Toxicology 187 (2003) 13
/24

www.elsevier.com/locate/toxicol

Formaldehyde-induced gene expression in F344 rat nasal

respiratory epithelium
Susan D. Hester a,c,*, Gina B. Benavides b, Lawrence Yoon b,
Kevin T. Morgan b, Fei Zou d, William Barry d, Douglas C. Wolf a
a

US Environmental Protection Agency, Research Triangle Park, NC, USA

b
GlaxoSmithKline, Inc., Research Triangle Park, NC, USA
c
Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, NC, USA
d
Department of Biostatistics, School of Public Health, University of North Carolina, Chapel Hill, NC, USA
Received 17 October 2002; received in revised form 24 December 2002; accepted 2 January 2003

Abstract
Formaldehyde (FA), an occupational and environmental toxicant used extensively in the manufacturing of many
household and personal use products, is known to induce squamous cell carcinomas in the nasal turbinates of rats and
mice and squamous metaplasia in monkey noses. Tissue responses to FA include a dose dependent epithelial
degeneration, respiratory cell hypertrophy, and squamous metaplasia. The primary target for FA-induced toxicity in
both rodents and monkeys is the respiratory nasal epithelium. FA increases nasal epithelial cell proliferation and
DNA
/protein crosslinks (DPX) that are associated with subsequent nasal cancer development. To address the acute
effects of FA exposure that might contribute to known pathological changes, cDNA gene expression analysis was used.
Two groups of male F344 rats received either 40 ul of distilled water or FA (400 mM) instilled into each nostril. Twentyfour hours following treatment, nasal epithelium was recovered from which total RNA was used to generate cDNA
probes. Significance analysis of microarrays (SAM) hybridization data using ClontechTM Rat Atlas 1.2 arrays revealed
that 24 of the 1185 genes queried were significantly up-regulated and 22 genes were significantly downregulated. Results
for ten of the differentially expressed genes were confirmed by quantitative real time RT PCR. The identified genes with
FA-induced change in expression belong to the functional gene categories xenobiotic metabolism, cell cycle, apoptosis,
and DNA repair. These data suggest that multiple pathways are dysregulated by FA exposure, including those involved
in DNA synthesis/repair and regulation of cell proliferation. Differential gene expression profiles may provide clues
that could be used to define mechanisms involved in FA-induced nasal cancer.
# 2003 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Gene expression; F344 rat; cDNA array; Nasal epithelium; Rodent nasal cancer; Formaldehyde

Abbreviations: FA, formaldehyde; DPX, DNA
/protein crosslink; Min, minute; mg, microgram; mM, millimolar; 8C, centrifuge; ml,
milliliter.

Supplementary data associated with this article can be found at doi:10.1016/S0300-483X(03)00008-8
* Corresponding author. Tel.: /1-919-541-1320; fax: /1-919-541-0694.
E-mail address: hester.susan@epa.gov (S.D. Hester).
0300-483X/03/$ - see front matter # 2003 Elsevier Science Ireland Ltd. All rights reserved.
doi:10.1016/S0300-483X(03)00008-8

14

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

1. Introduction
Formaldehyde (FA) is a common environmental
contaminant found in tobacco smoke, paint,
garments, medicinal and industrial products, and
is a component of diesel and gasoline exhaust
(Flyvholm and Andersen, 1993; Quievryn and
Zhitkovich, 2000). Although there is limited evidence that FA is a human carcinogen, it is well
known that FA induces tumors in rodents. FA
exposure leads to the formation of DNA
/protein
crosslinks (DPX), a major form of DNA damage
and the presence of DPX have been used as a
measure of delivered dose (Heck and Casanova,
1999; Heck et al., 1990).
FA is also cytotoxic in the rodent nose, inducing
sustained increases in cell turnover (proliferation)
at concentrations that induce tumors (Monticello
et al., 1991b; Morgan et al., 1986). Lesions
primarily develop in the epithelium lining the nasal
septum and the ventral meatus of the turbinates.
The types of lesions following an acute 24 h nasal
instillation of 400 mM FA were squamous metaplasia, ulceration, hyperplasia, and goblet cell
hypertrophy, which were observed at 3 weeks
post treatment (St. Clair et al., 1990). These lesions
occurred only in respiratory and transitional
epithelium lining the anterior nasal passages of
the rat nose. This suggests that respiratory and
transitional epitheliums are more susceptible to
FA damage than the other epithelial types present
in the nose.
In addition to DPX, FA can readily react with
DNA and RNA by forming cross-links that
connect bases exclusively through exocyclic amino
groups (Chaw et al., 1980; Kennedy et al., 1996;
Takahashi and Hashimoto, 2001). FA is known to
induce regenerative proliferation (Monticello et
al., 1991a,b) and may alter apoptosis (Szende et
al., 1995, 1998; Tyihak et al., 2001). An alteration
in the balance between cell growth and cell death
may contribute to the development of squamous
cell carcinoma in the rat nose. Biologically based
risk-assessment models (CIIT, 1999) have been
proposed which integrate levels of cell proliferation, cell death, and mutation rates associated with
FA treatment (Andersen et al., 1992; Conolly and
Andersen, 1993). These models reflect a mathe-

