hydroxylethylpiperazine-N'-2ethane sulfonic acid, was added to the noses and incubated at approximately 37°C for 40 min. Following incubation the mixtures were vigorously vortexed and the resulting cell suspensions collected in clean centrifuge tubes. The suspensions were centrifuged (200 x g, 10 min, 10°C) and the cell pellet was resuspended in 50 µl of PBS and 200 µl of RNALater (Qiagen, Valencia, CA). Nasal cell tissue samples were spun down in a microcentrifuge at full speed for 8–10 min to pellet cells. Each sample was homogenized and processed through the tissue protocol of the Qiagen R Neasy Micro RNA isolation kit (Qiagen) with an RNA elution volume of 17 µl. Resulting total RNA was quantified using NanoDrop ND-1000 (NanoDrop Technologies, Inc., Montchanin, DE) and quality was assessed using an Agilent 2100 Bioanalyzer (Palo Alto, CA).
RNA samples from day 1, day1 recovery, day 5, day 6, and day 15 groups, along with the instillation group, were processed for array analysis. Briefly, double-stranded cDNA was synthesized from 2 to 3 µg of total RNA using the GeneChip One-Cycle cDNA Synthesis Kit based on the manufacturer's protocol (Affymetrix, Santa Clara, CA). Synthesized cDNA template was transcribed to biotin-labeled cRNA using the GeneChip IVT Labeling Kit (Affymetrix). Fifteen µg of labeled cRNA was fragmented and hybridized to Affymetrix Rat Genome 230 2.0 arrays in the Hybridization Oven 640 (Affymetrix) for 16 h at 45°C. After hybridization, arrays were washed using the GeneChip Fluidics Station 450 (Affymetrix) and scanned with the GeneChip Scanner 3000 (Affymetrix). The gene expression omnibus accession number for these microarray studies is GSE7002 [NCBI GEO] .
Statistical evaluations.
Statistical analyses were performed using JMP and SAS Statistical Software (SAS, SAS campus Drive, Bldgs, Cary, NC, 27513) or other statistical programs, as deemed appropriate. A probability value of 0.05 was used as the critical level of significance within each statistical test. Microarray gene expression data were preprocessed using RMA (robust multichip averaging) with a log base 2 transformation (Irizarry et al., 2003a, b). Statistical analysis of gene expression microarray data was performed in R using the affylmGUI package (Smyth, 2004, 2005). To identify genes with significant changes in expression at each time point as a function of dose, all doses and time points associated were analyzed using a linear model with contrasts between animals at each exposure concentration and corresponding control. Probability values were adjusted for multiple comparisons using a false discovery rate of 5% (FDR = 0.05) (Reiner et al., 2003). Genes identified as statistically significant were subject to an additional filter by selecting only those genes that exhibited a 
1.5-fold change from the control group.
Benchmark dose analysis.
For the day 1 time point, genomic benchmark dose (BMD) analysis was performed as previously described (Thomas et al., 2007). For this analysis, the log2 transformed data were analyzed using a one-way ANOVA to identify probe sets on the array for which there was a difference in expression associated with treatment (dose). Probability values were adjusted for multiple comparisons using a FDR of 5% (Benjamini and Hochberg, 1995). Only probe sets identified as significant based on the ANOVA were used in the subsequent BMD analysis. The probe sets identified by the ANOVA were fit as continuous data to a series of four different dose-response models—linear, 2° polynomial, 3° polynomial, and power models. Each model was run assuming constant variance and the benchmark response factor was set to 1.349 multiplied by the standard deviation in the control animals. For model selection, a nested likelihood ratio test was performed on the linear, 2° polynomial, and 3° polynomial models. If the more complex model provided a significantly improved fit (p < 0.05), the more complex model was selected. If the more complex model did not provide a significantly improved fit (p 0.05), the simpler model was selected (Posada and Buckley, 2004). The Akaike information criterion (AIC) for the selected polynomial model was then compared with the AIC for the power model. The model with the lowest AIC (Akaike, 1973) was selected as the final model and was used to calculate a BMD and BMDL. To avoid model extrapolation, probe sets with a BMD value greater than the highest dose (i.e., 15 ppm) were removed from further analysis. The gene annotations and gene ontology (GO) classifications for all probe sets on the Affymetrix Rat Genome 230 2.0 array were downloaded from the Affymetrix web site. Affymetrix probe sets that demonstrated significant dose-response behavior and had a BMD < 15 ppm were converted into unique genes based on their National Center for Biotechnology Information Entrez Gene ID. When two or more probe sets were associated with a single gene, the BMDs and BMDLs for the individual probe sets were averaged to obtain a single BMD and BMDL. The Entrez Gene identifiers were then matched to their corresponding cellular component GO categories. GO categories with less than three genes were removed from the analysis. The three-gene cut-off was based on the minimum number required to define a mean and standard deviation for the category.
GO enrichment analysis.
