Disruptions of striatal circuits are found in patients with ASD 14, 18, 19 and ADHD 20, as well as genetic mouse models for the study of these disorders 54–57, and are related to core NDD pathologies, such as dysfunctional reward processing and altered locomotor activity. Of note, a male bias is observed in both human patients with these disorders and some of the respective mouse models. The molecular basis of male-specific striatal vulnerability, however, has not been identified.
Based on a novel strategy to identify genes-of-interest based on their spatial gene expression patterns, we identified 3 candidate genes from the 16p11.2 region, Taok2, Sez6l2, and Mvp. 31. Our hypothesis was that these 3 genes alone are sufficient to mediate sex-specific behavioral phenotypes that depend on striatal circuits. Our newly generated 3g del/+ mouse model displays sex-specific striatal behavioral and transcriptomic characteristics, resembling the phenotypes of 16p11.2 del/+ mice. These results demonstrate that our novel approach of selecting genes of interest based on their spatial gene expression patterns in structurally altered brain regions is a valid path to develop data-driven hypotheses that can be later tested by generating model organisms with a selective modification of the respective genes, providing novel insights into polygenic interactions underlying neurodevelopmental disorders.
Our behavioral studies reveal that a selective hemi-deletion of our 3 candidate genes leads to aspects of the striatal behavioral phenotypes also seen specifically in 16p11.2 del male mice. Previously, we found that male, but not female, 16p11.2 del/+ mice show impairments in reward-directed learning as well as motivation to work for rewards. Although 3g del/+ male mice show impaired motivation, we did not find an impairment in reward learning in 3g del/+ male mice. There are two potential explanations for the observed difference between the 3g del/+ mice and 16p11.2 del/+ mice. First, it could be because of a difference in FR protocol difficulty in the two studies. Previously, we applied the five-choice serial reaction time task (5-CSRTT) to 16p11.2 del/+ mice, a task uses 5 holes for correct 5-CSRTT performance 10. In this study, we tried to minimize the number of days in training and restricted access to one hole, reducing extraneous responding to non-active holes. Second, it is possible that the three genes are not sufficient to produce the entire phenotype. For instance, it was reported that a heterozygous deletion of KCTD13, a gene within 16p11.2, induced object recognition memory deficits 32, suggesting potential contribution of other genes related to the phenotypes of the 16p11.2 del/+ mice.
To identify the mechanisms underlying these behavioral phenotypes, we investigated transcriptomic changes in the striatum of 3g del/+ mice and 16p11.2 del/+ mice. The comparisons of striatal DEGs between male 3g del/+ mice and 16p11.2 del/+ mice revealed 248 overlapping DEGs. It is noteworthy that the overlapping DEGs between male 3g and 16p11.2 del/+ mice are statistically significantly enriched above chance (p < 1.890e-112), however, the overlapping DEGs identified in female 3g and 16p11.2 del/+ mice are not (p < 0.093). The pathway analysis of these 248 DEGs revealed that a ribosome-related pathway and translation regulation are significantly dysregulated, highlighting potential mechanisms that underlie the male-specific striatum-dependent behavioral alterations. A recent tissue-specific transcriptome profiling study on 16p11.2 del/+ mouse model also reported significant GO-term enrichments related to translation, in the striatum and cerebellum 58. Although they did not provide a sex specific analysis, these results support possible disruption of translational regulation in 16p11.2 del/+ mice. Reduced basal protein synthesis in the hippocampus of 16p11.2 del/+ mice were reported 59. The downregulation of ribosomal genes could affect translation efficiency leading to protein synthesis dysregulation and corresponding to presentation of NDDs 60–62. Recent studies of NDDs like Fragile X Syndrome and Tuberous Sclerosis have shown sex-specific differences in ribosomal function or translation regulation that may contribute to the sex-specific phenotypes 63–66. Further studies on translational regulation will provide insights into the sex-specific pathological mechanisms of NDD.
