Green tea extracts containing epigallocatechin-3-gallate modulate facial development in Down syndrome

Abstract

Trisomy of human chromosome 21 (Down syndrome, DS) alters development of multiple organ systems, including the face and underlying skeleton. Besides causing stigmata, these facial dysmorphologies can impair vital functions such as hearing, breathing, mastication, and health. To investigate the therapeutic potential of green tea extracts containing epigallocatechin-3-gallate (GTE-EGCG) for alleviating facial dysmorphologies associated with DS, we performed an experimental study with continued pre- and postnatal treatment with two doses of GTE-EGCG supplementation in a mouse model of DS, and an observational study of children with DS whose parents administered EGCG as a green tea supplement. We evaluated the effect of high (100 mg/kg/day) or low doses (30 mg/kg/day) of GTE-EGCG, administered from embryonic day 9 to post-natal day 29, on the facial skeletal development in the Ts65Dn mouse model. In a cross-sectional observational study, we assessed the facial shape in DS and evaluated the effects of self-medication with green tea extracts in children from 0 to 18 years old. The main outcomes are 3D quantitative morphometric measures of the face, acquired either with micro-computed tomography (animal study) or photogrammetry (human study). The lowest experimentally tested GTE-EGCG dose improved the facial skeleton morphology in a mouse model of DS. In humans, GTE-EGCG supplementation was associated with reduced facial dysmorphology in children with DS when treatment was administered during the first 3 years of life. However, higher GTE-EGCG dosing disrupted normal development and increased facial dysmorphology in both trisomic and euploid mice. We conclude that GTE-EGCG modulates facial development with dose-dependent effects. Considering the potentially detrimental effects observed in mice, the therapeutic relevance of controlled GTE-EGCG administration towards reducing facial dysmorphology in young children with Down syndrome has yet to be confirmed by clinical studies.

Figure 1

Principal Component Analyses of the global facial shape variation in Ts65Dn trisomic (TS) mouse models treated with high and low GTE-EGCG doses and euploid wild-type (WT) mice. Scatterplot of PC1 and PC2 axes with the corresponding percentage of total morphological variation explained is displayed along each axis. Convex hulls and horizontal color bars represent the ranges of variation within each group of mice. To visualize facial shape changes associated with positive and negative extremes of PC1 and PC2, skull CT reconstructions of mice occupying these positions are displayed. Note that although the complete skull is displayed for anatomical reference, only landmarks located on the colored facial region have been used for the analysis. Colors correspond to mice groupings. The red colored skull to the far right representing a TS mouse treated with a high dose of GTE-EGCG exhibited poor bone mineral density as evidenced by the nearly transparent regions of cranial vault bones. See Fig. S3 to further visualize facial shape changes associated to PC1, which is the axis that explains most morphological variation and separates groups based on genotype. Mice skull reconstructions were obtained from NRecon and Amira, and landmarking was performed with Amira.

Figure 2

Localized Euclidean Distance Matrix Analysis facial shape pairwise contrasts for each mouse group of untreated wildtype (WT) and trisomic (TS) mice and low and high dose GTE-EGCG-treated mice. Results from the baseline contrast (TS vs WT) can be compared to results from high and low GTE-EGCG dose experiment to assess the effect of the treatment on Ts65Dn facial skeleton development. Black solid lines represent linear facial measurements that are significantly different in the two compared groups. Numbers indicate the percentage of significantly different linear distances within each comparison. Mice skull reconstructions were obtained from NRecon and Amira.

Figure 3

Principal Component Analyses of the global facial shape variation in human children from different age groups. (a) Adolescents from 13 to 18 years old, and (b) Babies/toddlers to 3 years old. Scatterplots of PC1 and PC2 axes with the corresponding percentage of total morphological variance explained are displayed along each axis. Convex hulls represent the ranges of variation within each group of children. Anterior facial morphings associated with the extreme negative and positive values of PC1 are also shown. Orange solid lines represents facial phenotypes associated with Down syndrome, whereas blue solid line represents facial phenotype associated with euploid condition. Both shapes are compared to average facial shape, represented by a dashed grey line. Facial reconstructions were obtained from 3dMD and Agisoft PhotoScan, landmarks were recorded with PhotoModeler and morphings performed with Amira.

Figure 4

Localized Euclidean Distance Matrix Analysis facial shape pairwise contrasts and iterative bootstrapping tests of facial changes due to treatment for each age group. Black solid lines represent linear facial measurements that are significantly different in the two compared groups. First row: children with Down syndrome (DS) versus euploid children (EU); second row: children with Down syndrome supplemented with GTE-EGCG (DS + GTE-EGCG) versus euploid children (EU), third row: GTE-EGCG supplemented (DS + GTE-EGCG) versus untreated children with Down syndrome (DS), last row: iterative bootstrapping tests based on facial treatment scores (FTS). FTS is computed by contrasting the number of significant differences in facial traits between (1) euploid children and untreated children with Down syndrome and (2) euploid children and GTE-EGCG treated children with Down syndrome. Histograms represent the simulation results for each random group separately (top rows) as well as grouped together (bottom row). Each group contains an increasing number of GTE-EGCG treated Down syndrome cases. The red line shows the FTS score obtained with the complete set of observed cases, including all GTE-EGCG treated Down syndrome cases, in each group. P-values are provided for each group, and statistically significant comparisons are marked with **. Facial reconstructions were obtained from 3dMD and Agisoft PhotoScan.