Relative contributions of endogenous and exogenous formaldehyde to formation of deoxyguanosine monoadducts and DNA-protein crosslink adducts of DNA in rat nasal mucosa

Abstract

Understanding the dose-response for formaldehyde-induced nasal cancer in rats is complicated by (1) the uneven distribution of inhaled formaldehyde across the interior surface of the nasal cavity and, (2) the presence of endogenous formaldehyde (endoF) in the nasal mucosa. In this work, we used computational fluid dynamics (CFD) modeling to predict flux of inhaled (exogenous) formaldehyde (exogF) from air into tissue at the specific locations where DNA adducts were measured. Experimental work has identified DNA-protein crosslink (DPX) adducts due to exogF and deoxyguanosine (DG) adducts due to both exogF and endoF. These adducts can be considered biomarkers of exposure for effects of endoF and exogF on DNA that may be part of the mechanism of tumor formation. We describe a computational model linking CFD-predicted flux of formaldehyde from air into tissue, and the intracellular production of endoF, with the formation of DPX and DG adducts. We assumed that, like exogF, endoF can produce DPX. The model accurately reproduces exogDPX, exogDG, and endoDG data after inhalation from 0.7 to 15 ppm. The dose-dependent concentrations of exogDPX and exogDG are predicted to exceed the concentrations of their endogenous counterparts at about 2 and 6 ppm exogF, respectively. At all concentrations examined, the concentrations of endoDPX and exogDPX were predicted to be at least 10-fold higher than that of their DG counterparts. The modeled dose-dependent concentrations of these adducts are suitable to be used together with data on the dose-dependence of cell proliferation to conduct quantitative modeling of formaldehyde-induced rat nasal carcinogenicity.

Keywords: BBDR; CFD; DNA adducts; dosimetry; endogenous; formaldehyde; nasal.

Figure 1.

Computational mesh comparison. Views are of the right side of the nose, with the nostrils to the right, showing the tetrahedral mesh of the smoothed F344 rat model (top panels) used in the present work, and the hexahedral mesh of the original F344 rat model (bottom panels) used in the previous work (modified from Fig. 1 in Kimbell et al. (1997), used with permission). Insets show mesh close-ups for more detailed comparison.

Figure 2.

Lateral (top panels) and septal (bottom panels) views of the F344 rat nasal CFD model with the locations of the high tumor DPX site (left, magenta) and DG site (right, light blue) mapped onto the model. Epithelial types included in the surface model include dry squamous (dark grey), wet squamous (dark blue), respiratory/transitional (light grey), and olfactory (red cross-hatch). The DPX site was described by Casanova et al. (1994) and did not include any septal tissue; the DG site was defined based on descriptions by Lu et al. (2011) and Doyle-Eisele (2020) and consisted of nasal respiratory epithelium on the lateral and septal walls. The DG site encompassed a larger surface area that included regions outside of the main flow streams and therefore resulted in slightly lower average flux compared to the DPX site. Abbreviations: CFD, computational fluid dynamics; DG, deoxyguanosine; DPX, DNA protein cross-link.

Figure 3.

Formation of deoxyguanosine (DG) and DNA-protein crosslink (DPX) adducts in the rat nasal mucosa. Labile methyl groups and 1-carbon metabolism are responsible for the endogenous production of formaldehyde, whereas exogenous formaldehyde is absorbed from the nasal airway. Intracellular formaldehyde exists largely as its hydrate, formaldehyde acetal. In the nucleus, formaldehyde reacts with DNA to form DG and DPX adducts.

Figure 4.

Dose-response simulations for DG (lower curve) and DPX (upper curve) adducts. The DG data are from Lu et al. (2011) (Table 1). The DPX data are from Casanova et al. (1994) (Table 1). DG adducts were measured at 7 hr from the start of a single 6-h exposure. DPX were measured immediately after a 3-h exposure that followed 11 weeks and 4 days of exposure, 6 h/day/days per week. Abbreviations: DG, deoxyguanosine; DPX, DNA-protein crosslink.

Figure 5.

Simulation of the Yu et al. (2015) DG adduct data obtained during and after 28 days of exposure to 2.0 ppm formaldehyde. The lower, sawtooth curve is for exogenous DG adducts, whereas the upper curve is for endogenous DG adducts. Abbreviation: DG, deoxyguanosine.

Figure 6.

Simulated dose-response curves. Solid line and dashed line that crosses the solid line represent exogDPX and endoDPX, respectively. The concentration of exogDPX is predicted to cross that of endoDPX at about 2 ppm inhaled formaldehyde. Bottom 2 lines represent exogDG and endoDG. The concentration of exogDG is predicted to cross that for endoDG slightly below 6 ppm inhaled formaldehyde. Predicted concentrations of DPX adducts are over 10-fold higher than concentrations of DG adducts. Abbreviations: DG, deoxyguanosine; DPX, DNA-protein crosslink.

Figure 7.

Dose-response data for exogenous DG adducts (blue triangle) (Lu et al., 2011) and for exogenous DPX (orange circle) (Casanova et al., 1994). The lower, black dashed line represents the limit of detection (LOD) for DG adducts (5.26e−7 pmol/mm³) (Leng et al., 2019). Casanova et al. (1994) data were obtained from mucosal tissue in the high tumor region of the rat nose after 3 hr of exposure. The DPX datum at 0.32 ppm (black circle; Casanova et al., 1989) was obtained from the combined respiratory and olfactory mucosal tissue after 6 h of exposure; therefore, this 0.32 ppm datum would have been somewhat higher in value if it had been sampled in the high tumor region but lower in value if it had been sampled at 3 rather than 6 h. The curves show the predictions of an alternative model structure that includes a saturable extracellular clearance of formaldehyde (Campbell et al., 2020) that was able to reproduce the nonlinear low-dose behavior of DG adduct formation suggested by Leng et al. (2019). The black square shows the model prediction with zero order extracellular clearance (2.02E + 03 pmol/mm³/h) which had been set to produce exogenous DG adducts under the Leng et al. (2019) exposure scenario (28 consecutive day exposure for 6 h/day to 0.3 ppm). The dosimetry model was not able to reproduce the low-dose behavior reported for both DPX and DG adducts using a single set of model parameters. Abbreviations: DG, deoxyguanosine; DPX, DNA-protein crosslink.