Another impactful result of the KEGG enrichment analysis for 248 DEGs is tyrosine/serine/threonine phosphatase activity related genes including Dusp15 (Fig. 3F). Dusp15 plays an important role in regulation of extracellular signal-regulated kinase (ERK) activation 67. The link between Dusp15 and ERK leads us to the idea of a potential link between 3 genes and ERK pathway. Interestingly, our 3 candidate genes are all related to the regulation of ERK signaling. MVP has a well-established association with ERK signaling, serving as a scaffold protein for ERK 68, 69. Taok2 has been shown to interact with Septin7 at the postsynaptic density, which can activate ERK signaling 70, and Sez6l2 phosphorylates PKC 40, which is a known activator of ERK. Although our current study does not provide a clear mechanistic explanation of the combinational effects, one of our hypotheses is the involvement of ERK pathway in 3 gene mediated sex-specific phenotypes. The ERK pathway plays an important role in development, learning, and synaptic plasticity 71, 72. In addition, it has been shown that ERK signaling is involved in striatal behaviors 37, 73–75. We have observed reward-mediated hyperphosphorylated ERK1 in the dorsal striatum from 16p11.2 del/+ males, but not females consistent with the male-specific reward learning deficit in 16p11.2 del/+ mice 10, suggesting a role for ERK1 in striatum-dependent reward learning in males specifically. Therefore, we expect the connection between our 3 candidate genes and the ERK signaling pathway may provide an explanation for the molecular mechanisms of sex-specific striatal changes in 16p11.2 del/+ mice. This hypothesis is also supported by the strong connection between estrogen and the ERK signaling pathway. Estradiol influences memory consolidation via ERK signaling pathway in several brain regions 76–79. It has been shown that estrogen receptors interact with mGluR to activate ERK signaling 80, 81. Therefore, the connection between the 3 genes and ERK with estrogen may contribute to the sex-specific phenotypes of the 16p11.2 del/+ mice. Further studies will need to clarify the connection between ERK signal pathway and these 3 genes as an underlying mechanism of sex-specific phenotypes.
It has been reported that patients with 16p11.2 del syndrome show regional volumetric differences including accumbens, pallidum, caudate, and putamen, with increased FA in medial white matter 12, 53, 82, 83. In addition, 16p11.2 del/+ mice show increased volumes of several brain including striatum, nucleus accumbens, and globus pallidus 48, 84, 85. However, sex- and age-dependent brain structural changes in 16p11.2 deletion are not fully investigated. Previously, we have shown fiber tract changes in both male and female 16p11.2 del mice at 10 weeks of age 31. Here, we found decreased FA in 16p11.2 del/+ females compared to wt females at 6 weeks of age, however, not in males. These distinctive results indicate that FA changes in 16p11.2 del/+ mice are sex- and age-dependent, suggesting brain structural changes occur earlier in development in female 16p11.2 del/+ mice compared to male mice. In addition, the DTI results imply that the 3 gene hemi-deletion may not affect brain structural changes in the same way as 16p11.2 del. A previous study indicates that Taok2 hemi-deletion mice show increased total brain volume compared with wt mice 33, however, 16p11.2 del/+ mice display decreased total brain volume 48, 84. These opposite brain structural changes between Taok2 hemi-deletion mice and 16p11.2 del/+ mice may provide a hint to comprehend these distinctive brain structural changes between 3g del/+ mice and 16p11.2 del/+ mice. Uncovering the genetic mechanism of brain structural changes across development will require further study.
One important consideration for this study is whether the combinatorial effects of 3 genes on the behavioral phenotypes are the ‘sum of single gene effects’ or ‘combinatorial effects of 3 genes’. We believe it is the combinatorial effects of 3 genes based on previously described single gene mutation mouse studies. Although it has been shown that each gene (Taok2, Sez6l2, and Mvp) plays an important role in neuronal development as well as NDD 33–35, 39–44, 86, sex-specific phenotypes have not been observed in studies of single gene variant mouse lines. These studies suggest that the 3 gene hemi-deletion mediates male-specific phenotypes via combinatorial effects of 3 genes rather than just a summation of the effects of each gene. It is still possible that the combination of 2 genes may be sufficient to replicate certain phenotypes include male-specific phenotypes, so we cannot exclude the possibility that one out of the three genes is unnecessary to cause these effects. Here, we show that the 3 genes are sufficient to cause the male-specific phenotypes but not necessity of the 3 genes on the phenotypes. Therefore, further studies about necessity of each gene will be needed.
Overall summary
This study demonstrates a novel approach to dissect large chromosomal regions to identify candidate sets of genes-of-interest and the characterization of a respective gene set in a novel experimental system carrying a selective hemi-deletion of three genes in the 16p11.2 region. 3g del/+ animals recapitulate male-specific striatal phenotypes observed in 16p11.2 del/+ mice. Changes in the expression levels of genes associated with ribosomal processing seem to constitute the transcriptomic basis for male-specific striatal vulnerability to NDDs. These findings could pave the way towards novel therapeutic and preventive strategies for NDDs.