GO evaluations for classifying genes affected by exposure to formaldehyde were done using Database for Annotation, Visualization and Integrated Discovery Bioinformatics Resources at the National Institutes of Allergy and Infectious Disease (http://david.abcc.ncifcrf.gov/) with a 0.05 criterion for statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Inhalation Exposures
Mean daily concentrations of formaldehyde in the inhalation exposures were approximately 15–20% less than the target concentrations and were calculated using postexposure calibrations of the (infrared) analyzer output (Table 2).
|
Gross Observations and Histopathology
There were no significant differences in terminal body weights between formaldehyde-treated animals and concurrent controls. In this study, few treatment-related lesions were found and unequivocal treatment-related changes were evident only at the 6 ppm formaldehyde exposures (Table 3). A recently completed, 90-day repeated exposure inhalation study has found exposure related effects at 2 ppm at level I (Fig. 1) in the nose (Andersen et al., unpublished observations). In this present study, formaldehyde-induced lesions were observed primarily in the transitional and respiratory epithelium, were typically symmetrically distributed, and displayed an anterior to posterior gradient. The predominant lesions observed were inflammatory cell infiltration, epithelial hyperplasia, and epithelial squamous metaplasia. Minimal inflammatory infiltrates were frequently noted in controls and all treated groups. Consistent increases in the frequency and severity of this lesion were only evident in the 6 ppm exposure groups.
|
|
Day 1
At level I, animals treated with 6 ppm formaldehyde had a minimal to mild inflammatory cell infiltrate of the maxilloturbinate and often the nasoturbinate. At 2 ppm, a minimal inflammatory cell infiltrate of the maxilloturbinate was present at level I in six of eight animals.
Day 1 Recovery
Some animals from each of the four exposure concentrations had minimal inflammatory cell infiltration of the maxilloturbinate at level I. At 6 ppm, two of the eight animals had an inflammatory infiltrate of the nasoturbinate at level I. Epithelial hyperplasia of the lateral wall of level II was also observed.
Day 5
The most dramatic histological changes were evident at nose level I (i.e., inflammatory cell infiltration, epithelial cell hyperplasia, and squamous metaplasia of the respiratory epithelium). These changes were more pronounced in the maxilloturbinate than the nasoturbinate. Epithelial hyperplasia was evident at the 6 ppm exposure in the lateral wall of level II (Supplemental materials F3). A single rat from the 6 ppm exposure group had epithelial hyperplasia of the lateral wall at level III and two animals had hyperkeratosis of the maxilloturbinate at level I.
Day 6
Some animals from each of the four exposure groups had minimal inflammatory cell infiltration of the maxillotubinate at level I. Four animals exposed to 6 ppm formaldehyde had an inflammatory cell infiltrate of the nasoturbinate at level I. Eight animals at the 6 ppm dose level had epithelial hyperplasia of the lateral wall of level II.
Day 6 Recovery
Some animals from each of the four groups had an inflammatory cell infiltration of the maxilloturbinate at level I. Five animals receiving 6 ppm formaldehyde had an inflammatory infiltrate of the nasoturbinate at level I. Eight animals at the 6 ppm and two animals at the 2 ppm dose level had epithelial hyperplasia of the lateral wall of level II.
Day 15
Animals exhibited similar findings to the previous sacrifice times with a more pronounced inflammatory reaction in the anterior nose and epithelial hyperplasia of the lateral wall of level II in the 6 ppm formaldehyde animals. One animal of the 6 ppm exposure group had hyperkeratosis of the maxilloturbinate at level I. Two animals from the 2 ppm exposure group had epithelial hyperplasia of the lateral wall at level II.
High-Dose Inhalation versus Instillation
Although it was only a single exposure, the tissue responses with instillation were evident in more distal portions of the nose compared with 15 ppm exposure. At 15 ppm, all 8 animals had inflammatory infiltrate at the very tip of the nose and at level I. In addition, at level II, seven of eight animals showed inflammatory infiltrates, 2/8 had ulcerative lesions and 1/8 had epithelial hyperplasia. With instillation, the incidence of inflammatory infiltrates were 6/8, 4/8, 8/8 and 7/8, respectively for the tip, and levels I through III.
Cell Proliferation
Proliferation was evaluated from the immunostained nasal epithelium (Supplemental Materials F4). Because there were no significant differences between counts from the right and left sides of the nose for ULLI, the data were combined for the purpose of analysis. There was an increase in cell proliferation in control animals with time for levels I and II and III (Table 4). The histopathological and cell proliferation changes at levels II and III provide phenotypic anchoring for the changes in gene expression evaluated in these animals. Tissues from level I were not included in the areas taken for genomic studies.
|
Day 5
Significantly elevated levels of cell proliferation (ULLI) were observed at 6 ppm at level I and at all sites examined in levels II and III. A significant increase in the traditional LI was also observed at 6 ppm in level I (Table 4).
Day 15
At levels I, II, and III, no significant changes in the ULLI were observed at any of the doses examined. Significant changes were only observed in the LI at level I at 6 ppm (Table 4). Notably, the number of cells per length of basement membrane showed a significant decrease between day 5 and day 15 at level I in the 6 ppm group (Table 5). Both the ULLI and LI measures showed a U-shaped trend across the concentrations at level I in the front of the nose (Table 4). Although this U-shaped response was statistically significant, the variable incidence of mild, inflammatory infiltrate in these groups clouds the biological significance of the U-shaped responses.
|
Microarray Analysis
Looking across all time points and doses, there were approximately 100 genes that showed significantly altered expression. No significant gene expression changes were observed at the 0.7 ppm concentration and minimal changes at 2 ppm at two time points. The majority of changes were observed at 6 ppm. The concentration and time-dependent changes in these various genes altered by exposures to up to 6 ppm for the various exposure days can be best appreciated by viewing these changes in a heat map (Fig. 1).
Day 1
In the 2 ppm group, only 1 gene was significantly increased and none was decreased (Table 6). At 6 ppm, 24 genes were increased and 18 genes were decreased.
|
Day 1 Recovery
No significant changes were observed at any concentration.
