Evolutionary and functional analyses of cytochrome P4501A promoter polymorphisms in natural populations

Abstract

The functional importance of variable, transcriptional regulatory sequences within and among natural populations is largely unexplored. We analysed the cytochrome P4501A (CYP1A) promoter in three populations of the minnow, Fundulus heteroclitus, because two SNPs in the promoter and first intron of CYP1A are under selection in populations adapted to pollutants. To define the importance of these SNPs, 1630 bp of the CYP1A promoter and first intron and exon were sequenced in eight individuals from three populations: a population from a polluted environment resistant to some aromatic pollutants and two flanking reference populations. CYP1A is induced by many aromatic pollutants, but in populations adapted to pollutants, CYP1A has been shown to be refractory to induction. We were interested in understanding whether variation in the CYP1A promoter explains mechanism(s) of adaptation to these aromatic pollutants. The CYP1A promoter was extremely variable (an average of 9.3% of the promoter nucleotides varied among all populations) and exhibited no fixed differences between populations. As CYP1A is poorly inducible in adapted fish, we hypothesized that CYP1A promoter regions might vary functionally between populations. Unexpectedly, in vitro analysis showed significantly greater transcription from CYP1A promoters found in the population from the polluted environment relative to promoters found in both reference populations. Thus, despite extensive variation among populations and lack of fixed differences between populations, individuals from a polluted environment have significantly enhanced promoter activity. These data demonstrate that intraspecific variation, which provides the raw material for natural selection to act on, can occur while maintaining promoter function.

Figure 1. F. heteroclitus collection sites

Superfund site (New Bedford Harbor, MA) is denoted by a star, and north and south clean, reference sites are denoted by circles.

Figure 2. Sliding window comparisons of average nucleotide substitutions per site within F. heteroclitus populations (Pi1 and Pi2) and between populations (Dxy)

Plots for the sliding window use a 50 bp-wide window and 10 bp step size.

Figure 3. Sliding window plots of Tajima's D, Fu and Li's D and Fu and Li's F

A. Tajima's D calculations, B. Fu and Li's D calculations, and C. Fu and Li's F calculations using a sliding window with a 50 bp-wide window and a 10 bp step size for each F. heteroclitus population. Solid and dashed lines represent the estimated values (D or F) and 99% confidence interval, respectively, obtained through 10,000 coalescent simulations. Green line: Sandwich population. Blue line: New Bedford Harbor population. Red line: Pt. Judith population. Asterisk (*) indicates the significant deviations of D.

Figure 4. Induction of luciferase activity from F. heteroclitus CYP1A reporter gene constructs from a New Bedford Harbor and Sandwich individual

CYP1A promoter strength was assayed in PLHC-1 cells treated with 3-MC. Asterisks (*) indicate a significant fold induction over control (p < 0.05).

Figure 5. Average fold induction of the CYP1A promoter over control for Sandwich, New Bedford Harbor and Point Judith populations

N = 4/population at a 1 μM concentration of 3-MC. A one-way ANOVA found a significant difference in the average fold induction between populations (p < 0.0001) and a Tukey's HSD post-hoc test determined that New Bedford Harbor was significantly different from Sandwich (p < 0.0001) and Point Judith (p < 0.0001), but Sandwich and Point Judith were not significantly different from each other (p = 0.258). Significant induction in the New Bedford Harbor population compared to the reference populations is denoted by an asterisk (*).