Genetic Influences on Behavioral Outcomes After Childhood TBI: A Novel Systems Biology-Informed Approach

Abstract

Objectives: To test whether genetic associations with behavioral outcomes after early childhood traumatic brain injury (TBI) are enriched for biologic pathways underpinning neurocognitive and behavioral networks. Design: Cross-sectional evaluation of the association of genetic factors with early (~ 6 months) and long-term (~ 7 years) post-TBI behavioral outcomes. We combined systems biology and genetic association testing methodologies to identify biologic pathways associated with neurocognitive and behavior outcomes after TBI. We then evaluated whether genes/single nucleotide polymorphism (SNPs) associated with these biologic pathways were more likely to demonstrate a relationship (i.e., enrichment) with short and long-term behavioral outcomes after early childhood TBI compared to genes/SNPs not associated with these biologic pathways. Setting: Outpatient research setting. Participants:140 children, ages 3-6:11 years at time of injury, admitted for a TBI or orthopedic injury (OI). Interventions: Not Applicable. Main Outcome Measures: Child behavior checklist total problems T score. Results: Systems biology methodology identified neuronal systems and neurotransmitter signaling (Glutamate receptor, dopamine, serotonin, and calcium signaling), inflammatory response, cell death, immune systems, and brain development as important biologic pathways to neurocognitive and behavioral outcomes after TBI. At 6 months post injury, the group (TBI versus OI) by polymorphism interaction was significant when the aggregate signal from the highest ranked 40% of case gene associations was compared to the control set of genes. At ~ 7 years post injury, the selected polymorphisms had a significant main effect after controlling for injury type when the aggregate signal from the highest ranked 10% of the case genes were compared to the control set of genes Conclusions: Findings demonstrate the promise of applying a genomics approach, informed by systems biology, to understanding behavioral recovery after pediatric TBI. A mixture of biologic pathways and processes are associated with behavioral recovery, specifically genes associated with cell death, inflammatory response, neurotransmitter signaling, and brain development. These results provide insights into the complex biology of TBI recovery.

Keywords: behavioral outcomes; genetics; pediatrics; systems biology; traumatic brain injury.

Figure 1

Quality control flow diagram and selection of genes and variants on exome chip to be used in analyses.

Figure 2

Number of case and control genes and excluded genes on exome chip after quality control (QC) was completed.

Figure 3

Network representation of select enriched biological processes, pathways and genes. Select enriched biological processes and pathways (p <0.05) are represented as blue rectangles along with their associated training set genes (orange ellipse) and top 10% case genes (purple ellipse). Only genes (training or case sets) that are associated with terms identified by ToppGene suite as being influential to brain injury recovery are shown in the network. Complete details of the functional enrichment results can be found in Supplementary Table.

Figure 4

Differential response to outcome by injury type (TBI vs. OI) for short (A) and long (B) term of behavioral outcomes. Centiles represent the portion of case genes used in the genetic association analysis: 0% includes that only the training list was included, 5th percentile includes the training list plus the top 5% highest ranked genes, 10th centile includes the training list plus the top 10% of ranked genes and so forth until all ranked genes were included (i.e., 100th centile). Vertical axis represents the number of single nucleotide polymorphisms. Box plots represent the number of significant SNPs in the 10,000 runs of control gene SNPs. The dot indicates that number of nominal associations (p < 0.05) identified in case genes. Enrichment is indicated when a greater number of genetic associations are present in case vs. control genes; therefore, when the number of associations in case genes (represented by the dot) exceeded the upper 95th percentile threshold in the run of control genes. In panel A, at the 40th centile and above, the dot is above the 95th percentile indicating that there are more case-gene SNPs significantly associated with short-term behavioral outcomes than what would be expected by chance, indicating enrichment. In panel B, the number of case gene SNP associations are below number of associations in the control set, indicating that case genes are unlikely to be enriched for pathways specific to long-term behavioral outcomes after TBI compared to OI. Because this figure represents the point estimates for the interaction term of group (TBI vs. OI) with SNPs, these findings demonstrate that there is differential outcomes in the TBI vs. the OI group in the short-term rather than the long-term.

Figure 5

Short (A) and Long-term behavioral outcomes (B) when controlling for injury type (TBI vs. OI). Centiles represent the portion of case genes used in the genetic association analysis: 0% includes that only the training list was included, 5th percentile includes the training list plus the top 5% highest ranked genes, 10th centile includes the training list plus the top 10% of ranked genes and so forth until all ranked genes were included (i.e., 100th centile). Vertical axis represents the number of single nucleotide polymorphisms. Box plots represent the number of significant SNPs in the 10,000 runs of control gene SNPs. The dot indicates that number of nominal associations (p < 0.05) identified in case genes. Enrichment is indicated when a greater number of genetic associations are present in case vs. control genes; therefore, when the number of associations in case genes (represented by the dot) exceeded the upper 95th percentile threshold in the run of control genes. In panel A, no enrichment is identified for short-term outcomes. In panel B, at the 10th to the 60th centile, there are more case-gene SNPs associated with long-term behavioral outcomes than what would be expected by chance. Because this figure demonstrates main effects controlling for group (TBI vs. OI), these findings indicate that a complex network of genes/polymorphisms is associated with long-term behavioral outcomes after traumatic injury of any sort (i.e., TBI or OI).