A systems approach reveals that the myogenesis genome network is regulated by the transcriptional repressor RP58

Abstract

We created a whole-mount in situ hybridization (WISH) database, termed EMBRYS, containing expression data of 1520 transcription factors and cofactors expressed in E9.5, E10.5, and E11.5 mouse embryos--a highly dynamic stage of skeletal myogenesis. This approach implicated 43 genes in regulation of embryonic myogenesis, including a transcriptional repressor, the zinc-finger protein RP58 (also known as Zfp238). Knockout and knockdown approaches confirmed an essential role for RP58 in skeletal myogenesis. Cell-based high-throughput transfection screening revealed that RP58 is a direct MyoD target. Microarray analysis identified two inhibitors of skeletal myogenesis, Id2 and Id3, as targets for RP58-mediated repression. Consistently, MyoD-dependent activation of the myogenic program is impaired in RP58 null fibroblasts and downregulation of Id2 and Id3 rescues MyoD's ability to promote myogenesis in these cells. Our combined, multi-system approach reveals a MyoD-activated regulatory loop relying on RP58-mediated repression of muscle regulatory factor (MRF) inhibitors.

Figure 1. Myogenic Expression of RP58 Based on the WISH Database

(A) Flow chart of database construction. (B) Representation of annotated gene expression. Gene expression detected by WISH was evaluated in several anatomical structures. (Left) Lateral views of embryos at E9.5, 10.5, and 11.5. Hybridization signals detected in colored areas were evaluated visually. Magnification of each photo is not uniform. (Upper Right) Anatomical drawing of the brain at each stage. (Bottom Right) Dorsal view of trunk. Boxed area is enlarged and detailed structures are visualized by WISH performed with markers of each structure. PSM, presomitic mesoderm. (C) Signal intensity is evaluated as follows: 2, intense (as in the posterior part of the forelimb bud); 1, moderate (as in Progress Zone, PZ); or blank, no signal or not determined. Areas showing no detectable signal, such as the regions of brain, are annotated “Not Stained,” and staining that could not be interpreted is evaluated as “Not Determined” and indicated as blank cells in Table S4. (D) Gene expression patterns in the developing limb bud are shown in detail, and colored areas are individually evaluated. AER, apical ectodermal ridge; PZ, progress zone.

Figure 2. RP58 Expression in Mouse Limb Bud and in C2C12 Culture

(A) Expression patterns of muscle-related transcription factors (Pax3, Myf5, MyoD, and Myog) and RP58 in mouse limb bud at E9.5, 10.5, and 11.5, as determined by WISH. At the bottom are sections of an E10.5 (Left) or E11.5 (Right) embryo showing RP58 expression. Abbreviations: m, myotome; drg, dorsal root ganglia; nt, neural tube; dmm, dorsal muscle mass; vmm, ventral muscle mass; fl, forelimb. Bar, 500 μm. (B) Real-time PCR analysis of RP58 and myogenesis-associated genes in C2C12 cells cultured in GM or DM for 0, 2, or 4 days (d). Error bars, SEM (n = 3). (C) Western blotting for RP58 during C2C12 cell myogenic stages.

Figure 3. RP58 Knockdown Inhibits Myogenesis

(A) Western blotting for RP58 in C2C12 lines stably expressing RP58 shRNA (shRP58) or control shRNA (Ctl) on differentiation day 2. (B) Immunocytochemistry for MyHC and DAPI staining in C2C12 lines stably expressing RP58 shRNA (shRP58) or control shRNA (Ctl) on differentiation day 4. Bar, 200 μm. (C) Percentage of MyHC-positive nuclei in RP58 knockdown (shRP58) or control (Ctl) C2C12 cells of (B) in three independent fields. Error bars, SD (n = 3). (D) H&E staining of hindlimb muscles from WT and RP58^−/− mice (E18.5). Higher magnifications of boxed regions in left panels are shown in right panels. Arrowheads indicate an abnormal population of mononucleate cells in RP58^−/− skeletal muscle tissues. Bar, 50 μm. (E) Haematoxylin and eosin staining of diaphragm from WT and RP58^−/− mice (E18.5, arrow). Lu, lung; Li, liver. Bars, 50 μm.