Day 5
At 2 ppm, 14 genes were increased and 1 was decreased. At 6 ppm, 28 genes were significantly altered, 24 were increased, and 4 were decreased. Eight genes that had altered expression in the 2 ppm exposure were also altered at 6 ppm.
Day 6
Transcriptional changes were only observed at 6 ppm. A total of nine genes were increased and no genes were decreased.
Day 15
At 6 ppm, 23 genes were increased and 31 were decreased.
High-Dose Inhalation versus Instillation
Both instillation of 400mM and inhalation of 15 ppm formaldehyde altered many more genes than were affected at 6 ppm (Table 6). Instillation also altered more than three times as many genes as the 15 ppm exposure. One clear difference between 15 ppm inhalation and instillation was the degree of overlap among altered genes (Supplemental materials F5). Of the total transcripts increased by 15 ppm formaldehyde by inhalation, 74% (346/465) were also increased by instillation. For the total transcripts decreased by 15 ppm formaldehyde by inhalation, 78% (221/280) were decreased by instillation. At least with the dosage used, 40 µl of a 400mM solution into each nostril, instillation had a much more widespread effect on transcript levels and caused histological responses more distally in the nose. Of the genes increased by instillation, only 32% (346/1086) were increased by inhalation of 15 ppm; for the genes decreased by instillation only 15% (221/1466) were decreased by inhalation.
GO analyses were conducted for the intersection of genes increased by both 15 ppm and instillation. These genes showed enrichment of a large number of GO categories (Supplemental Materials S1). Considering the irritant and cytotoxic properties of formaldehyde, many of the categories that showed enrichment on the basis of biological processes (BP) were not surprising, that is, response to wounding, control and induction of apoptosis, inflammation pathways, and receptor tyrosine kinase signaling. Based on cellular component GO categories, the genes increased by instillation and not by 15 ppm were enriched in many extracellular space and external plasma membrane components (Table 7).
|
BMD Analysis
The BMD analyses provide an estimate of the exposure concentration causing a defined, increase in response to formaldehyde inhalation on the basis of changes in GO categories. An earlier evaluation of the 6-h inhalation microarray results from 0.7 through 15 ppm provided BMD values based on genes from various GO categories (Thomas et al., 2007
). GO classification is based on three main categories: BP, molecular function (MF), and cell component (CC). The original evaluation was completed on the basis of BP and MF. Using the same methods (Yang et al., 2007), we conducted the BMD analysis on this same data set based on cellular components. The full suite of BMD values for GO categories based on cellular components is in supplementary materials to this paper (Supplemental Materials S2). The 15 categories with the lowest BMD values are shown in Table 8.
|
Evaluation of Dose-Response Characteristics for Individual Genes
A second, semiquantitative evaluation was conducted plotting the day 1 dose-response relationships for those genes that were changed at various times and concentrations in the repeated dose study (Fig. 1). These dose-response bar graphs (Figs. 2
–5) show behavior for all the inhalation concentrations and the instillation, considering the instillation to represent a treatment of greater intensity than even the 15 ppm inhalation. Four types of dose response are illustrated: genes that consistently increased over the concentration range (Fig. 2); genes that were significantly increased only above 6 ppm (Fig. 3); genes that tended to U-shaped behavior, initially increasing with concentration and then decreasing at 15 ppm or with instillation (Fig. 4); and genes that tended to U-shaped behavior, initially decreasing with concentration and then increasing at 15 ppm or with instillation (Fig. 5).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Sensitivity of Genomic Responses
This acute and repeated dose study has provided results to compare formaldehyde-induced genomic responses in nasal epithelium with corresponding histopathological changes in these tissues over a range of concentrations for exposures of up to 3 weeks in duration. With the exception of the day 1 time point, the study design focused primarily on genomic changes that occurred at or below the threshold for carcinogenicity (6 ppm). Two chronic inhalation bioassays have been conducted previously; each included a 6 ppm group. There was only one animal with a nasal tumor in the 6 ppm in the two studies combined (Kerns et al., 1983
). Evaluation in the region of concentration below those that are carcinogenic in a large proportion of exposed animals could possibly extend the dose-response curve for precursor responses to lower concentrations. However, the genomic changes did not prove to be more sensitive than tissue responses. No significant gene changes were observed at 0.7 ppm and very few were observed at 2 ppm, and those were primarily seen only at day 5. With formaldehyde, genomic markers provide, at best, only modestly more sensitive measures of tissue response compared with histology. With some of the genes with U-shaped dose responses, a trend toward altered expression at 0.7 ppm was evident, but the changes themselves were not statistically significant (Figs. 4 and 5). Interestingly, the BMD estimates for cell proliferation (4.9 ppm) and tumors (6.4 ppm) reported previously are only 1.5–3 times higher (Schlosser et al., 2003) than the BMD estimates for the most sensitive responses in our cellular component GO category analysis (Table 8).
Temporal Alteration of Tissue Response to Formaldehyde
Based on the comparison of the time course of cell proliferation and histopathology, there appeared to be three key time-dependent responses. First, there is the initial exposure(s) of the nasal epithelium to the irritant in a naïve rat (day 1). Second, at the end of the 5 days of exposure (day 5) there are significant gene changes, even at 2 ppm. All of the 6 ppm animals examined at day 5 showed, inflammatory infiltrates, and hyperplasia and squamous metapalsia at level II. Cell proliferation, measured as ULLI, was increased in this group at the 6 ppm group in level II regions. Third, at the end of 3 weeks (day 15), cell proliferation, based on ULLI, is at control levels and none of the 6 ppm animals have squamous metaplasia, and there are no significant alterations in genes at 2 ppm.