matical approach based on the two-stage model of

cancer development proposed by Moolgavkar et
al. (1988). The model used to assess human risk to
environmental pollutants, states that any agent
that alters cell division or rates of cell differentiation will cause an increase in the number of
susceptible cells and thus, an increase in a cancerous outcome. To date, only cell proliferation and
DPX data have driven FA cancer risk assessments.
The endpoints needed to complete the model,
namely cell death rates associated with FA exposure, have yet to be reported. The assessment of
cell death in the epithelia lining the nasal turbinates poses a unique challenge because the lining is
only one to two cells thick. If a respiratory cell
underwent apoptosis it would immediately slought
off and, therefore, go undetected. Conventional
methods to quantitate cell death would not be
accurate. Since measurements of death rates of the
nasal epithelium following FA exposure are not
feasible, we chose to evaluate gene expression
profiles to gauge whether acute changes in activity
of genes that control apoptosis were induced by
FA.
The evaluation of respiratory tract toxicity
following exposure to chemicals involves delivering the agent to animals by inhalation. This is a
natural route of entry in the exposed animal and is
a preferred method to introduce toxicants to the
respiratory tract. However, this method cannot
always be used because of cost and the need of
specialized equipment. Alternatively, direct instillation of the test substance to the respiratory tract
has been employed in many studies (Evans and
Hastings, 1992; Jeffrey et al., 2002; Wagner et al.,
2001). The advantages of direct instillation are that
the chemical is delivered to the target site (respiratory cells lining the nasal cavity) and it is a
relatively simple procedure to carry out without
requiring special equipment. An acute dose of a
chemical in aqueous solution can be infused with
relative ease and little stress to the animal.
Although there are distinct differences in distribution, clearance, and retention of chemicals when
administered by nasal instillation versus inhalation, nasal instillation can be a useful and relatively inexpensive method for investigating specific
questions involving respiratory toxicants.

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

The concentration of FA used in our study

induces regenerative cell proliferation, but is not
high enough to produce necrosis. St. Clair et al.,
(1990), using nasal instillation of 400 mM FA
showed increased cell proliferation of respiratory
epithelium with minimal cytotoxicity. Since the
present study represents the first look at global
gene expression changes following FA exposure, a
simple design of two groups of animals (control
and treated) was utilized. For review of experimental design decisions and interpretation of gene
expression data from toxicogenomic experiments
see (Hamadeh et al., 2002). In the present study, a
single concentration and time point were used to
examine transcriptional changes associated with
acute FA exposure.

2. Materials and methods

15

phase separation was carried out after thawing the

cellular/TrizolTM samples in a warm water bath
(65
/70 8C) for 10 min. Chloroform was added to
each sample and tubes were shaken followed by
10 000 /g centrifugation at 8 8C for 20 min. The
phenol/chloroform phase separates to the bottom
and the upper aqueous phase contains the RNA,
which was precipitated using ice cold isopropyl
alcohol followed by centrifugation at 10 000/g
for 20 min. The isopropyl supernatant was removed leaving the RNA pellet. The RNA was
resuspended with 75% alcohol, mixed by inversion,
then centrifuged at 10 000/g for 30 min. RNA
was dissolved in 100 ml of DEPC water and heated
at 70 8C for 3 min with intermittent vortexing.
The RNA solution was then transferred to siliconized Eppendorf tubes and the concentration
(absorbance at 260 nm; 1A260 unit of single
stranded RNA /40 mg/ml) was adjusted to 2 mg/
ml with diethylene pyrocarbonate (DEPC) water.

2.1. Animal dosing

Rats were maintained in plastic cages (two per
cage) with filter purified tap water and feed ad
libitum. On day of treatment, 40 ml aliquots of
water or FA (400 mM) were instilled into each
nostril using a pipette.
2.2. Epithelial cell extraction from anterior rat
noses
Eight 60-day-old F344 male rats (Charles River
Laboratory, Raleigh, NC) were euthanized by
receiving CO2 asphyxiation and nasal respiratory
cells were recovered and processed as previously
described (Hester et al., 2002). Briefly, TrizolTM
reagent was instilled into each nostril through
polyethylene tubing and incubated for 10 min.
Using a syringe attached to the tubing, the cellular
TrizolTM solution was removed, placed in a microfuge tube, and immediately frozen in liquid nitrogen for subsequent total RNA isolation and
quantitation.
2.3. Total RNA extraction
Total RNA was extracted from respiratory cells
as previously reported (Hester et al., 2002). Briefly,

2.4. Electrophoresis of the RNA

5 mg of the resultant RNA sample was loaded
into a denaturing 1% agarose gel containing FA
according to the protocol described in Lehrach et
al. (1977) for electrophoresis. Observation of
rRNA subunit bands at 18S and 28S indicate the
presence of intact RNA.

