Development of Erf-Mediated Craniosynostosis and Pharmacological Amelioration

Abstract

ETS2 repressor factor (ERF) insufficiency causes craniosynostosis (CRS4) in humans and mice. ERF is an ETS domain transcriptional repressor regulated by Erk1/2 phosphorylation via nucleo-cytoplasmic shuttling. Here, we analyze the onset and development of the craniosynostosis phenotype in an Erf-insufficient mouse model and evaluate the potential of the residual Erf activity augmented by pharmacological compounds to ameliorate the disease. Erf insufficiency appears to cause an initially compromised frontal bone formation and subsequent multisuture synostosis, reflecting distinct roles of Erf on the cells that give rise to skull and facial bones. We treated animals with Mek1/2 and nuclear export inhibitors, U0126 and KPT-330, respectively, to increase Erf activity by two independent pathways. We implemented both a low dosage locally over the calvaria and a systemic drug administration scheme to evaluate the possible indirect effects from other systems and minimize toxicity. The treatment of mice with either the inhibitors or the administration scheme alleviated the synostosis phenotype with minimal adverse effects. Our data suggest that the ERF level is an important regulator of cranial bone development and that pharmacological modulation of its activity may represent a valid intervention approach both in CRS4 and in other syndromic forms of craniosynostosis mediated by the FGFR-RAS-ERK-ERF pathway.

Keywords: Ets2 repressor factor; craniosynostosis; pharmacological modulation.

Figure 1

The development and penetrance of craniosynostosis phenotype in Erf^loxp/− mouse model. (A) Representative microCT images of mouse skulls at the indicated postnatal time points. The black arrow indicates frontal bone defect. The yellow arrow indicates the deformity of the frontal suture. Cranial sutures in top left panel are named in yellow (P., posterior, A., anterior) and bones in red letters. White scale bars, 5 mm. (B) Graph showing the percentage of the total length of each suture that is either closed (positive y-axis) or open in comparison to control (negative y-axis) considering all the animals at the indicated developmental stages. (C) The percentage of mice exhibiting a defect in the indicated suture for each developmental group is shown. Positive values in the y-axis are used when the defect is a premature closure of the suture and negative values when the defect involves a widely open suture with a gap, as shown by the black arrow in (A). The PF suture at P25, although displaying deformities, is neither closed nor more open compared to the control state. Each measurement in (B,C) is based on four male mice except for the P45 group, which includes three male mice. The P65 group includes the 6 mice from Figure 5A below. (D,E) The ratio of phosphates (bone), collagen (Phe), proline (Pro) and hydroxyproline (Hyp) shown in (D) and crystallinity shown in (E) was evaluated by Raman spectroscopy of coronal sutures from P65 Erf-competent (Erf^loxP/+) and Erf-insufficient (Erf^loxP/−) mice. Significant differences were observed in the mineralization of the suture but not the components of the suture. Spectra recorded from the area of the suture with the minimal I961/I1005 ratio and the collagen content relative to mineral (Supplementary Figure S2 and Table S3) ensured the sutures were not fully ossified. All components of the calvarium fragments shown were measured using metrology associated with polarized Raman spectroscopy presented, in detail, in Supplementary Table S3. Data were analyzed with unpaired t-test with two-tailed distribution. * p < 0.05, ** p < 0.01.

Figure 2

Elimination of Erf in cells of neuroectodermal origin causes a frontal bone development defect but not craniosynostosis. Representative skull microCT scans of Nestin-cre/+;Erf^loxP/loxP P65 mice carrying homozygous deletion of Erf in cells of neuronal origin and age-matched Nestin-cre/+ mouse. Although compromised development of frontal bones (black arrow) is evident in two of the three animals tested and facial deformity in all three animals, cranial sutures do not display synostosis. Cre recombinase activity in the neuronal tissues was verified on postnatal day 2 through crossing with the Gt(ROSA)26Sor^{tm4(ACTB-tdTomato,-EGFP)Luo} mouse line.