In the day 15 group, there were changes in proliferation noted at level I, a region not included in the tissues taken for genomics. In this group, cell counts along the epithelium were reduced and the LI was increased (Table 4). Another characteristic noted was a tendency to show a mild, but not statistically significant U-shaped behavior at day 15 for cell proliferation at level I (Table 4). This tendency toward U-shaped dose response for proliferation was noted previously (Conolly et al., 2003).
Temporal Alterations in Gene Changes with Formaldehyde
The heat map (Fig. 1) clusters gene changes. The day 1 section contains five columns, including both the 15 ppm and instillation groups. All other groups have only three exposure concentrations. Category A genes tend to be increased in the 6 ppm day 15 group. The changes in these genes were more spotty at day 5 and day 6. At day 19, the increases are mostly seen at the 6 ppm exposures with fewer trends at 0.7 or 2 ppm. Category A included genes associated with responses to stress (Hmox1), cell proliferation (Areg), cell survival (Tnfsfr12a), early response genes (Fosl1), and sulfur reduction (Txnrd1 and Srxn1). These genes also tend to be upregulated at day 1. Genes in category B were increased at 2 and 6 ppm at day 1 but decreased at higher exposure levels. These genes also tended to be increased at day 5, but were less responsive at day 6 and day 15. These gene changes, the most sensitive to formaldehyde inhalation, appear to be associated with a transient tissue response. Downregulated genes did not offer as clear an explanation for tissue responses to formaldehyde. Broadly, category C represented genes that were downregulated consistently by formaldehyde concentration of up to 6 ppm for day 1, day 5 and day 15. Category D genes appeared to be less responsive at 15 ppm and instillation than for 2 or 6 ppm.
When considered together with the cellular responses, the overall patterns of gene expression noted in the heat map (Fig. 1) appear to represent two separate phases of responses. At day 1 for 6 ppm and above, cell injury and response to injury (category A and C genes) prevail. Day 1 responses also have a process with U-shaped response genes (category B). At day 5, the tissue is primarily undergoing adaptation to the irritant stress and has diminished response to high concentration injury at the 6 ppm with more contributions to the genomic signature from the genes altered in a U-shaped pattern. By day 15, the tissue has undergone adaptation, is less responsive to low concentrations, and predominately has the higher dose response to injury also seen at day 1 at 6 ppm (Table 3). These time-dependent gene changes are also concurrent with observations of squamous metaplasia, present at day 5 and absent at day 15. Among the genes in category B with U-shaped dose-response curves, was wee homolog. Wee1 is a kinase that inhibits progression through cell cycle (Okamoto and Sagata, 2007) and may play a role in altered cell proliferation in these animals. Map3k1 (a mitogen-activated protein [MAP] kinase kinase kinase; also called MEKK1) phosphorylates several different proteins, with greatest activity for MKK4, a kinase for the c-Jun NH2-terminal kinases (Xia et al., 2007). This kinase mediates degradation of c-Jun in response to cellular stress and affects apoptotic processes. These gene changes at day 5 for Wee1 and Map3k1, indicating slower progression through the cell cycle, appear inconsistent with increases in ULLI at day 5 (Table 4). However, it bears emphasis that measurements of cell proliferation, as conducted in our studies, integrate information over 3 days. The gene changes are more a snapshot in time. Day 5 alteration in Wee1 and Map3k1 may be more influenced by the instantaneous transcript levels, whereas replication rate represents the averaging of processes over days 3 through 5 of exposure.
The classes of genes with U-shaped dose-response characteristics represent the most sensitive responses to formaldehyde exposures at day 1 and day 5. Among the upregulated genes in this category are several that show clearer responses at 2 ppm, such as Per2, Pum1_predicted, Atrx, Slc25a15 and Arhgap5_predicted (Fig. 4). The downregulated genes in Figure 5—Map3k1, Galnt7, Sec22l2 and Arntl—show these dose related trends at 2 ppm. Per2 and Arntl are PAS-proteins involved in circadian rhythm control (Karman and Tischkau, 2006; Verwey et al., 2007). Pum1 is an RNA-binding protein essential for stem cell maintenance and self-renewal in lower organisms and probably in mammals as well (Spassov and Jurecic, 2003). Atrx, also called Xh2, is a likely binding protein for annexin V-binding which promotes survival of rat cells in vitro (Ohsawa et al., 1996; Park et al., 2004). Slc25a15, solute carrier family member 25 (mitochondrial carrier; ornithine transporter), also called ORNT1, is involved in urea metabolism (Camacho et al., 1999). Argap5_predicted (also known as p190) is a Rho family guanosine triphosphatase (GTPase) that regulates actin stress fiber dynamics and may be linked to completion of cytokinesis and generation of viable cell progeny in replication (Su et al., 2003). Several genes were increased at 2 ppm at day 5—including serpine1, sprouty protein, Prelp, and Xlkd1_predicted. Serpine 1 is a serine proteinase inhibitor; sprouty proteins are inhibitors of fibroblast growth factor signaling (Hashimoto et al., 2002); Prelp appears to serve as a linker between tissue matrix and cell surface proteoglycans (Bengtsson et al., 2000); and Xlkd1 (also LYVE-1), a gene that encodes a type I integral membrane glycoprotein that acts as a receptor, binds to both soluble and immobilized hyaluronan, and is present in lymphatic vessel endothelium (Groger et al., 2007). Of the downregulated genes with U-shaped responses GalNAc transferase 7 controls the initiation of mucin-type O-linked protein glycosylation and transfer of N-acetylgalactosamine to serine and threonine amino acid residues(Bennett et al., 1999). Finally, Sec22l2 is involved in trafficking between cytoplasm and Golgi apparatus (Hay et al., 1996). Thus, most of these genes affected with maximal responses at 2 ppm or trends to increases at 2 ppm are associated with cell membrane, external aspect of the cell membrane, or cell architecture. These genes are not statistically affected at the lower inhaled concentrations at day 15. The most sensitive responses (i.e., those apparent at the lowest concentrations) are consistent with actions of formaldehyde on the periphery of the cell or at the plasma membrane. Responses associated with cellular stress, that is, Hmox1, Tnfsfr12a, or Srxn1 (sulfiredoxin), are only altered at the higher inhaled concentrations.