3. Synthesis and labeling of cDNA

About 5 mg total RNA was used to generate a
probe for hybridizations as described in Hester et
al. (2002). Briefly, seven (four control and three
treated) cDNA probes were generated by reverse
transcribing the RNA with primers in the presence
of 33P. Unincorporated label was removed on a
MicrospinTM Sepharose G-50 gel filtration column
(Amersham Pharmacia Biotech, Piscataway, NJ).
For hybridization of the labeled product, 250 000
Cerenkov cpm were used. ClontechTM software was
used to analyze gene expression levels from the
hybridization membranes.

16

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

4. Real-time quantitative PCR

Quantification of mRNA was done using Molecular Probes Ribogreen RNA Quantification
kit. To prevent amplification of genomic DNA
sequences, all RNA samples were treated with
DNase I and diluted to 10 ng/ul. Taqman probes
were labeled with FAM (carboxyfluorescein) as
the reporter dye on the 5? position and TAMRA
(carboxytetramethylrhodamine) as the quencher
dye on the 3? position. cDNA was made using
moloney murine leukemia virus (MMuLV) reverse
transcriptase. All reactions were carried out in a
single tube reaction setup on an ABI PRISM 7700
sequence detection system (Applied Biosystems,
Inc). The following temperature profile was used:
30 min at 48 8C for reverse transcription and 10
min at 95 8C for reverse transcription inactivation
and AmpliTaq Gold activation, 40 cycles of 15 s at
94 8C and 1 min at 60 8C. To check for possible
contamination in the reaction mix, no template
control (NTC) wells without RNA were used. The
cycle threshold Ct (i.e. ten times the standard
deviation of the mean baseline emission calculated
during PCR cycles 3 to 15) was used to calculate
relative amounts of target RNA. The delta Ct
method was used to calculate relative fold expression levels, as described by Applied Biosystems.
Primers and probes for ten rat genes (aldr, b2mg,
calp, cof, czsod, p38mapk, gst, nmor, sc1, waf1)
were purchased from Keystone Biosource (Camarillo, CA). Final concentration of all primers were
900 nM and probes were at 200 nM.
After the reaction was complete, a graph of
fluorescent intensity versus cycle number was
created using the above software. For each gene
assayed, a Ct is placed at the intersection of the
linear portion of the curve, which reflects the
exponential amplification. Samples, which do not
reach the threshold line, are considered not above
background. The average Ct for each gene is
calculated by subtracting the Ct of the sample
RNA from the control RNA for the same time
measurement. This value is known as the delta Ct
and reflects the relative expression of the treated
sample compared with control and becomes the
exponent in the calculation for amplification
240ct, equivalent to fold change in expression.

When the delta Ct is negative, the final calculation

is negative, and is interpreted as down regulation
of that gene.

5. Statistical analysis of microarray data

The mean and variance of the Adjusted signal
intensity (ASI) were calculated and plotted (Fig. 1)
for each gene across the control and FA-treated
samples. There is a clear trend that the variation in
expression of a gene increased along with the
intensity of the gene. Therefore, intensities were
log (base 2) transformed to stabilize the variation
across intensity levels. Next, the data was globally
normalized within each membrane by first subtracting the median intensity and then scaling by
the median absolute difference (MAD); normalizing by the median and MAD for each membrane
is more robust against extreme values than by the
classical measures, mean and standard deviation.
Once the data were normalized, statistically significant, differentially expressed genes were obtained using the statistical software, SAM
(significance analysis of microarrays) (Tusher et
al., 2001). Because of the large number of genes
incorporated into microarray experiments, adjusting for multiple testing is necessary when assessing
statistical significance. Few experiments have enough power to detect significance with a stringent
Bonferroni correction when the number of comparisons is large. In SAM, the effect of multiple
comparisons is controlled through the false discovery rate (FDR; Benjamini and Hochberg,
1995), which is defined as the expected proportion
of false positives among the rejected hypotheses;
for example, an FDR of 5% in testing differential
gene expression would indicate that about 5% of
the identified genes were spurious. Through a
permutation scheme, SAM estimates the FDR
for a set of genes whose test statistics deviate
from their expected value by more than a given
threshold level; by adjusting the threshold level
one can achieve a desired FDR. Due to the
limitation in sample size of our data, threshold
levels were chosen that resulted in FDRs of 11.3
and 5.3% as the nearest approximations of 10 and
5%, respectively.