Figure 3

Nuclear accumulation of Erf after MEK1/2 or Xpo1 inhibition and the effect on osteogenic differentiation. Evaluation of the active concentration of inhibitors in cellular assays. (A,B) Mek1/2 inhibition by U0126 (A) and Xpo1 inhibition by KPT-330 (B) were evaluated by monitoring the subcellular localization of GFP-Erf in HeLa cells after 2 h exposure to the drugs. The average of at least 3 independent experiments is shown. (C) Calcium deposition per cell of cell cultures derived from coronal and sagittal sutures growing for 28 days in osteogenic medium in the presence of 2 μΜ U0126 and 10 nM KPT-330. Cell numbers were estimated using MTT assay by the absorbance at 600 nm and calcium deposits by the absorbance of Alizarin Red S at 405 nm after its extraction from the cells. Three independent biological experiments, with two experimental replicates each, were performed. In all cases, data were analyzed with one-way ANOVA followed by Dunnett’s (two-sided) test to compare all groups (treatments) against control group. * p < 0.05, *** p <0.001.

Figure 4

Local administration of Mek1/2 and Xpo1 inhibitors does not have system-wide effects. (A) Animals were injected either with 0.5 mg/kg of the inhibitors subcutaneously (subcut) over the skull or 50% DMSO or with 5 mg/kg of the inhibitors intraperitoneally (Intraper), per animal treatment protocol. Blood was collected after 6 h without sacrificing the animals, and the effect of the inhibitors was evaluated by the localization of GFP-Erf after the addition of 1% mouse serum in the cell cultures. KPT-330, (left panel). U0126, (right panel). The experiment was performed in triplicate. * p < 0.05, ** p < 0.01. (B) Effect on the weight of the animals at P65 per treatment protocol with 0.5 mg/kg of the inhibitors subcutaneously over the skull (left panel) or 5 mg/kg of the inhibitors intraperitoneally (right panel). Seven to ten animals were included in each group, and the data were analyzed using one-way ANOVA followed by Bonferroni correction for comparisons among the groups and unpaired (two-sided) t-test for comparisons between the two genotypes in each group; * p < 0.05, ** p < 0.01.

Figure 5

Mek1/2 and Xpo1 inhibitors can ameliorate Erf-related craniosynostosis. (A) Volume renderings (3 images on the left) and transverse section image midway through the right coronal suture (rightmost image) derived from microCT scans of representative Erf-competent (Erf^loxP/+) and Erf-insufficient (Erf^loxP/−) animals at P65. Littermates were treated intraperitoneally on alternate days from P5 to P65 with 5 mg/kg of U0126 (UI) or KPT-330 (KI) or subcutaneously with 0.5 mg/kg of U0126 (US) or KPT-330 (KS) over the skull or with the inhibitor solvent only (DMSO). All animals in a litter received the same treatment. Dorsal anterior (left), dorsal posterior (middle) and lateral (right) images of representative animal skulls are shown. Volumetric images were constructed using the same opacity and luminescence settings. Red arrows indicate animals with dental malocclusion. Yellow arrows indicate the coronal suture. White scale bars, 5 mm; blue scale bars, 1 mm. (B) Evaluation of suture synostosis from the microCT scans of Erf^loxp/− treated animals. The percentage of mice exhibiting a defect in the indicated sutures for each treatment group is shown in the left panel. The right panel indicates the percentage of the total length of each suture that was fully ossified, considering all the animals in each group. The evaluation was performed independently by two scientists blind to genotype and treatment. Each measurement is based on four mice, with the exception of KI treatment group, which includes three mice and the DMSO group, which includes six mice. All groups have one female mouse except the DMSO group, which has two.

Figure 6

Mek1/2 U0126 and Xpo1 KPT-330 inhibitors improve skull morphology of the Erf^loxP/− mice. (A) For the semiautomated analysis of the curvature of the skull, the lambda and bregma points were marked on the volume renderings of the microCT scans (upper panels), and then all skulls were programmatically aligned to a reference normal skull (lower left image), and a transverse section at ¼ of the lambda/bregma axis (measured from the bregma) was used to fit the ellipse. Cranial sutures in top left panel are named in yellow letters. (B) Curvature of the skull as evaluated by the b/a ratio of the fitted ellipse. Average of 3–5 male animals per group treated as in (A) above. Data were analyzed with one-way ANOVA followed by Dunnett’s (two-sided) test to compare all groups (treatments) against control group (Erf^loxP/+). (C) The overall width of the calvarium was measured directly from the microCT volume rendering with Bruker DataViewer 1.5.6.2 software.