Allele interaction--single locus genetics meets regulatory biology

Abstract

Background: Since the dawn of genetics, additive and dominant gene action in diploids have been defined by comparison of heterozygote and homozygote phenotypes. However, these definitions provide little insight into the underlying intralocus allelic functional dependency and thus cannot serve directly as a mediator between genetics theory and regulatory biology, a link that is sorely needed.

Methodology/principal findings: We provide such a link by distinguishing between positive, negative and zero allele interaction at the genotype level. First, these distinctions disclose that a biallelic locus can display 18 qualitatively different allele interaction sign motifs (triplets of +, - and 0). Second, we show that for a single locus, Mendelian dominance is not related to heterozygote allele interaction alone, but is actually a function of the degrees of allele interaction in all the three genotypes. Third, we demonstrate how the allele interaction in each genotype is directly quantifiable in gene regulatory models, and that there is a unique, one-to-one correspondence between the sign of autoregulatory feedback loops and the sign of the allele interactions.

Conclusion/significance: The concept of allele interaction refines single locus genetics substantially, and it provides a direct link between classical models of gene action and gene regulatory biology. Together with available empirical data, our results indicate that allele interaction can be exploited experimentally to identify and explain intricate intra- and inter-locus feedback relationships in eukaryotes.

Figure 1. Example illustrating the classical definition of gene action and the proposed definition of allele interaction.

(A) Phenotype axis with five genotypic values for five genotypes (two homozygote, one heterozygote and two hemizygotes) for a biallelic locus X. (B) The classical definition of gene action (eqn. (2)). Additive genotypic value is defined as half the distance between the two homozygotes while dominance genotypic value is defined as the difference between the heterozygote genotypic value and the midpoint between the two homozygote genotypic values. (C) Proposed definition of allele interaction (eqn. (1)). The allele interaction value for the heterozygote is defined as the difference between the heterozygote genotypic value and the sum of the two hemizygote genotypic values.

Figure 2. Allele interaction sign motifs and dominance generated by negative and positive autoregulation.

(A) Relative frequency of the 18 different allele interaction sign motifs with positive and negative autoregulation. (B) The frequency distributions of negative dominance, no dominance, and positive dominance for the displayed sign motifs. (C) Box plots of the scaled dominance values () for the various sign motifs. See text for equations and parameter value ranges. 50000 simulations were run for each type of autoregulation, 47186 and 12054 valid datasets are shown for negative and positive autoregulation, respectively.

Figure 3. Allele interaction sign motifs generated by nonmonotonic autoregulation.

Relative frequency of the 18 different sign motifs with feedback regulation and the nonmonotonic dose-response relationships in eqn. (7). 10000 simulations were run for each type of autoregulation, results for 9631 and 6527 valid datasets are shown for /\-shaped and \/-shaped dose-response functions, respectively (see Methods for parameter values and details).

Figure 4. Allele interaction sign motifs generated by two-element feedback loops.

Relative frequencies of allele interaction sign motifs displayed by four different two-element feedback loops when genetic variation is only present in locus 1. In both panels blue bars represent the steady state expression levels assayed at locus 1 (polymorphic), while red bars represent steady state expression levels assayed at locus 2 (nonpolymorphic). (A) Results for negative two-element loops. Dark colors: Locus 2 acts negatively on locus 1 (9454 valid datasets). Bright colors: Locus 2 acts positively on locus 1 (9346 valid datasets). (B) Results for positive two-element loops. Dark colors: Both loci act positively on each other (279 valid datasets). Bright colors: Both loci act negatively on each other (9411 valid datasets). Parameter ranges were the same as for the one-locus simulations. 10000 simulations were run for each type of two-element feedback loop.

Figure 5. Overdominance generated by nonmonotonic autoregulation.

Frequencies of partial dominance/additive gene action (red bars) and overdominance (blue bars) for autoregulation with nonmonotonic dose-response relationship. Frequencies of type of dominance are shown within each sign motif, see Fig. 2 for the corresponding overall sign motif frequencies. Results for /\-shaped and \/-shaped dose-response functions are shown in (A) and (B), respectively (see Methods for details).