Our BMD analysis based on cellular components and the earlier analyses on MF and BP ontology categories were also consistent with this interpretation. The BMD method, utilizing GO category classification, analyzes behavior of related genes rather than simply collecting information on changes in individual genes or by categorizing the dose-response behavior for groups of genes empirically (Figs. 2–5
). The mean BMD had been selected as the primary point of comparison based on its ability to aggregate behavior of genes within a GO category. In an earlier evaluation of these day 1 results, the mean BMD for molecular process GO categories was found to be closely associated with the phenotypic effects (Thomas et al., 2007). The BMD10% for the cell LI based on multiple time points was 4.91 ppm (Schlosser et al., 2003). The GO category-based BMD for the positive regulation of cell proliferation was 5.68 ppm. Here, the analysis was extended to include CC categories.
The CC GO categories with the lowest BMDs were associated with the cell membrane, extracellular components, maintenance of cell architecture, and anchorage to cell membrane (Table 8; Supplemental material S 2). These low concentration effects and their persistence beyond Day1 indicate that the initial tissue responses to inhaled formaldehyde likely arise from extracellular reactivity of formaldehyde at or in the vicinity of the cell membrane. Increasing the exposure concentrations leads to elevated intracellular and extracellular formaldehyde resulting in more diverse responses of the affected epithelial cells.
GO Analysis of High-Dose Formaldehyde Responses
The 15 ppm exposure caused changes in a much larger number of genes than did the 6 ppm. The GO categories enriched (Table 7) included transcription, DNA-dependent; response to stress; cell regulation of apoptosis, cell defense; transmembrane signaling and NF-
B cascades. These changes are consistent with irritant damage from formaldehyde and response to damage, including enhanced inflammatory signaling and cell proliferation, at this higher concentration. The earlier analysis of all the day 1 results from 0.7 to 15 ppm produced BMD values based on GO categories for MF and BP (Thomas et al., 2007). Categories for inflammation, apoptosis and cell cycle control were less sensitive markers of exposure than those for certain biochemical processes, including organic acid transport, sodium transport, amino acid transport and ionic homeostasis.
The comparison of instillation with the 15 ppm inhalation exposure also indicated that there are alternate targets at the high exposure concentrations with correspondingly higher tissue exposures. Instillation affected a much larger number of genes and there was enrichment in genes and processes associated with extracellular tissues and the external plasma membrane. Based on our results, the instillation dose used here and in other studies (Hester et al., 2003, 2005) appears to be too high to allow meaningful comparisons even with the high concentration formaldehyde exposures from the bioassays.
The gene expression changes at 6 ppm for day 1 and throughout the repeated dose study corroborate the inflammatory pathology in the nose. At 15 ppm and in the instillation group, these gene changes for response to wounding and inflammation were even more pronounced. Local inflammatory responses of epithelial tissues in the respiratory tract may have more systemic consequences through mediation by tissue cytokines that can be released from the nasal epithelium and circulate through the body. A recent proteomic study in rats exposed to 5 or to 10 ppm formaldehyde, 6 h/day for 2 weeks found changes in plasma proteins, including two cytokines, interleukin-4 and interferon-
, that were dose related (Im et al., 2006). Because inhaled formaldehyde at these concentrations causes no measurable change in blood formaldehyde (Heck et al., 1982), these responses are probably a secondary consequence of the nasal inflammatory process initiated by concentrations of inhaled formaldehyde sufficiently high to initiate inflammatory responses in the nose.
Genomic studies have also evaluated dose related gene expression changes in whole rat lungs after inhalation exposures of 5 and 10 ppm formaldehyde for 2 weeks (Sul et al., 2007). Under these exposure conditions, the only histopathology that has been noted previously in subchronic and chronic rat inhalation studies were in the nasal cavity and in the proximal trachea (Kerns et al., 1983). The gene expression pattern in lung which had only 21 genes with dose-dependent changes was markedly different from our results at any of the exposure concentrations leading to tissue responses within the nose. These tissue responses in the absence of any histological changes in bulk lung tissue are likely to arise from systemic consequences of circulating cytokines associated with inflammation in the target areas within the nasal cavity.
