Analysis of neuroanatomical differences in mice with genetically modified serotonin transporters assessed by structural magnetic resonance imaging

Abstract

Background: The serotonin (5-HT) system has long been implicated in autism spectrum disorder (ASD) as indicated by elevated whole blood and platelet 5-HT, altered platelet and brain receptor and transporter binding, and genetic linkage and association findings. Based upon work in genetically modified mice, 5-HT is known to influence several aspects of brain development, but systematic neuroimaging studies have not previously been reported. In particular, the 5-HT transporter (serotonin transporter, SERT; 5-HTT) gene, Slc6a4, has been extensively studied.

Methods: Using a 7-T MRI and deformation-based morphometry, we assessed neuroanatomical differences in an Slc6a4 knockout mouse on a C57BL/6 genetic background, along with an Slc6a4 Ala56 knockin mouse on two different genetic backgrounds (129S and C57BL/6).

Results: Individually (same sex, same background, same genotype), the only differences found were in the female Slc6a4 knockout mouse; all the others had no significant differences. However, an analysis of variance across the whole study sample revealed a significant effect of Slc6a4 on the amygdala, thalamus, dorsal raphe nucleus, and lateral and frontal cortices.

Conclusions: This work shows that an increase or decrease in SERT function has a significant effect on the neuroanatomy in 5-HT relevant regions, particularly the raphe nuclei. Notably, the Slc6a4 Ala56 knockin alone appears to have an insignificant, but suggestive, effect compared to the KO, which is consistent with Slc6a4 function. Despite the small number of 5-HT neurons and their localization to the brainstem, it is clear that 5-HT plays an important role in neuroanatomical organization.

Keywords: 5-HT; 5HTT; Brain; Dorsal raphe; Magnetic resonance imaging; Neurodevelopment; Serotonin; Slc6a4.

Fig. 1

Bar graphs showing the absolute volume for the total brain and seven summary regions encapsulating the entire brain volume. The asterisk represents a significant difference between the single sex group versus its same background wild-type counterpart. The dagger symbol represents a significant difference between the full group (males and females) versus its same background wild-type. Slc6a4 KO (N = 39, 19 WT (9M, 10F), 20 KO (10M, 10F); Slc6a4 Ala56 KI (B6) (N = 40, 20 WT (10M, 10F), 20 KI (10 M, 10F); Slc6a4 Ala 56 KI (129) (N = 42, 21 WT (11M, 10F), 21 KI (11M, 10F)

Fig. 2

Heatmap showing the effect size differences in 159 different regions throughout the brain for the Slc6a4 KO group and the full Slc6a4 Ala56 KI group regardless of background. The effect size is calculated as the difference in means divided by the standard deviation of the WT group (Effect Size = (μ_MUT − μ_WT)/σ_WT); it is measured in units of standard deviation. The heatmap is organized from the anterior of the brain at the top to the posterior at the bottom. Any region in blue is smaller in the mutant versus wild-type and any in red are larger. The asterisk indicates significance at a q < 0.05, The dot indicates a trend at p < 0.05. Slc6a4 KO (N = 39, 19 WT (9M, 10F), 20 KO (10M, 10F); Slc6a4 Ala56 KI (B6) (N = 40, 20 WT (10M, 10F), 20 KI (10M, 10F); Slc6a4 Ala 56 KI (129) (N = 42, 21 WT (11M, 10F), 21 KI (11M, 10F)

Fig. 3

This figure shows the significant voxelwise differences between the Slc6a4 KO and its corresponding WT. Anything shown in blue/cyan is significantly smaller than the WT and anything in red/orange is significantly larger. Slc6a4 KO (N = 39, 19 WT (9M, 10F), 20 KO (10M, 10F)

Fig. 4

a–d This figure shows significant voxelwise differences as measured using an ANOVA while factoring in both the sex and background of the mice. All 121 mice were used in this comparison. The panel on the right displays a coronal flythrough highlighting the location of volume differences found due to the Slc6a4 gene in both absolute and relative volume. For this figure, the C57Bl6/J mice from the Slc6a4 KO group were normalized to the C57Bl6/J mice from the Slc6a4 Ala56 KI group (see the “Methods” section). Slc6a4 KO (N = 39, 19 WT (9M, 10F), 20 KO (10M, 10F); Slc6a4 Ala56 KI (B6) (N = 40, 20 WT (10M, 10F), 20 KI (10M, 10F); Slc6a4 Ala 56 KI (129) (N = 42, 21 WT (11M, 10F), 21 KI (11M, 10F)

Fig. 5

Association between the dorsal raphe nucleus (DRN) fiber-tract connectivity and SERT volume differences. a Shown in vertical panels are coronal slices through the mouse brain from anterior (top row) to posterior (bottom row), see top right of figure for location of slices relative to brain. Vertical panel columns correspond to i) anatomy (via MRI), ii) absolute volume differences between groups (F-statistic), iii) neuronal tracers projecting anterograde from the dorsal raphe nucleus, and iv) voxels which show significant differences at a FDR q < 0.05 (F > 4.56) that also express a tracer signal