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

17

Fig. 1. Distribution of the raw intensities for each gene. The variation in intensities across treatment (gray triangles) and control
samples (black circles) is plotted against the mean intensity for all 1185 genes. It is evident, that the variation observed within each gene
was dependent on the average intensity. Despite the smaller sample size, the variability observed in the treatment group (n/3) was
smaller on average, than in the control group (n/4).

6. Results
6.1. SAM analysis of gene expression
The ASI were obtained from the image analysis
for 1185 genes from four control and three FAtreated samples. The data were first transformed
and normalized as described in Section 5, SAM
was then used to identify a set of genes that were
differentially expressed across treated and control
samples. By setting a threshold level of 0.187,
SAM predicted a FDR of 11.3% for 46 significantly altered genes; 24 genes increased in expression due to FA treatment and 22 genes decreased
in expression (Fig. 2). Among the genes listed in
Tables 1 and 2, the fold change in raw intensities
ranged from 6.1 to 2.4 (Table 1) among upregulated genes and 10
/2.5 in the down-regulated
genes (Table 2). By setting a more stringent
threshold level of 0.232 we obtain a set of 14
significant genes with FDR 5.3%. This set of genes
corresponded to the top 13 up-regulated genes in
Table 1 and the most down-regulated gene in
Table 2.

Genes listed in Table 1 were statistically evaluated and sorted from the highest test statistic
value to the lowest. Ten of the most highly
expressed genes for FA-treated animals included
NMDA receptor, inducible nitric oxide synthase,
macrophage inflammatory protein 1 alpha and 2,
Wilms tumor protein, tumor necrosis factor
ligand, methyl-CpG binding protein 2, GABA
receptor, Fos-responsive gene 1, and presomatotrophin. Of the 24 significantly up-regulated genes
six were receptors, six were involved in extracellular cell signaling, and four were oncogene/tumor
suppressor genes suggesting pathways that could
be affected by FA treatment. In contrast, in Table
2, of the 22 genes significantly down-regulated,
five were involved in ion channel regulation and
four were involved in protein turnover which
suggests that an early response to FA treatment
may be impaired ion channel function and interruption of protein processing. Many phase I and
phase II genes regulating xenobiotic metabolism
and oxidative stress, the family of cytochrome
P450 and glutathione, respectively, were altered in
treated versus control.

18

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

Fig. 2. Identification of differentially expressed genes. Using SAM, the observed test statistic is plotted against the statistic expected by
chance for 1185 genes. Genes that were expressed higher in the FA-treated samples appear on the positive side of the x -axis. At the
threshold, D/0.187 (drawn as dashed lines) SAM predicts 46 genes (24 positive and 22 negative) as being differentially expressed. The
FDR was estimated as 11.3%.

7. Comparison of atlas rat toxicology II array and

real-time PCR data
A subset of ten genes with a range of expression
levels from high to low were selected for quantitative real-time PCR analysis to validate the
direction of gene expression observed on the
microarrays. These ten genes included aldehyde
dehydrogenase (aldr), calpactin (calp), cofillin
(cof), copper
/zinc superoxide dismutase (czsod),
p38 map kinase (p38mapk), gluthione (gst),
NAD(p)H quinone oxidoreductase (nmor), sodium chloride 1 (sc1), and p21 waf1/CIP 1
(cyclin-dependent protein kinase inhibitor)
(waf1). Fig. 3 displays gene expression changes
present in both the Atlas Toxicology II arrays and
TaqManTM assays. Ten genes were confirmed but
arrays under-estimated, on average, the true
expression differences as revealed by quantitative
RT-PCR. Eight of the ten TaqManTM confirmed

genes on the array were estimated as increased

signal and one gene, copper
/zinc superoxide
dismutase (CZSOD), showed a reduced signal.

7.1. Apoptosis gene expression

Fig. 4 shows gene expression levels of nine
apoptosis genes assayed in the FA and control
groups. None of these genes were defined as
statistically
different.
However,
there
was a general trend of less apoptotic gene expression levels in the FA group compared with control.
The nine genes were representative of three of the
major apoptosis-regulating pathways including
receptor-mediated (Fas-L), caspase (Caspase 3),
and the mitochondrial-associated bcl2 family of
genes (bcl2-x, bcl2-associated oncogene, BCLX,
BAD).

