Egr2-dependent microRNA-138 is dispensable for peripheral nerve myelination

Abstract

Recent studies have elucidated the crucial role for microRNAs in peripheral nerve myelination by ablating components of the microRNA synthesis machinery. Few studies have focused on the role of individual microRNAs. To fill this gap, we focused this study on miR-138, which was shown to be drastically reduced in Dicer1 and Dgcr8 knockout mice with hypomyelinating phenotypes and to potentially target the negative regulators of Schwann cell differentiation. Here, we show that of two miR-138 encoding loci, mir-138-1 is the predominant locus transcribed in Schwann cells. mir-138-1 is transcriptionally upregulated during myelination and downregulated upon nerve injury. EGR2 is required for mir-138-1 transcription during development, and both SOX10 and EGR2 bind to an active enhancer near the mir-138-1 locus. Based on expression analyses, we hypothesized that miR-138 facilitates the transition between undifferentiated Schwann cells and myelinating Schwann cells. However, in conditional knockouts, we could not detect significant changes in Schwann cell proliferation, cell cycle exit, or myelination. Overall, our results demonstrate that miR-138 is an Egr2-dependent microRNA but is dispensable for Schwann cell myelination.

Figure 1

miR-138 is downregulated upon nerve injury. Quantitative RT-PCR expression levels of selected SC genes and miR-138, 338-3p and 146b in injured nerves ((A) crushed nerve, (B) transected nerve) (gray bars) and contralateral controls (black bars) (*p < 0.05, **p < 0.001, ***p < 0.005). For crushed nerve experiments, sciatic nerves of wildtype mice were harvested five days post-surgery. For transected nerve experiments, sciatic nerves were harvested four days post-surgery.

Figure 2

mir-138-1 is the predominant locus transcribed in SC. (A) Schematic representation of the microRNA allele. A promoter-less lacZ reporter with an internal ribosome entry site (IRES) with a polyA (pA) signal, β-actin-driven neomycin selection marker with a pA signal, and a microRNA stem-loop flanked by loxP sites were targeted into the microRNA locus. The mice can be crossed with either germline- or tissue-specific Cre transgenic mice. We crossed mir-138-1 mice and mir-138-2 mice with germline deleter β-actin::Cre⁺ mice to produce mice with a lacZ-tagged deleted allele that lacked the β-actin promoter-neomycin casette (del). (B) Teased sciatic nerves of heterozygous lacZ-tagged mir-138-1^del/wt mice and heterozygous lacZ-tagged mir-138-2^del/wt mice at newborn and P14, stained with Xgal (blue) and Nuclear Fast Red (pink/ light red) (N ≥ 3). Arrow heads point to SC with oval-shaped nuclei. Insert shows one SC at high magnification. (C) miR-138 quantitative RT-PCR of P14 mir-138-1^del/del and mir-138-2^del/del sciatic nerves. miR-138 level is 28 fold lower in the sciatic nerves of the mir-138-1^del/del homozygous mutants than the controls (N ≥ 3, p = 0.0019), while miR-138 level in the mir-138-2^del/del homozygous mutants does not significantly differ from the controls (***p < 0.005).

Figure 3

Egr2 is required for mir-138-1 transcription during development and both SOX10 and EGR2 bind to an active enhancer near the mir-138-1 locus. (A) Quantitative RT-PCR expression levels of selected SC genes (Mpz, Egr2 and Sox2) and microRNA candidates (miR-138, 338-3p and 146b) in Egr2-null sciatic nerves (gray bars) and littermate controls (black bars) at P5 (*p < 0.05, **p < 0.001, ***p < 0.005). (B) Shown are whole mount Xgal staining of P7 mir-138-1::LacZ and Egr2^−/−; mir-138-1::LacZ sciatic nerves. Blue Xgal staining reports endogenous transcription activities at mir-138-1 locus. Transcription activity at mir-138-1 at P7 is detectable in control sciatic nerve but is not detectable in Egr2^−/− sciatic nerves. (C) The UCSC genome browser view shown in the figure displays peaks of SOX10 and EGR2 binding based on analysis of the ChIP-Seq data from pooled P15 rat sciatic nerves. Histone H3 K27Ac data is a track of active enhancers also obtained from P15 sciatic nerves. The peaks are co-localized (red arrow) and within 100 kb from mir-138-1 locus (red box).

Figure 4

Sciatic nerves of P4 mir-138-1 cKO mice appear to be morphologically similar to the controls. (A) Semi-thin cross sections of P4 mir-138-1 cKO and control sciatic nerves stained with Toluidine blue. In both mutants and controls, normal numbers of myelin sheaths are observed in the nerves. (B) Electron microscopic analyses of P4 mir-138-1 cKO and control sciatic nerves in cross sections. In both mutants and controls, myelinating SC are mainly observed. (C) Percentage (%) of axons in axon diameter ranges (μm). There is no significant difference in axon diameter distribution in P4 control and mir-138-1 cKO nerves. (D) Scatter plot of g-ratio (axon diameter/fiber diameter) versus axon diameter. There is no significant difference in the relationship between axon diameter and myelin thickness in P4 control and mir-138-1 cKO nerves. (E) Box-and-whisker plot of g-ratios measured from the P4 control and mir-138-1 cKO axons. >400 axons were measured in each sample. (F) Percentage (%) of axons in g-ratios ranges. There is no significant difference in myelin thickness distribution between P4 mir-138-1 cKO and control nerves. N = 3 in (A–F).

Figure 5

mir-138-1 cKO sciatic nerves do not display alterations in proliferation, cell cycle exit, and in cell numbers. (A) EdU (green), EGR2 (red) and DAPI (blue) staining of P4 mutant and control sciatic nerves. (B) Quantification of EGR2 expression (EGR2+/DAPI+) and cell proliferation rate (EdU+/DAPI+) in P4, and quit fraction (Ki67-EdU+/EdU+) and cell numbers (DAPI+/area) in P4 mutant and control sciatic nerves. There is no significant difference detected in the number of EGR2-expressing cells (N ≥ 4; P4, p = 0.7847), cell proliferation rate (N ≥ 4; P4, p = 0.6942), quit fraction (N ≥ 4; P4, p = 0.1500), or cell number per unit area (N ≥ 4; P4, p = 0.4910). Black bar = controls. Gray bar = mir-138-1 cKOs.

Figure 6

Sciatic nerves of adult mir-138-1/-2 cKOs are myelinated normally. (A) Semi-thin cross sections stained with Toluidine blue and (B) Electron microscopic photos of sciatic nerves from adult mir-138-1/-2 cKO mice and the littermate controls. In both mutants and controls, myelin sheaths are observed in the nerves. (C) Scatter plot of g-ratios and axon diameters shows no significant difference in myelin thickness and axon size relationship between adult mir-138-1/-2 cKO mutants and controls (N = 3).