Trisomy 21 results in Down's syndrome, but little is known about how a 1.5-fold increase in gene dosage produces the pleiotropic phenotypes of Down's syndrome. Here we report that two genes, DSCR1 and DYRK1A , lie within the critical region of human chromosome 21 and act synergistically to prevent nuclear occupancy of NFATc transcription factors, which are regulators of vertebrate development. We use mathematical modelling to predict that autoregulation within the pathway accentuates the effects of trisomy of DSCR1 and DYRK1A, leading to failure to activate NFATc target genes under specific conditions. Our observations of calcineurin-and Nfatc-deficient mice, Dscr1- and Dyrk1a–overexpressing mice, mouse models of Down's syndrome and human trisomy 21 are consistent with these predictions. We suggest that the 1.5-fold increase in dosage of DSCR1 and DYRK1A cooperatively destabilizes a regulatory circuit, leading to reduced NFATc activity and many of the features of Down's syndrome. More generally, these observations suggest that the destabilization of regulatory circuits can underlie human disease.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
We thank S. L. Schreiber, W. Mobley and K. Tanda for discussion and comments on the manuscript; W. Becker for Dyrk1a expression constructs; E. Olson for the anti-DCSR1 (MCIP1) antibody; R. S. Williams for the Dscr1 (MCIP1) expression construct; K. Stankunas, G. Krampitz and C. Shang for help with histology on Dyrk1a and Dscr1 transgenic mice; E. Wang for mass spectrometric analysis; W. Mobley and K. Zhan for providing Ts65Dn mice; and F. Wang, members of the Crabtree laboratory, J. Lee, M. Dionne and S. Arron for discussions. We thank the Stanford Center for Innovation in In Vivo Imaging (NCI Small Animal Imaging Resource Program Grant), the Stanford Imaging Facility and the Stanford Proteomics and Integrated Research Facility. These studies were supported by the Howard Hughes Medical Institute and NIH grants to G.R.C., and by the Christopher Reeve Paralysis Foundation (I.A.G.). M.M.W. is supported by a Stanford Graduate Fellowship and an HHMI predoctoral fellowship, C.-P.C. by grants from AHA and the NIH, J.R.A. by a postdoctoral fellowship from the Berry Foundation, J.J.H. and S.K.K. by the ADA, H.W. by a Damon Runyon Cancer Research Foundation postdoctoral fellowship and a Muscular Dystrophy Association research development grant, and T.M. by KAKENHI from JSPS and MEXT and by a grant from JST BIRD. Author Contributions The order of listing of the authors J.R.A., M.M.W., A.P. and I.A.G. does in no way reflect their relative contribution to this work. I.A.G. and G.R.C. are responsible for the original concept. I.A.G. generated the Nfatc mutant mice (Fig. 1, Supplementary Figs 2–5 and Table 1), J.R.N. the Cnb1 mutant mice (Supplementary Fig. 8c), and H.W. and L.C. the Dyrk1a/Dscr1 transgenic mice (Fig. 3). I.A.G., M.M.W., C.-P.C., X.G., J.R.N., J.J.H., S.K.K., N.Y. and T.M. analysed mutant mice (Figs 1, 3 and Supplementary Figs 2–5 and Table 1). M.M.W. performed the skull morphometry studies (Fig. 1a–c and Supplementary Figs 2–4) and helped with the analysis of Ts1Cje and Ts65Dn mice (Supplementary Fig. 9). I.A.G. performed the neuron signalling experiments (Fig. 2a, b, e, f and Supplementary Figs 7, 8b), biochemical analysis of human Down's syndrome samples (Fig. 4c), calcineurin mutant mice (Supplementary Fig. 8c), Ts1Cje and Ts65Dn mice (Supplementary Fig. 9) and Dyrk1a/Dscr1-overexpressing mice (Fig. 3a) as well as the Nfatc4 promoter studies (Supplementary Fig. 8). A.P. generated and solved the mathematical model (Fig. 4a, b, d and Supplementary Discussion B). U.F. provided the clinical samples (Fig. 4c). J.R.A. conducted the in vitro kinase (Fig. 2c, d and Supplementary Fig. 7) and 293T (Supplementary Fig. 6) assays and DSCR1/DYRK1A quantifications (used in Supplementary Discussion B), made the anti-DYRK1A antiserum and helped H.W. and I.A.G. to genotype some of the Dyrk1a/Dscr1-overexpressing mice. G.R.C., I.A.G., J.A.A., M.M.W. and A.P. wrote the manuscript and I.A.G., M.M.W., A.P. and G.R.C. generated the figures.
Increased expression of DSCR1 and DYRK1a on chromosome 21 destabilizes the NFAT genetic circuit
Craniofacial phenotype of NFATc2/c4 DKO mice
Measurement of cranial dimensions
Measurement of mandible dimensions.
NFATc mutant mice have defects in placental vascularization, annular pancreas and agangionic megacolon.
a, DYRK1a is localized to the nucleus. b, DYRK1a and DSCR1 inhibit NFAT transcriptional activity in a dose-dependent manner. c, DYRK1a kinase activity prevents nuclear accumulation of NFATc1.
a, Diagram of NFATc4 constructs. b, Serine-to-alanine mutations of in the SRR- and SP-region of NFATc4 result in Ca2+-independent nuclear localization of EGFP-NFATc4. c, DYRK1a targets the third serine in the SP1 region, permitting processive phosphorylation of the second, then the first serine by GSK-3.
a, NFAT binding sites in the NFATc4 promoter. b, The NFATc4 promoter in cortical neurons is regulated by CaN/NFAT activity. c, positive feedback regulation of NFATc1 and c4.
Immunoblot of DYRK1a, HSP-90, NFATc4 and DSCR1 in whole cell extracts from E11.5 heads and E13.5 cerebral cortex from trisomic and control Ts1Cje embyros.
Discussion of the Down syndrome critical region and additional references.
A Mathematical Model of the NFAT Genetic Circuit
This file contains additional details of the methods used in this study.