Dose-Dependent Transitions with Formaldehyde
Our genomic results, both for the repeated and the acute dose studies, are supportive of a qualitative schema for the toxicity of formaldehyde that has dose-dependent contributions of extracellular and intracellular concentrations of formaldehyde at various inhaled concentrations (Fig. 6). This scheme considers dose-dependent input rates of formaldehyde to various tissue compartments and clearance from these compartments due to diffusion, metabolism and reactivity. The internal milieu of the cell has higher glutathione and higher formaldehyde dehydrogenase than does the extracellular space. At low rates of intake from inhalation to the epithelium there should be a concentration gradient with higher concentrations in the extracellular spaces. These extracellular regions would be the first to respond to inhaled formaldehyde. From the genomic results, this transition appears to occur between 0.7 and 2 ppm. As inhaled concentrations increase and cellular processes for formaldehyde detoxification become saturated, the inter-compartmental differences in concentration will be lower and intracellular targets would become increasingly important. This dose-dependent transition likely occurs between 2 and 6 ppm. At very high input rates, as occurs with 15 ppm or instillation, there are more extensive excursions of formaldehyde concentrations and widespread damage in the extracellular and the intracellular environment.
|
Nel and colleagues have developed the concept of hierarchical oxidative stress arising from exposures to diesel exhaust particles and other particles (Nel et al., 2006
; Xia et al., 2006; Xiao et al., 2003). In the hierarchy, sufficiently low doses have no discernible effect. The first measurable response is upregulation of cellular-anti-oxidants, a process that is essentially adaptive to permit the tissue to be more robust in the face of oxidant stress. As concentrations increase further, the anti-oxidant control is overwhelmed leading to cellular toxicity and activation of inflammatory pathways. At even higher concentration, activation of apoptotic pathways lead to programmed cell death pathways. Our results with formaldehyde gene expression are also consistent with a hierarchical response to plasma membrane perturbations from formaldehyde progressing through several dose-dependent transitions. The genomic signatures related to these transitions are for cell membrane and extracellular components, then inflammation and cellular stress, and eventually apoptosis. Importantly, these hierarchical models indicate that the tissue responses at low dose concentrations are qualitatively different from those at high concentrations and linear extrapolations or extrapolations that specify similar modes of action at high and low doses would be inappropriate. As more gene expression results become available for formaldehyde responses in the nasal epithelium, it should be possible to more completely map response stress pathways that are affected by changes in cell membrane integrity. Another important component of a mode of action based risk assessment that accounts for dose dependencies of cellular effects, including gene changes, is a tissue dosimetry model for formaldehyde that considers contributions of both endogenous production and metabolism of formaldehyde together with exogenous formaldehyde delivered to nasal tissues by inhalation. The dosimetry model should be developed to understand when inhaled formaldehyde causes significant increases in tissue formaldehyde above endogenous levels (Fig. 6).
Enhancing Sensitivity
Based on our study design tissue were harvested from a fairly large region toward the front of the rat nose. Formaldehyde is highly soluble in epithelial tissues, due to rapid hydration of parent chemical in aqueous environments. High solubility leads to proximal to distal gradients of tissue absorption during inhalation. Some increase in sensitivity for detecting genomic changes might be possible by more deliberate microdissection of tissues at the edge of the area selected for microarray analysis to obtain tissue from more proximal regions. It was our intention here to evaluate tissues representative of the general region from which tumors occurred rather than attempting a specific analysis of results in the most proximal section of the nose with the highest exposure. Further research could focus on examining the gradient of gene expression in a group of animals exposed to 0.7 or 2 ppm to see whether significant differences exists from the front to the back of the larger sections used in the present study. This present study was also limited in time period of observation (3 weeks) and concentration range tested (only up to 6 ppm for the extended exposure periods). We are now completing a 90-day repeated exposures inhalation study at 0, 0.7, 2, 6, 10, and 15 ppm to more fully evaluate the persistence of the tissue changes seen through day 15 and to provide a more complete dose-response analysis including concentration ranges that were carcinogenic in over half the exposed animals. It will be interesting to see if the patterns of transcript changes for day 15 are stable throughout a longer experiment and how the dose-response curves for various gene categories change with duration of exposure to formaldehyde.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Supplementary data are available online at http://toxsci.oxfordjournals.org/.
|
Funding |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
The Long Range Research Initiative of the American Chemistry Council supported aspects of the development of bioinformatics infrastructure
|
ACKNOWLEDGMENTS |
|---|
We thank the Formaldehyde Council Inc. (Arlington, VA) for support of this project and their forward-thinking, proactive approach to product stewardship for formaldehyde. We particularly express our thanks to Drs Stewart Holm and Robert Golden for many valuable discussions throughout the course of this work and in preparation of the final manuscript. We thank Drs Jingbo Pi and Courtney Woods for careful review of the review copy of the paper and many suggestions for improving the paper.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFundingREFERENCES |
|---|
Akaike H. Information theory and an extension of the maximum likelihood principle. In: 2nd International Symposium on Information Theory—Petrovand BN, Casiki F, eds. (1973) Budapest: Academiai Kiado. 267–281.
Bengtsson E, Aspberg A, Heinegard D, Sommarin Y, Spillmann D. The amino-terminal part of PRELP binds to heparin and heparan sulfate. J. Biol. Chem. (2000) 275:40695–40702.
Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R. Stat. Soc. B (1995) 57:289–300.
