Environmentally Relevant Levels of Depleted Uranium Impacts Dermal Fibroblast Proliferation, Viability, Metabolic Activity, and Scratch Closure

Abstract

Uranium (U) is a heavy metal used in military and industrial settings, with a large portion being mined from the Southwest region of the United States. Uranium has uses in energy and military weaponry, but the mining process has released U into soil and surface waters that may pose threats to human and environmental health. The majority of literature regarding U's human health concern focuses on outcomes based on unintentional ingestion or inhalation, and limited data are available about its influence via cutaneous contact. Utilizing skin dermis cells, we evaluated U's topical chemotoxicity. Employing soluble depleted uranium (DU) in the form of uranyl nitrate (UN), we hypothesized that in vitro exposure of UN will have cytotoxic effects on primary dermal fibroblasts by affecting cell viability and metabolic activity and, further, may delay wound healing aspects via altering cell proliferation and migration. Using environmentally relevant levels of U found in water (0.1 μM to 100 μM [UN]; 23.8-23,800 ppb [U]), we quantified cellular mitosis and migration through growth curves and in vitro scratch assays. Cells were exposed from 24 h to 144 h for a time-course evaluation of UN chemical toxicity. The effects of UN were observed at concentrations above and below the Environmental Protection Agency threshold for safe exposure limits. UN exposure resulted in a dose-dependent decrease in the viable cell count; however, it produced an increase in metabolism when corrected for the viable cells present. Furthermore, cellular proliferation, population doubling, and percent closure was hindered at levels ≥10 μM UN. Therefore, inadvertent exposure may exacerbate pre-existing skin diseases in at-risk demographics, and additionally, it may substantially interfere in cutaneous tissue repair processes.

Keywords: Navajo Nation; cellular respiration; cytotoxicity; dermis cell; growth curve; scratch assay; skin; statistical modeling; uranyl nitrate; wound healing.

Figure 1

Metabolic activity results of UN exposured hDFn. UN decreases PrestoBlue$${}^{textTM}$$ readings in cultured human dermal cells at levels below and above the EPA MCL. Bar charts are faceted by exposure time representing the mean metabolic activity percentage ±StdErr of each treatment group (n = 6 done in sextuplicate). All UN concentrations are in $$\mathsf{\mu }$$M with -0- representing control group. Statistically significant differences between groups are illustrated by compact letter display with different letters at the top representing a balanced Tukey pairwise t-tests where p-value ≤ 0.05. Group means followed by a common letter were not different at 5% level of significance. Vertical red line illustrates the EPA MCL threshold for uranium (0.126 $$\mathsf{\mu }$$M).

Figure 2

Viable cell count of UN exposed hDFn. Bar charts are faceted by exposure time representing the mean viable cell count percentage ±StdErr of each treatment group (n = 6 done in sextuplicate). All UN concentrations are in $$\mathsf{\mu }$$M with -0- representing control group. At all days UN levels below EPA MCL caused a significant decrease in viable cell count. Statistically significant differences between groups are illustrated by compact letter display with different letters at the top representing a balanced Tukey pairwise t-tests where p-value ≤ 0.05. Group means followed by a common letter were not different at 5% level of significance. Vertical red line illustrates the EPA MCL threshold for uranium (0.126 $$\mathsf{\mu }$$M).

Figure 3

Metabolic ratio of UN exposed hDFn. Bar charts represents the mean relative metabolic ratio percentage ±StdErr of each treatment group (n = 6 done in sextuplicate for each condition). All UN concentrations are in $$\mathsf{\mu }$$M with -0- representing control group. Statistically significant differences between groups are illustrated by compact letter display with different letters at the top representing a balanced Tukey pairwise t-tests where p-value ≤ 0.05. Group means followed by a common letter were not different at 5% level of significance. Vertical red line illustrates the EPA MCL threshold for uranium (0.126 $$\mathsf{\mu }$$M).

Figure 4

(A) Fibroblast cell population reduced when exposed to increasing levels of UN over 144 hour growth curve (n = 6 for each level of UN). Cell counts were identically and independently collected. Contrasts were performed by a Tukey pairwise t-test across significant variables from an interaction two-way ANOVA. Significance bars, marked with “*”, refer to significant Tukey pairwise t-tests between the conditions at the beginning and the end of the bar. Days where no significant differences were observed are labeled with N.S. (B) Modeled regression lines of ln(cell estimate) growth over 144 h (R$${}_{adj}^2$$ = 0.8788). Colors of least squares (LS) lines differentiate UN levels. LS best fit lines represent predicted Ln(cell estimate) based on UN concentration and Ln(hour). Points represent mean cell estimate for each group ±StrErr. Red vertical line represents the largest percent cell estimate difference observed between 100 $$\mathsf{\mu }$$M and control. Images were taken using light microscopy and used to illustrate differences in cell population after 72 h of exposure (100× total magnification).

Figure 5

(A) Cell population doublings over a 144 h growth curve (n = 6 for each group). Contrasts were performed by a Tukey pairwise t-test across significant variables from an interaction two-way ANOVA. Significance bars, marked with “*”, refer to significant Tukey pairwise t-tests between the conditions at the beginning and the end of the bar. Time points where no significant differences were observed are labeled with N.S. (B) Polynomial regression modeling of hDFn population doublings (R$${}_{adj}^2$$ = 0.9314). Points represent mean PD ±StrErr. Red vertical lines indicates largest PD difference observed between control and 100 $$\mathsf{\mu }$$M. Colors of least squares (LS) lines differentiate UN levels.

Figure 6

(A) Percent closure differences over a 24 h scratch assay (n = 8 for each condition). Contrasts were performed by a Tukey pairwise t-test across significant variables from an additive two-way ANOVA. Significance bars, marked with “*”, refer to significant Tukey pairwise t-tests between the conditions at the beginning and the end of the bar. (B) Interaction modeling of percent closure across treatments (R$${}_{adj}^2$$ = 0.8663). Displayed points represent the mean of each group ±StdErr. Red vertical line represents largest difference observed 4 h post-scratch between control and 10 $$\mathsf{\mu }$$M. Inverted light microscopy of hDFn migration at 0, 12, and 24 h post scratch (100× total magnification). All initial scratch widths were 1 mm ± 0.1 mm, with no significant differences of beginning scratch width at hour 0.

Figure 6

(A) Percent closure differences over a 24 h scratch assay (n = 8 for each condition). Contrasts were performed by a Tukey pairwise t-test across significant variables from an additive two-way ANOVA. Significance bars, marked with “*”, refer to significant Tukey pairwise t-tests between the conditions at the beginning and the end of the bar. (B) Interaction modeling of percent closure across treatments (R$${}_{adj}^2$$ = 0.8663). Displayed points represent the mean of each group ±StdErr. Red vertical line represents largest difference observed 4 h post-scratch between control and 10 $$\mathsf{\mu }$$M. Inverted light microscopy of hDFn migration at 0, 12, and 24 h post scratch (100× total magnification). All initial scratch widths were 1 mm ± 0.1 mm, with no significant differences of beginning scratch width at hour 0.