Myelination defects in the medial prefrontal cortex of Fkbp5 knockout mice

Abstract

The hypothalamic-pituitary-adrenal (HPA) axis plays a principal role in stress response regulation and has been implicated in the etiology of stress-related disorders. The HPA axis regulates the normal synthesis and release of glucocorticoids; dysregulation of the HPA axis causes abnormal responses to stress. FK506-binding protein 5 (FKBP5), a co-chaperone of heat shock protein 90 in the glucocorticoid receptor (GR) molecular complex, is a key GR sensitivity regulator. FKBP5 single nucleotide polymorphisms are associated with dysregulated HPA axis and increased risk of stress-related disorders, including posttraumatic stress disorder (PTSD) and depression. In this study, we profiled the microRNAs (miRNAs) in the medial prefrontal cortex of Fkbp5 knockout (Fkbp5^-/- ) mice and identified the target genes of differentially expressed miRNAs using sequence-based miRNA target prediction. Gene ontology analysis revealed that the differentially expressed miRNAs were involved in nervous system development, regulation of cell migration, and intracellular signal transduction. The validation of the expression of predicted target genes using quantitative polymerase chain reaction revealed that the expression of axon development-related genes, specifically actin-binding LIM protein 1 (Ablim1), lemur tyrosine kinase 2 (Lmtk2), kinesin family member 5c (Kif5c), neurofascin (Nfasc), and ephrin type-A receptor 4 (Epha4), was significantly decreased, while that of brain-derived neurotrophic factor (Bdnf) was significantly increased in the brain of Fkbp5^-/- mice. These results suggest that axonal development-related genes can serve as potential targets in future studies focused on understanding the pathophysiology of PTSD.

Keywords: Nfasc; PTSD; small RNA-seq; target prediction.

FIGURE 1

Target prediction and functional annotation of predicted target genes of significantly differentially expressed miRNAs in the brain of Fkbp5^−/− mice. A, Multidimensional scaling plots of wild‐type (WT) and knockout (KO) mice based on small RNA sequencing data. WT4 is considered to be an outlier. Blue, WT; Red, KO. B, Heatmap representing 41 significantly differentially expressed miRNAs in the brain of Fkbp5^−/− mice compared to that in the brain of WT mice (P < .05). Each column represents one mouse. Each row represents one miRNA. Blue and red colors indicate the decreased and increased expression levels, respectively, in Fkbp5^−/− mice relative to WT. The color scale is shown at the top. C, Venn diagram of the predicted target genes of differentially expressed miRNAs (mmu‐miR‐139‐3p) obtained from five online databases. The gray portion includes genes that were obtained from two or more databases. D, Target genes of differentially expressed miRNAs were functionally categorized under the biological process (GO_BP), cellular component (GO_CC), molecular function (GO_MF), and KEGG pathway (KEGG_PATHWAY). The horizontal axis shows the enrichment score of each cluster. Node color represents the Benjamini adjusted P‐value and node size represents the gene count. Adjusted P‐values < 0.1 (Benjamini) were considered significant

FIGURE 2

Validation of mRNA expression of the predicted target genes by quantitative real‐time PCR. The expression patterns of the predicted target genes associated with axonal guidance during neuronal development were examined in the brain of Fkbp5^−/− mice. Results show that the expression of Ablim, Epha4, Kif5c, Lmtk2, Nfasc, and Slit2 was significantly decreased and that of Bdnf was significantly increased in the medial prefrontal cortex of Fkbp5^−/− mice. All Ct values were normalized to those of Gapdh. Data are shown as mean ± SEM (‐dCt) of each gene in wild‐type (WT, n = 3) and Fkbp5^−/− (KO, n = 4) mice. *P < .05 upon comparison of Fkbp5^−/− mice to the WT using unpaired t test

FIGURE 3

The expression of Neurofascin in the brain tissue of Fkbp5^−/− mice. A, Predicted binding sites of miR‐139‐3p to the Nfasc 3ʹ UTR. Luciferase activity was significantly decreased in HEK 293T cells transfected with miR‐139‐3p mimics. Firefly luciferase activity was normalized to Renilla luciferase activity. B, The expression of two Nfasc isoforms was measured by quantitative real‐time PCR. Both isoforms were significantly downregulated in the medial prefrontal cortex of Fkbp5^−/− mice. C, Immunostaining of fresh frozen tissues showing the decreased expression of NFASC and its distinct signal pattern in the medial prefrontal cortex of Fkbp5^−/− mice compared to that in the medial prefrontal cortex of wild‐type (WT) mice (scale bar = 40 µm). D, Polar plot depicting the angle and length of the NFASC signal (upper, WT; lower, KO). E, Immunostaining of freely floating fixed brain tissue sections of Fkbp5^−/− mice. Double immunofluorescence analysis of the expression of NFASC (green) and CASPR (red) in the corpus callosum showed significantly decreased NFASC expression in the brain of Fkbp5^−/− mice (WT n = 4, KO n = 3; scale bar = 40 µm). *P < .05 upon comparison of Fkbp5^−/− mice to the WT using unpaired t test

FIGURE 4

The expression of marker genes associated with axonal development and myelination. A, The expression of marker genes associated with axonal development and myelination was measured by quantitative real‐time PCR. Results show the significantly decreased expression of Mpb and Mag in the medial prefrontal cortex of Fkbp5^−/− mice. All Ct values were normalized to those of Gapdh. Data are shown as mean ± SEM (‐dCt) of each gene in wild‐type (WT, n = 9) and Fkbp5^−/− (KO, n = 11) mice. B, Immunostaining of free‐floating fixed tissue sections showing decreased MBP expression in the medial prefrontal cortex of Fkbp5^−/− mice compared to that in the medial prefrontal cortex of wild‐type (WT) mice (scale bar = 50 µm). C, Western blot showing significantly decreased MBP expression in the medial prefrontal cortex of Fkbp5^−/− mice. Data shown as mean ± SEM (fold change) in WT (n = 8) and KO (n = 8). *P < .05 upon comparison of Fkbp5^−/− mice to the WT using unpaired t test

FIGURE 5

Alteration of myelin status in the cortex and corpus callosum of Fkbp5^−/− mice. A, Representative images showing fluoromyelin staining in the cortex and corpus callosum of Fkbp5^−/− mice (4×, scale bar = 250 µm; 20×, scale bar = 50 µm). B, Fluoromyelin intensity was measured using ImageJ. Fluoromyelin intensity in both the cortex and corpus callosum was significantly decreased in Fkbp5^−/− mice. Cortex (4×), 55.2% (P = 2.77 × 10⁻³); Cortex (20×), 60.1% (P = 2.88 × 10⁻³); Corpus callosum (4×), 70.9% (P = 9.68 × 10⁻⁴); Corpus callosum (20×), 56.4% (P = 3.49 × 10⁻⁴). Data are shown as mean ± SEM (percentage) in wild‐type (WT, n = 6) and Fkbp5^−/− (KO, n = 4) mice. *P < .05 upon comparison of Fkbp5^−/− mice to the WT using unpaired t test