Bennett EP, Hassan H, Hollingsworth MA, Clausen H. A novel human UDP-N-acetyl-D-galactosamine:polypeptide N-acetylgalactosaminyltransferase, GalNAc-T7, with specificity for partial GalNAc-glycosylated acceptor substrates. FEBS Lett. (1999) 460:226–230.[CrossRef][Web of Science][Medline]
Buckley LA, Jiang XZ, James RA, Morgan KT, Barrow CS. Respiratory tract lesions induced by sensory irritants at the RD50 concentration. Toxicol. Appl. Pharmacol. (1984) 74:417–429.[CrossRef][Web of Science][Medline]
Camacho JA, Obie C, Biery B, Goodman BK, Hu CA, Almashanu S, Steel G, Casey R, Lambert M, Mitchell GA, et al. Hyperornithinaemia-hyperammonaemia-homocitrullinuria syndrome is caused by mutations in a gene encoding a mitochondrial ornithine transporter. Nat. Genet. (1999) 22:151–158.[CrossRef][Web of Science][Medline]
Casanova M, Morgan KT, Gross EA, Moss OR, Heck HA. DNA-protein cross-links and cell replication at specific sites in the nose of F344 rats exposed subchronically to formaldehyde. Fundam. Appl. Toxicol. (1994) 23:525–536.[CrossRef][Web of Science][Medline]
Conolly RB, Kimbell JS, Janszen D, Schlosser PM, Kalisak D, Preston J, Miller FJ. Biologically motivated computational modeling of formaldehyde carcinogenicity in the F344 rat. Toxicol. Sci. (2003) 75:432–447.
Groger M, Niederleithner H, Kerjaschki D, Petzelbauer P. A previously unknown dermal blood vessel phenotype in skin inflammation. J. Invest. Dermatol (2007) 127:2893–2900.[CrossRef][Web of Science][Medline]
Hashimoto S, Nakano H, Singh G, Katyal S. Expression of Spred and Sprouty in developing rat lung. Gene Expr. Patterns (2002) 2:347–353.[CrossRef][Medline]
Hay JC, Hirling H, Scheller RH. Mammalian vesicle trafficking proteins of the endoplasmic reticulum and Golgi apparatus. J. Biol. Chem. (1996) 271:5671–5679.
Heck HD, Casanova-Schmitz M, Dodd PB, Schachter EN, Witek TJ, Tosun T. Formaldehyde (CH2O) concentrations in the blood of humans and Fischer-344 rats exposed to CH2O under controlled conditions. Am. Ind. Hyg. Assoc. J. (1985) 46:1–3.[Web of Science][Medline]
Heck HD, White EL, Casanova-Schmitz M. Determination of formaldehyde in biological tissues by gas chromatography/mass spectrometry. Biomed. Mass. Spectrom. (1982) 9:347–353.[CrossRef][Web of Science][Medline]
Hester SD, Barry WT, Zou F, Wolf DC. Transcriptomic analysis of F344 rat nasal epithelium suggests that the lack of carcinogenic response to glutaraldehyde is due to its greater toxicity compared to formaldehyde. Toxicol. Pathol. (2005) 33:415–424.
Hester SD, Benavides GB, Sartor M, Yoon L, Wolf DC, Morgan KT. Normal gene expression in male F344 rat nasal transitional and respiratory epithelium. Gene (2002) 285:301–310.[CrossRef][Web of Science][Medline]
Hester SD, Benavides GB, Yoon L, Morgan KT, Zou F, Barry W, Wolf DC. Formaldehyde-induced gene expression in F344 rat nasal respiratory epithelium. Toxicology (2003) 187:13–24.[CrossRef][Web of Science][Medline]
Im H, Oh E, Mun J, Khim JY, Lee E, Kang HS, Kim E, Kim H, Won NH, Kim YH, et al. Evaluation of toxicological monitoring markers using proteomic analysis in rats exposed to formaldehyde. J. Proteome Res. (2006) 5:1354–1366.[CrossRef][Web of Science][Medline]
Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. (2003a) 31:e15.
Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics (2003b) 4:249–264.[Abstract]
Karman BN, Tischkau SA. Circadian clock gene expression in the ovary: Effects of luteinizing hormone. Biol. Reprod. (2006) 75:624–632.
Kerns WD, Pavkov KL, Donofrio DJ, Gralla EJ, Swenberg JA. Carcinogenicity of formaldehyde in rats and mice after long-term inhalation exposure. Cancer Res. (1983) 43:4382–4392.
Kimbell JS, Gross EA, Joyner DR, Godo MN, Morgan KT. Application of computational fluid dynamics to regional dosimetry of inhaled chemicals in the upper respiratory tract of the rat. Toxicol. Appl. Pharmacol. (1993) 121:253–263.[CrossRef][Web of Science][Medline]
Monticello TM, Miller FJ, Morgan KT. Regional increases in rat nasal epithelial cell proliferation following acute and subchronic inhalation of formaldehyde. Toxicol. Appl. Pharmacol. (1991) 111:409–421.[CrossRef][Web of Science][Medline]
Monticello TM, Morgan KT, Hurtt ME. Unit length as the denominator for quantitation of cell proliferation in nasal epithelia. Toxicol. Pathol. (1990) 18:24–31.[Web of Science][Medline]
Monticello TM, Swenberg JA, Gross EA, Leininger JR, Kimbell JS, Seilkop S, Starr TB, Gibson JE, Morgan KT. Correlation of regional and nonlinear formaldehyde-induced nasal cancer with proliferating populations of cells. Cancer Res. (1996) 56:1012–1022.
National Research Council. Guide for the care and use of labaratory animals. (2006) Washington Street, Suite 1550, Conshohockena, PA, 19428.
