Anxiety and fear in a cross of C57BL/6J and DBA/2J mice: mapping overlapping and independent QTL for related traits

Abstract

Anxiety, like other psychiatric disorders, is a complex neurobehavioral trait, making identification of causal genes difficult. In this study, we examined anxiety-like behavior and fear conditioning (FC) in an F(2) intercross of C57BL/6J and DBA/2J mice. We identified numerous quantitative trait loci (QTL) influencing anxiety-like behavior in both open field (OF) and FC tests. Many of these QTL were mapped back to the same chromosomal regions, regardless of behavior or test. For example, highly significant overlapping QTL on chromosome 1 were found in all FC measures as well as in center time measures in the OF. Other QTL exhibited strong temporal profiles over testing, highlighting dynamic relationship between genotype, test and changes in behavior. Next, we implemented a factor analysis design to account for the correlated nature of the behaviors measured. OF and FC behaviors loaded onto four main factors representing both anxiety and fear behaviors. Using multiple QTL modeling, we calculated the percentage variance in anxiety and fear explained by multiple QTL using both additive and interactive terms. Quantitative trait loci modeling resulted in a broad description of the genetic architecture underlying anxiety and fear accounting for 14-37% of trait variance. Factor analysis and multiple QTL modeling showed both unique and shared QTL for anxiety and fear; suggesting a partially overlapping genetic architecture for these two different models of anxiety.

Figure 1. 3-D LOD score plot for significant QTL on chromosome 1 exhibiting a temporal relationship with changes in freezing on Day 1

A). 3-D representation of LOD scores for day 1 FC in 30-s time bins. The QTL is not significant until the first tone-CS presentation (T1). During each subsequent 30-s time bin (z-axis) the LOD score increases, peaking at the second CS presentation (T2). B). Relationship between %freezing (bars) and LOD scores (open symbols) on day 1 illustrates the relationship between changes in behavior and the emergence of the QTL. C). Relationship between %freezing (bars) and LOD scores (open symbols) on day 3. In contrast to day 1, the QTL precedes changes in behavior. The QTL is significant when freezing is low and does not increase during CS presentation when %freezing increases. *Significant LOD score, p < 0.05 (LOD score > 3.8).

Figure 2. Genome-wide LOD scores for factors derived from the factor analysis of open field and fear conditioning data

Dashed line represents the genome-wide significance threshold (p < 0.05).

Figure 3. Genome-wide LOD scores for open field behavior (OF)

Single QTL analysis results for OF measures: distance traveled in the periphery (cm/10min; top chart), distance traveled in the center (cm/10min; middle chart) and % time in the center of the arena (bottom chart). Grey QTL traces indicate LOD scores when sex is included as and interactive covariate. Black (no covariate) and grey (sex as interactive covariate) LOD thresholds represent significance at p < 0.05.

Figure 4. Genome-wide LOD scores for fear conditioning (FC) freezing behavior

Single QTL analysis results for FC measures: %freezing to tone day 1, %freezing to context, %freezing to altered context and %freezing to tone day 3. Grey QTL traces indicate LOD scores when sex is included as and interactive covariate. Black (no covariate) and grey (sex as interactive covariate) LOD thresholds represent significance at p < 0.05.

Figure 5. Stepwise QTL models of OF behavior with QTL interactions

Models selection results for distance traveled in the periphery and center and corresponding profile LOD score plots. Interactions are represented as QTL products (i.e. (1*2)). ^! indicates a second QTL position on a previously identified chromosome.

Figure 6

Stepwise QTL models of FC behavior with QTL interactions. Models selection results for %freezing to context, %freezing to altered context and %freezing to tone day 3. Interactions are represented as QTL products (i.e. (1 Ö 5)). ! Second QTL position on a previously identiï¬ ed chromosome.