Otitis media phenotype of Dp1Tyb mice. a,

Representation of Hsa21 (in grey), the orthologous regions on Mmu10, Mmu17 and Mmu16 (in light blue) and the region of the duplication in Dp1Tyb mice (in dark blue). b, Haematoxylin-eosin stained transverse sections through the middle ear of 3 week (3wk), 4 week (4wk), 8 week (8wk) and 16-week-old (16wk) wild-type (WT) and Dp1Tyb mice. MEC: middle ear cavity; TM: tympanic membrane; PG: polypoid growth. Scale bars: 1 mm. c, Comparison of the thickness of the epithelial lining of the middle ear for each genotype and age (3wk: WT n = 7 mice (3 female and 4 male), Dp1Tyb n = 7 mice (3 female and 4 male); 4wk: WT n = 5 mice (2 female and 3 male), Dp1Tyb n = 4 mice (2 female and 2 male); 8wk: WT n = 6 mice (3 female and 3 male), Dp1Tyb n = 6 mice (3 female and 3 male); 16wk: WT n = 4 mice (1 female and 3 male), Dp1Tyb n = 7 mice (3 female and 4 male). d, Broadband click stimuli ABR thresholds in the right ears of 2-month-old wild-type (WT) and Dp1Tyb mice. The graph shows elevated mean thresholds in Dp1Tyb mice (n = 6 mice, 3 female and 3 male) compared to wild-type (WT) mice (n = 6 mice, 3 female and 3 male). Bars: standard error of mean. P-values were determined using two-tailed t-test. ** p<0.01; *** p<0.001.

Normal ear morphology and craniofacial dysmorphology in Dp1Tyb mice.

a, Histological analysis of the hair cells in 2-months old mice from both genotypes. No difference was observed between the genotypes. MEC: middle ear cavity. Scale bars: 1mm and 100 μm. b, Scanning electron micrographs showing hair cell organisation and morphology in the mid turn of the cochlea of wild-type (WT) and Dp1Tyb mice. Three rows of OHCs and a row of IHCs were observed in both wild-type and mutant mice. Scale bars: 10 μm. OHC, outer hair cell; IHC, inner hair cell. c, Microscopic views of tympanic membrane from wild-type (WT) and Dp1Tyb mice. No visible fluid behind the tympanic membrane and easily recognizable malleus was detected in the wild-type ears and cloudy appearance of the eardrum was detected in all inspected Dp1Tyb middle ears. Both wild-type and mutant ears had intact tympanic membranes. d, μCT images of the ear of a wild-type (WT) and a Dp1Tyb mouse showing no visible difference in the size of the malleus (in red), incus (in blue) and stapes (in magenta). M, malleus; I, incus; S, stapes. e, Images of a Dp1Tyb mouse and a wild-type littermate showing a different shape of the skull and different shape and position of the pinna in the mutant mouse. The arrow indicates the pinna in the Dp1Tyb mouse. f, Dorsoventral view of a 2-month-old wild-type mouse skull showing the measurements used to study the craniofacial defect of the Dp1Tyb mice and graphic comparison of the ratios of the nasal bone length (NL), frontal bone length (FL) and the frontal bone width (FW) to the full dorsal length of the skull (SL) for each genotype. The graph shows significant difference in the width of the Dp1Tyb mice (n = 6 mice, 3 female and 3 male) compared to wild-type mice (WT) (n = 8 mice, 3 female and 5 male). Bars: standard error of mean. P-values were determined using two-tailed t-test ** p<0.01. g, Images of dissected ears from wild-type (WT) and Dp1Tyb mice. Arrows indicate the Eustachian tube measurements taken for this study. Comparison of the length and the width of the bony part of the Eustachian tubes in wild- type (WT) and Dp1Tyb mice. Bars: standard error of mean. P-values were determined using two-tailed t-test. No significant difference was detected.

OM phenotypes of Dp4Tyb, Dp5Tyb, Dp6Tyb and Ts1Rhr mice.

a, Representation of Hsa21 (in grey), the orthologous regions on Mmu10, Mmu17 and Mmu16 (in light blue) and the region of the duplication in Dp1Tyb, Dp9Tyb, Dp2Tyb, Dp3Tyb, Dp4Tyb, Dp5Tyb, Dp6Tyb and Ts1Rhr mice (in dark blue). b, Haematoxylin- eosin stained transverse sections through the middle ear of 2-month-old wild type (WT) and mutant mice showing middle ear inflammation in Dp5Tyb and Ts1Rhr mice. Note that the OM phenotype observed in the Ts1Rhr mouse displayed is unilateral in the left ear. MEC: middle ear cavity. Scale bars: 1 mm. c, Broadband click stimuli ABR thresholds in the right ears of 2-month-old wild-type (WT) and mutant (Dp5Tyb and Ts1Rhr) mice. The graph shows elevated mean thresholds in Dp5Tyb mice (n = 5 mice, 3 female and 2 male) compared to wild-type (WT) mice (n = 4 mice, 2 female and 2 male) and in Ts1Rhr mice (n = 5 mice, 2 female and 3 male) compared to wild-type (WT) mice (n = 5 mice, 1 female and 4 male). Bars: standard error of mean. P-values were determined using two-tailed t-test. * p≤0.05; ** p<0.01.