Table 1
24 significantly up-regulated genes
Treatmenta Controla Fold
change

Test-statistic

Family

D11c

0.658

0.580

6.086

1.794

Receptors

A08f
E02k
D14k
A10f
E02m

0.116
0.251
0.154
0.331
0.184

0.759
0.906
0.673
1.200
1.009

3.856
5.763
3.536
3.785
3.804

1.415
1.404
1.281
1.157
1.144

A06l
D10d
D01l
E03j
D07m
A09e

0.390
1.115
0.153
0.697
0.534
0.463

0.368
1.937
0.790
1.393
1.201
1.144

2.950
4.938
2.641
3.207
2.904
2.900

1.117
1.083
0.991
0.988
0.987
0.986

D14f
D12i
E02n
A11a
E03d
F12h

0.626
0.674
1.025
0.852
0.616
0.076

1.270
0.060
1.631
0.195
1.345
0.654

2.866
2.371
3.513
2.456
3.680
2.404

0.985
0.914
0.897
0.887
0.886
0.880

A11h

0.882

0.307

2.236

0.878

A11g
D05m
B01a
C05b
A05b

0.439
0.762
0.331
0.005
0.651

1.110
1.376
0.361
0.591
1.420

2.590
2.633
2.364
2.454
2.599

0.869
0.868
0.853
0.852
0.839

Description

N -methyl-D-aspartate receptor subtype 2B (NMDAR2B; NR2B); glutamate receptor subunit

epsilon 2B (GRIN2B)
Immune
Inducible nitric oxide synthase (INOS); type II NOS (NOS2)
Extrac/sign Macrophage inflammatory protein 1 alpha (Small inducible cytokine A3)
Extrac/sign Macrophage inflammatory protein 2 (MIP2)
Onco/tumo Wilms tumor protein homolog 1 (WT1)
Extrac/sign Tumor necrosis factor (ligand) superfamily, member 6 (apoptosis (APO-1) antigen ligand 1) (Fas
antigen ligand)
Tx fact/dBP Methyl-CpG-binding protein 2 (MECP2)
Receptors
Gamma-aminobutyric-acid receptor alpha 2 subunit (GABA(A) receptor alpha 2; GABRA2)
Receptors
Fos-responsive gene 1
Extrac/sign Presomatotropin
Receptors
Somatostatin receptor 3 (SSTR3; SS3R)
Immune
34A transformation-associated protein; TAP-related matrix metalloproteinase 10 (MMP10);
stromelysin 2 (SL2); transin 2
Extrac/sign Nerve growth factor 8A (VGF8A)
Receptors
Metabotropic glutamate receptor 6 (GRM6; MGLUR6)
Extrac/sign Brain natriuretic peptide (BNP); 5-kDa cardiac natriuretic peptide (NPPB); ISO-ANP
Onco/tumo Fos-like antigen 1
Extrac/sign Follicle-stimulating hormone beta subunit
Recept by
Adenosine A2A receptor (ADORA2A)
act
Onco/tumo Neogenin; DCC netrin receptor-related protein; immunoglobulin gene superfamily member;
former tumor suppressor protein candidate
Onco/tumo Deleted in colcorectal cancer (rat homolog)
Receptors
5-hydroxytryptamine (serotonin) receptor 5A
Stress res
Cytochrome P450 1b1
Metab pat
Acetyl-CoA carboxylase (ACAC; ACC); biotin carboxylase
Cell sur ag NK lymphocyte receptor; NKR-P1B

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

Gene
ID

Sample means are calculated from the normalized data. Fold Change is calculated from the raw data.
a
Mean signal intensities.

19

20

Table 2
Tenty-two significantly down-regulated genes
Treatmenta Controla Fold
change

Test-statistic

Family

Description

E09m

/2.352

/1.356

9.701

/1.439

Phosphorylase B kinase gamma subunit

F07j
B04i
F06l
E14n

/2.284
/1.910
/0.568
/2.357

/1.318
/1.004
0.208
/1.499

7.999
4.272
3.312
4.625

/1.159
/1.124
/1.094
/1.060

C07g
B14i

/0.238
/1.969

0.369
/1.206

2.489
3.533

/1.026
/1.012

Mod/
transd
Prot turnov
Ion chann
Prot turnov
Mod/
transd
Metab pat
Traff/target

A07k
C05d
B12i
B13h
B10m
F07i
B01i
B13k
E14l

/1.286
/0.768
/2.448
/1.514
/1.566
/1.658
/0.087
/0.994
0.212

/0.503
/0.009
/1.692
/0.787
/0.986
/0.824
0.632
/0.399
0.819

3.385
2.846
12.641
2.926
2.488
2.900
3.000
2.531
2.656

/0.951
/0.940
/0.936
/0.934
/0.928
/0.927
/0.918
/0.917
/0.903

F09j
C11k
B02d
B10n

/1.498
/2.197
/1.803
/1.537

/0.809
/1.352
/1.065
/0.829

3.393
2.802
2.853
3.079

/0.884
/0.872
/0.871
/0.869

B02j

/1.621

/0.972

2.617

/0.866

C08c

/1.596

/0.973

2.855

/0.854

Proprotein convertase subtilisin/kexin type 2

Sodium channel protein 6 (SCP6)
Secretory granule neuroendocrine, protein 1 (7B2 protein)
Ras-related protein RAB26