Figure4. High-Throughput TransfectionScreening of RP58 Initiators

(A) HTS scheme. (B) Luciferase activity of pGL4.12, pGL4.12-RP58-1.6K, or pGL4.12-RP58-3.4K reporters in 293T cells tranected with MyoD-, NEUROD1-, NEUROG1-, NEUROG2-expressing vectors or empty vector. Error bars, SD (n = 3). (C) (Upper panel) Reporter vector pGL4.12-RP58-1.6K and the corresponding upstream region of RP58. Relative positions and sequences of E-boxes E1, E2, and E3, which are potential MyoD binding sites, are shown with mutations introduced into these regions. (Lower panel) Luciferase activity of various reporters, mutated either individually or in combination, in 293T cells transfected with MyoD-expressing or empty vector. Error bars, SD (n = 3). (D) Quantitative ChIP analysis using anti-MyoD or anti-acetyl-histone 3 antibody (α-Ac-H3) on E1 and E2, potential MyoD binding sites on RP58 promoter, in C2C12 cells cultured in GM or DM. Error bars, SEM (n = 3). (E) Immunohistochemistry of transverse sections of fore-limb level somites in E10.5 wild-type (WT) or RP58^−/− mice using anti-MyoD or -Myog antibodies. Counterstaining was performed using DAPI. Bar, 100 μm.

Figure 5. Microarray Analysis for RP58 shRNA-Expressing C2C12 Cells

(A) Venn diagram of genes upregulated by treatment with RP58 shRNA (“shRNA”) and downregulated as C2C12 cells differentiate (“timecourse”). (B) Real-time PCR of Id2, Id3, and Ckm in C2C12 cells cultured in growth (GM) or differentiation (DM) medium at 0, 2, or 4 days (d) or in control- (Ctl) or RP58- (sh) shRNA-expressing C2C12 cells. Error bars, SEM (n = 3). (C) Real-time PCR of Id2 and Id3 in C2C12 cells infected with RP58-expressing or control adenovirus. Error bars, SEM (n = 3). (D) Real-time PCR of Id2, Id3, and Ckm in diaphragm of RP58 KO or WT mice (E18.5). Error bars, SEM (n = 3). (E) Immunohistochemistry of transverse section of wild-type (RP58^+/+) and RP58 KO (RP58^−/−) mice at E10.5 (Left panel) and E13.5 (Right panel), using anti-Id2 (for both E10.5 and E13.5) or -Id3 (for E10.5) antibodies. Counterstaining was performed using DAPI. In E13.5, boxed areas are enlarged in the right panel. Bars, 100 μm (E10.5), and 200 μm (E13.5), respectively.

Figure 6. Id2 and Id3 Are Direct Targets of RP58

(A) Location and sequence (in mouse, human, and opossum) of highly conserved putative RP58 binding sites in Id2 (upper) or Id3 (bottom) promoters. (B) Luciferase activity of TK-Luc (Empty), BS10-Luc (BS10), Id2-Luc (Id2), or Id3-Luc (Id3) reporters in C2C12 cells transfected with RP58-expressing or empty vector. Error bars, SD (n = 3). (C) Quantitative ChIP assay in C2C12 cells infected with Flag-RP58-expressing adenovirus. Error bars, SEM (n = 3). (D) (Upper) Immunofluorescence for MyHC expression in WT or RP58 null 10T1/2 fibroblasts in which the myogenic program is activated upon ectopic expression of MyoD on differentiation day 3 (3d). (Lower) Real-time PCR of Ckm, Id2, and Id3 in WT or RP58 null 10T1/2 fibroblasts expressing MyoD on differentiation day 3 (3d). Error bars, SEM (n = 3). (E) (Upper) Immunofluorescence for MyHC expression in WT or RP58 null 10T1/2 fibroblasts in which the myogenic program is activated upon ectopic expression of MyoD and siRNA-mediated downregulation of Id2 and Id3 on differentiation day 2 (2d). (Lower) Real-time PCR of Ckm (Left) and Id2, Id3 (Right) in WT or RP58 null 10T1/2 fibroblasts in the same conditions described above on differentiation day 2 (2d). Error bars, SEM (n = 3). (F) Proposed myogenesis regulatory network.