Localisation and verification of the major gene, Dyrk1a, involved in OME in DS.

a, Representation of Hsa21 (in grey), the orthologous regions on Mmu10, Mmu17 and Mmu16 (in light blue) and the region of the duplication in all DpTyb mutants (in dark blue). b, The pie charts represent percentage of mutant mice with no effusion (in yellow) and with effusion in one (in orange) or in both ears (in red) for all the DpTyb lines tested along with Ts1Rhr, Dp2Yey and Dp3Yey. c, Normalisation of Dyrk1a dosage in Dp3Tyb mice leads to restoration of the wild-type phenotype. The graphs show the incidence of OM in wild-type (WT), DpTyb lines, and double mutant mice, carrying a DpTyb allele and heterozygous for a KO allele at either the Dykr1a or Hmgn1 gene. First panel, mice produced from Dp1Tyb mice crossed to Dyrk1a knockout mice at the Francis Crick Institute - wild-type (WT), Dp1Tyb mice, and double mutant mice Dp1Tyb Dyrk1a+/-. Second panel, mice produced from Dp1Tyb mice crossed to Dyrk1a knockout mice at the MRC Harwell Institute (H) - wild-type (WT), Dp1Tyb mice, and double mutant mice Dp1Tyb Dyrk1a+/-. Third panel, mice produced from the Dp3Tyb mice crossed to Dyrk1a knockout mice in the Francis Crick Institute - wild-type (WT), Dp3Tyb mice, and double mutant mice Dp3Tyb Dyrk1a+/-. Fourth panel, mice produced from the Dp5Tyb mice crossed to Hmgn1 knockout mice at the Francis Crick Institute - wild-type (WT), Dp5Tyb mice, and double mutant mice Dp5Tyb Hmgn1+/-. P-values were determined using two-tailed t-test. * p≤0.05; ** p<0.01; *** p<0.001.

Expression of DYRK1A and associated pathway members.

a, Immunohistochemical staining of transverse sections through the middle ear of mutant and wild-type (WT) mice with antibodies against DYRK1A, phospho SMAD2 (pSMAD2), SMAD3, VEGF-A, IL-6 and IL-10. Scale bar = 25 µm. EL: epithelial lining; MEC: middle ear cavity; MEF: middle ear fluid. b, Bar graphs of the immunohistochemistry shown in a, represented as percentage of middle ear epithelial cells that express the protein. To quantify the results middle ear epithelial cells were counted in six ears from each genotype for each antibody except for pSMAD2 and SMAD3 for the Dp3Tyb, Dp3Tyb Dyrk1a+/- mice and wild-type littermates (WT) where the cells were counted in eight ears. Error bars show standard error of the mean (SEM). Significance was established using two-tailed unpaired t-tests. P-values were denoted according to **** < 0.0001, *** < 0.001, ** < 0.01, * < 0.05. c, Negative control (no primary antibody) for the immunohistochemistry shown in (a). d, Mesoscale Discovery (MSD) assay using blood from Dp5Tyb mice (n = 12 mice, 11 female and 1 male), blood from wild-type (WT) littermates (n = 12 mice, 8 female and 4 male) and middle ear fluid from Dp5Tyb mutants (n = 12 mice, 11 female and 1 male). The results presented in the graphs are from two independent experiments for all antibodies. Each point corresponds to a single sample which has been measured repeatedly and averaged (see data pre- processing in methods). For each group in each panel the height of the coloured bar indicates the mean and whiskers extend to +/- 2 SEM. For each pairwise comparison in a panel a two-sided Mann–Whitney–Wilcoxon (MWW) test was performed and raw p- values were denoted according to: **** < 0.0001, *** < 0.001, ** < 0.01, * < 0.05. To control for multiple testing, we applied the Benjamin-Hochberg procedure to the complete set of 24 p-values arising from all the MWW tests across all assays and pairs of groups. Rejecting the null in all starred cases (*, **, *** and ***) controlled the FDR below 5%.

Cilia loss in the middle ear cavities of Dp5Tyb mice and littermate controls.

Scanning electron microscopy (SEM) images from the dorsal and ventral region of the middle ear of a, 2-week-old and b, 2-month-old Dp5Tyb mice and wild-type (WT) littermates. Images were taken of both ears of each mouse (2-week-old Dp5Tyb n = 3 mice, 2 female and 1 male, WT n = 3 mice, 2 female and 1 male; 2-month-old Dp5Tyb n = 6 mice, 3 female and 3 male, WT n = 4 mice, 2 female and 2 male). Representative SEM images of middle ears were chosen for the panels. For the 2-week-old time point the dorsal and ventral images of wild-type ears and the dorsal images of the Dp5Tyb ears are from three separate mice. The 2-week-old ventral images of the Dp5Tyb ears are from two separate mice. For the 2-month-old time point the dorsal images of wild-type ears are from six different ears of five mice, the ventral images of wild-type mice - six different ears from three mice. Both, dorsal and ventral images of middle ear images of 2-month- old Dp5Tyb mice represented in the figure, at the time when the mice have fully developed OM, are from six different ears from five mice (for the dorsal) and from six different ears for four mice (for the ventral). Loss of cilia in the middle ear of epithelial cells was detected only in the ears of 2-month-old mutants (highlighted in the figure with red borders) but not in 2-week-old Dp5Tyb mice. A complete lawn of cilia is no longer present in the highlighted ventral images of the mutant mice. Images taken at x2500 magnification. Scale bar = 10 µm.