Cytochrome P450 IVA8 (CYP4A8); P450-KP1; P450-PP1

Fatty acid-binding protein 9 (FABP9); testis lipid-binding protein (TLBP); 15-kDa perforatorial
protein (PERF15)
Cell cycle
Cyclin-dependent kinase 4 inhibitor 2B (CDKN2B); p14-INK4B; p15-INK5B
Metab pat Brain long-chain fatty acid-CoA ligase (LACS); acyl-CoA synthetase
Extra matr Myelin-associated glycoprotein
Traff/target Gastric intrinsic factor
Ion chann Aquaporin 3 (AQP3); 31.4-kDa water channel protein
Prot turnov Dipeptidyl peptidase IV (DPPIV; DPP4); bile canaliculus domain-specific membrane glycoprotein
Stress res
Plasma glutathione peroxidase (GSHPX-P; GPX3); selenoprotein
Traff/target Syntaxin binding protein 1
Mod/
Ras-related protein RAB16
transd
Prot turnov Interleukin 1beta converting enzyme
Translation Ribosomal protein S19
Ion chann Solute carrier family 2 A2 (gkucose transporter, type 2)
Ion chann Synaptic vesicle amine transporter (SVAT); monoamine transporter; vesicular amine transporter 2
(VAT2)
Ion chann Neuronal acetylcholine receptor protein alpha 6 subunit (NACHRA6; ACRA6); cholinergic
receptor nicotinic alpha polipeptide 6 (CHRNA6)
Metab pat Cytochrome P-450 2C23, arachidonic acid epoxygenase

Sample means are calculated from the normalized data. Fold change is calculated from the raw data. For down-regulated genes, fold change is represented as control
over treated.
a
Mean signal intensities.

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

Gene
ID

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

21

Fig. 3. Comparison of TaqManTM to ClontechTM array results. A subset of ten genes was verified using quantitative real time PCR.
Nine of the ten gene expression values were in good agreement. One gene value, calpactin, was under-estimated on the array compared
with the TaqManTM result.

Fig. 4. Apoptosis Genes. Mean expression levels of nine genes involved in regulating apoptosis. Seven of the nine genes on the FAtreated group showed reduced expression values compared with control group; AO1g-Annexin V, B13c-Annexin I, C12h-Bcl2associated death promoter (BAD), C12i-Bcl2-associated X protein (BAX), C12j-Bcl2 associated oncogene, BCL2-L, C12k-BCL2,
C12m activator of apoptosis (HRK), FO9i- Caspase 3, and EO2m-Fas-ligand.

22

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

8. Discussion
FA induces squamous metaplasia leading to
squamous cell carcinoma after 12 months or
longer of exposure. In the present study, we
evaluated the acute response of epithelial cells 24
h after FA treatment. Our results indicated that
the expression levels of genes in several functional
categories were altered, including those participating in xenobiotic metabolism, cell cycle regulation,
DNA synthesis and repair, oncogene, and apoptosis. Xenobiotic metabolism genes showed a
general trend towards increased expression. Unexpectedly, two cytochrome P450 family members
were down-regulated, cytochrome P450 IVA8, and
cytochrome P450 2C23, genes which are usually
involved in metabolizing FA (Dahl and Hadley,
1991).
Genes were considered differentially expressed if
they exceeded the calculated threshold of 0.187,
with up-regulated genes falling on the positive side
and down regulated genes appearing in the negative quadrant (Fig. 2). However, gene expression
changes, which were not statistically significant
cannot be considered to be irrelevant but rather
not included in this analysis. For example, genes
expressed at low levels in the FA-treated group
that exhibit a small change in expression could still
be responsible for a large biologic effect, especially
if their gene products control critical pathways
such as cell growth or cell death.
Seven of the nine genes regulating apoptosis
showed less expression in the FA group compared
with the control group, however, none of these
genes were identified as being statistically different. This result may be interpreted as a negative
finding, however, the effect of FA exposure on
respiratory epithelium is unknown. Therefore, any
information of changes in gene expression levels
after FA treatment is both novel and informative.
The post-treatment changes in gene expression
levels may well be subtle ones rather than large
changes. In addition, the respiratory cell may be
responding to acute FA exposure by altering
metabolism preferentially to deal with the acute
toxicity of this chemical. Just as changes in
reparative cell proliferation secondary to FA
treatment require time to develop, changes in