Nel A, Xia T, Madler L, Li N. Toxic potential of materials at the nanolevel. Science (2006) 311:622–627.
Ohsawa K, Imai Y, Ito D, Kohsaka S. Molecular cloning and characterization of annexin V-binding proteins with highly hydrophilic peptide structure. J. Neurochem. (1996) 67:89–97.[Web of Science][Medline]
Okamoto K, Sagata N. Mechanism for inactivation of the mitotic inhibitory kinase Wee1 at M phase. Proc. Natl. Acad. Sci. U. S. A. (2007) 104:3753–3758.
Park DJ, Pask AJ, Huynh K, Renfree MB, Harley VR, Graves JA. Comparative analysis of ATRX, a chromatin remodeling protein. Gene (2004) 339:39–48.[CrossRef][Web of Science][Medline]
Posada D, Buckley TR. Model selection and model averaging in phylogenetics: Advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Syst. Biol. (2004) 53:793–808.
Reiner A, Yekutieli D, Benjamini Y. Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics (2003) 19:368–375.
Schlosser PM, Lilly PD, Conolly RB, Janszen DB, Kimbell JS. Benchmark dose risk assessment for formaldehyde using airflow modeling and a single-compartment, DNA-protein cross-link dosimetry model to estimate human equivalent doses. Risk Anal. (2003) 23:473–487.[CrossRef][Web of Science][Medline]
Schulte A, Bernauer U, Madle S, Mielke H, Herbst U, Richter-Reichhelm H-B, Appel KE, Gundert-Remy U. Assessment of the Carcinogenicity of Formaldehyde [CAS No. 50-00-0] (2006) Berlin: Bundesinstitut fur Riskobewertung.
Slikker W Jr, Andersen ME, Bogdanffy MS, Bus JS, Cohen SD, Conolly RB, David RM, Doerrer NG, Dorman DC, Gaylor DW, et al. Dose-dependent transitions in mechanisms of toxicity: Case studies. Toxicol. Appl. Pharmacol. (2004a) 201:226–294.[CrossRef][Web of Science][Medline]
Slikker W Jr, Andersen ME, Bogdanffy MS, Bus JS, Cohen SD, Conolly RB, David RM, Doerrer NG, Dorman DC, Gaylor DW, et al. Dose-dependent transitions in mechanisms of toxicity. Toxicol. Appl. Pharmacol. (2004b) 201:203–225.[CrossRef][Web of Science][Medline]
Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. (2004) 3. Article 3.
Smyth GK. Limma: Linear models for microarray data. In: Bioinformatics and Computational Biology Solutions using R and Bioconductor—Gentleman R, Carey V, Dudoit S, Irizarry RA, Huber W, eds. (2005) Springer, New York, 397–420.
Spassov DS, Jurecic R. Mouse Pum1 and Pum2 genes, members of the Pumilio family of RNA-binding proteins, show differential expression in fetal and adult hematopoietic stem cells and progenitors. Blood Cells Mol. Dis. (2003) 30:55–69.[CrossRef][Web of Science][Medline]
Su L, Agati JM, Parsons SJ. p190RhoGAP is cell cycle regulated and affects cytokinesis. J. Cell Biol. (2003) 163:571–582.
Sul D, Kim H, Oh E, Phark S, Cho E, Choi S, Kang HS, Kim EM, Hwang KW, Jung WW. Gene expression profiling in lung tissues from rats exposed to formaldehyde. Arch. Toxicol. (2007) 81:589–597.[CrossRef][Web of Science][Medline]
Thomas RS, Allen BC, Nong A, Yang L, Bermudez E, Clewell HJ 3rd, Andersen ME. A method to integrate benchmark dose estimates with genomic data to assess the functional effects of chemical exposure. Toxicol. Sci. (2007) 98:240–248.
Verwey M, Khoja Z, Stewart J, Amir S. Differential regulation of the expression of Period2 protein in the limbic forebrain and dorsomedial hypothalamus by daily limited access to highly palatable food in food-deprived and free-fed rats. Neuroscience (2007) 147:277–285.[CrossRef][Web of Science][Medline]
Xia T, Kovochich M, Nel A. The role of reactive oxygen species and oxidative stress in mediating particulate matter injury. Clin. Occup. Environ. Med. (2006) 5:817–836.[Medline]
Xia Y, Wang J, Xu S, Johnson GL, Hunter T, Lu Z. MEKK1 mediates the ubiquitination and degradation of c-Jun in response to osmotic stress. Mol. Cell. Biol. (2007) 27:510–517.
Xiao GG, Wang M, Li N, Loo JA, Nel AE. Use of proteomics to demonstrate a hierarchical oxidative stress response to diesel exhaust particle chemicals in a macrophage cell line. J. Biol. Chem. (2003) 278:50781–50790.
Yang L, Allen BC, Thomas RS. BMDExpress: A software tool for the benchmark dose analyses of genomic data. BMC Genomics (2007) 8:387.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Rouquie, C. Friry-Santini, F. Schorsch, H. Tinwell, and R. Bars Standard and Molecular NOAELs for Rat Testicular Toxicity Induced by Flutamide Toxicol. Sci., May 1, 2009; 109(1): 59 - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Boverhof, B. B. Gollapudi, J. A. Hotchkiss, M. Osterloh-Quiroz, and M. R. Woolhiser Evaluation of a toxicogenomic approach to the local lymph node assay (LLNA) Toxicol. Sci., February 1, 2009; 107(2): 427 - 439. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||