apoptosis gene expression may require longer

exposures over time to saturate the detoxifying
capacity of the nasal cell. So perhaps the optimum
time-frame to observe apoptosis gene expression
changes would certainly be longer than 24 h as it
takes 4 days to 6 weeks of continuous exposure for
increased cell proliferation to occur (Monticello et
al., 1991a).
FA is a potent respiratory irritant and is capable
of stimulating a number of receptors related to the
trigeminal nerve and localized in the nasal epithelium (Babiuk et al., 1985; Cassee et al., 1996). In
the present study, seven of the 24 significantly upregulated genes (Table 1) were classified as receptors. Stimulation of these receptors represent one
of many defense mechanisms of the rodent respiratory tract (Babiuk et al., 1985). Other protective mechanisms include a decrease in minute
ventilation, local cellular metabolism the mucociliary apparatus, and DNA repair (Swenberg et
al., 1983). Chemicals that can induce these protective mechanisms are considered to be sensory
irritants because they can stimulate the trigeminal
nerve fibers, cause a burning sensation, and
diminish respiratory rates (Alarie, 1973). Interestingly, the gene expression data in this report show
that four of the seven most highly expressed genes
were neuropeptides including NMDAR2B, serotonin 5A, GABA(A), and metabotrophic glutamate receptor which could reflect a response to
sensory irritation caused by FA.
In summary, there were more up-regulated
genes in the FA treated group than down-regulated genes. Some genes, such as cell receptors, are
reported for the first time to have changed
expression in FA-exposed respiratory cells lining
the nasal turbinates, which is consistent with the
phenotypic response of these cells. Several genomic techniques, including subtractive hybridization,
array
based
comparative
genomic
hybridization (aCGH), mRNA differential display
and serial analysis of gene expression (SAGE) are
being utilized to identify gene expression profiles
which could account for differentially expressed
genes after chemical or xenobiotic treatment.
From the results reported here, it is clear that
multiple genetic changes are induced after FA
exposure. Genetic events that account for progres-

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

sion from squamous metaplasia to squamous cell

carcinoma are yet to be identified. Genes playing
more significant roles in common or overlapping
pathways are candidates to be assessed and
analyzed further over time. Results of the above
experiments hold the promise of revealing critical
gene interactions, which may aid in the discovery
of potential molecular events operative in the
pathogenic transition of epithelium exposed to
FA. Once gene pathways are identified, the next
key step will be to conduct bioassays that could
validate gene expression changes and will reinforce
insights provided by the gene profiling experiments.

Acknowledgements
This manuscript has been reviewed and approved for publication by the Environmental
Protection Agency and does not necessarily reflect
the views of the agency. Mention of trade names or
commercial products does not constitute endorsements or recommendations for use. Special thanks
to Dr. Jeffrey Ross, Dr. Julian Preston, Dr. David
Threadgill, Dr. Donald Delkar, and Dr. Marila
Cordeiro-Stone for review and advice of this
manuscript.

References
Alarie, Y., 1973. Sensory irritation by airborne chemicals. CRC
Crit. Rev. Toxicol. 2, 299
/363.
Andersen, M.E., Krishnan, K., Conolly, R.B., McClellan,
R.O., 1992. Biologically based modeling in toxicology
research. Arch. Toxicol. Suppl. 15, 217
/227.
Babiuk, C., Steinhagen, W.H., Barrow, C.S., 1985. Sensory
irritation response to inhaled aldehydes after formaldehyde
pretreatment. Toxicol. Appl. Pharmacol. 79, 143
/149.
Benjamini, Y., Hochberg, Y., 1995. Controlling the false
discovery rate: a practical and powerful approach to multiple testing. J. Royal Stat. Soc. Ser. B 57, 289
/300.
Cassee, F.R., Arts, J.H., Groten, J.P., Feron, V.J., 1996.
Sensory irritation to mixtures of formaldehyde, acrolein,
and acetaldehyde in rats. Arch. Toxicol. 70, 329
/337.
Chaw, Y.F., Crane, L.E., Lange, P., Shapiro, R., 1980.
Isolation and identification of cross-links from formaldehyde-treated nucleic acids. Biochemistry 19, 5525
/5531.
CIIT, 1999. Formaldehyde: hazard characterization and dose
/
response for carcingenecity by route of inhalation. Chemical

23

Industry Institute of Toxicology, Review edition, Research

Trianagle Park, NC.
Conolly, R.B., Andersen, M.E., 1993. An approach to mechanism-based cancer risk assessment for formaldehyde. Environ. Health Perspect. 6, 169
/176.
Dahl, A.R., Hadley, W.M., 1991. Nasal cavity enzymes
involved in xenobiotic metabolism: effects on the toxicity
of inhalants. Crit. Rev. Toxicol. 21, 345
/372.
Evans, J., Hastings, L., 1992. Accumulation of Cd(II) in the
CNS depending on the route of administration: intraperitoneal, intratracheal or intranasal. Fundam. Appl. Toxicol.
19, 275
/278.
Flyvholm, M.A., Andersen, P., 1993. Identification of formaldehyde releasers and occurrence of formaldehyde and
formaldehyde releasers in registered chemical products.
Am. J. Ind. Med. 24, 533
/552.
Hamadeh, H.K., Amin, R.P., Paules, R.S., Afshari, C.A., 2002.
An overview of toxicogenomics. Curr. Issues Mol. Biol. 4,
45
/56.
Heck, H., Casanova, M., 1999. Pharmacodynamics of formaldehyde: applications of a model for the arrest of DNA
replication by DNA-protein cross-links. Toxicol. Appl.
Pharmacol. 160, 86
/100.
Heck, H.D., Casanova, M., Starr, T.B., 1990. Formaldehyde
toxicity */new understanding. Crit. Rev. Toxicol. 20, 397
/
426.
Hester, S.D., Benavides, G.B., Sartor, M., Yoon, L., Wolf,
D.C., Morgan, K.T., 2002. Normal gene expression in male
F344 rat nasal transitional and respiratory epithelium. Gene
285, 301
/310.
Jeffrey, A.M., Luo, F.Q., Amin, S., Krzeminski, J., Zech, K.,
Williams, G.M., 2002. Lack of DNA binding in the rat nasal
mucosa and other tissues of the nasal toxicants roflumilast,
a phosphodiesterase 4 inhibitor, and a metabolite, 4-amino3,5-dichloropyridine, in contrast to the nasal carcinogen 2,6dimethylaniline. Drug Chem. Toxicol. 25, 93
/107.
Kennedy, G., Slaich, P.K., Golding, B.T., Watson, W.P., 1996.
Structure and mechanism of formation of a new adduct
from formaldehyde and guanosine. Chem. Biol. Interact.
102, 93
/100.
Lehrach, H., Diamond, D., Wozney, J.M., Boedtker, H., 1977.
RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamination.
Biochemistry 16, 4743
/4751.
Monticello, T.M., Miller, F.J., Morgan, K.T., 1991a. Regional
increases in rat nasal epithelial cell proliferation following
acute and subchronic inhalation of formaldehyde. Toxicol.
Appl. Pharmacol. 111, 409
/421.
Monticello, T.M., Renne, R., Morgan, K.T., 1991b. Chemically
induced cell proliferation in upper respiratory tract carcinogenesis. Prog. Clin. Biol. Res. 369, 323
/335.
Moolgavkar, S.H., Dewanji, A., Venzon, D.J., 1988. A
stochastic two-stage model for cancer risk assessment. I.
The hazard function and the probability of tumor. Risk
Anal. 8, 383
/392.
Morgan, K.T., Jiang, X.Z., Starr, T.B., Kerns, W.D., 1986.
More precise localization of nasal tumors associated with

24

S.D. Hester et al. / Toxicology 187 (2003) 13
/24

chronic exposure of F-344 rats to formaldehyde gas.

Toxicol. Appl. Pharmacol. 82, 264
/271.
Quievryn, G., Zhitkovich, A., 2000. Loss of DNA-protein
crosslinks from formaldehyde-exposed cells occurs through
spontaneous hydrolysis and an active repair process linked
to proteosome function. Carcinogenesis 21, 1573
/1580.
St. Clair, M.B., Gross, E.A., Morgan, K.T., 1990. Pathology
and cell proliferation induced by intra-nasal instillation of
aldehydes in the rat: comparison of glutaraldehyde and
formaldehyde. Toxicol. Pathol. 18, 353
/361.
Swenberg, J.A., Barrow, C.S., Boreiko, C.J., Heck, H.D.,
Levine, R.J., Morgan, K.T., Starr, T.B., 1983. Non-linear
biological responses to formaldehyde and their implications
for carcinogenic risk assessment. Carcinogenesis 4, 945
/
952.
Szende, B., Tyihak, E., Szokan, G., Katay, G., 1995. Possible
role of formaldehyde in the apoptotic and mitotic effect of
1-methyl-ascorbigen. Pathol. Oncol. Res. 1, 38
/42.

Szende, B., Tyihak, E., Trezl, L., Szoke, E., Laszlo, I., Katay,
G., Kiraly-Veghely, Z., 1998. Formaldehyde generators and
capturers as influencing factors of mitotic and apoptotic
processes. Acta Biol. Hung. 49, 323
/329.
Takahashi, H., Hashimoto, Y., 2001. Formaldehyde-mediated
modification of natural deoxyguanosine with amines: onepot cyclization as a molecular model for genotoxicity.
Bioorg. Med. Chem. Lett. 11, 729
/731.
Tusher, V.G., Tibshirani, R., Chu, G., 2001. Significance
analysis of microarrays applied to the ionizing radiation
response. Proc. Natl. Acad. Sci. USA. 98, 5116
/5121.
Tyihak, E., Bocsi, J., Timar, F., Racz, G., Szende, B., 2001.
Formaldehyde promotes and inhibits the proliferation of
cultured tumor and endothelial cells. Cell Prolif. 34, 135
/
141.
Wagner, J.G., Hotchkiss, J.A., Harkema, J.R., 2001. Effects of
ozone and endotoxin coexposure on rat airway epithelium:
potentiation of toxicant-induced alterations. Environ.
Health Perspect. 4, 591
/598.