Genetic dissection of triplicated Hsa21 orthologs produces differential skeletal phenotypes in Down syndrome mouse models

Abstract

Down syndrome (DS) phenotypes result from triplicated genes, but it is generally unknown how specific three copy human chromosome 21 (Hsa21) orthologous genes or interactions between genes affect these traits. A mouse mapping panel genetically dissecting Hsa21 syntenic regions was used to investigate the contributions and interactions triplicated Hsa21 orthologous genes on mouse chromosome 16 (Mmu16). Four-month-old femurs of male and female Dp9Tyb, Dp2Tyb, Dp3Tyb, Dp4Tyb, Dp5Tyb, Dp6Tyb, Ts1Rhr, and Dp1Tyb;Dyrk1a+/+/- mice were analyzed by micro-computed tomography and 3-point bending to assess skeletal structure and mechanical properties. Male and female Dp1Tyb mice, with the entire Hsa21 homologous region of Mmu16 in three copies, display specific bone deficits similar to humans with DS and were used as a baseline comparison for the other strains in the panel. Bone phenotypes varied based on triplicated gene content, sex, and bone compartment. Three copies of Dyrk1a played a sex-specific, essential role in trabecular deficits and may interact with other genes to influence cortical deficits related to DS. Triplicated genes in Dp9Tyb and Dp2Tyb mice improved some skeletal deficits. As triplicated genes may both improve and worsen bone deficits, it is important to understand the interaction between and molecular mechanisms of skeletal alterations affected by these genes.

Full text

© 2023. Published by The Company of Biologists Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution and reproduction in any medium
provided that the original work is properly attributed
Genetic dissection of triplicated Hsa21 orthologs produces
differential skeletal phenotypes in Down syndrome mouse models
Kourtney Sloan1, Jared Thomas1, Matthew Blackwell1, Deanna Voisard1, Eva Lana-Elola2,
Sheona Watson-Scales2, Daniel L. Roper3, Joseph M. Wallace4, Elizabeth M. C. Fisher5,
Victor L. J. Tybulewicz2 and Randall J. Roper1,*
1Department of Biology, Indiana University-Purdue University, Indianapolis, IN, USA
2The Francis Crick Institute, London, UK
3Data Analytics Computing, Lehi, UT, USA
4Department of Biomedical Engineering, Indiana University-Purdue University, Indianapolis, IN,
USA
5UCL Institute of Neurology, London, UK
*Corresponding Author: rjroper@iupui.edu
ORCHID ID: 0000-0002-9860-5037
Keywords: Down syndrome, Trisomy 21, Skeletal deficits, Animal models, Genetics
SUMMARY STATEMENT
A Down syndrome mouse mapping panel showed that triplicated Dyrk1a plays a sex-
specific role in some skeletal phenotypes. Triplicated genes may interact and cause improved or
worsened skeletal measurements.
ABSTRACT
Down syndrome (DS) phenotypes result from triplicated genes, but it is generally
unknown how specific three copy human chromosome 21 (Hsa21) orthologous genes or
interactions between genes affect these traits. A mouse mapping panel genetically dissecting
Hsa21 syntenic regions was used to investigate the contributions and interactions triplicated
Hsa21 orthologous genes on mouse chromosome 16 (Mmu16). Four-month-old femurs of male
and female Dp9Tyb, Dp2Tyb, Dp3Tyb, Dp4Tyb, Dp5Tyb, Dp6Tyb, Ts1Rhr, and
Dp1Tyb;Dyrk1a+/+/- mice were analyzed by micro-computed tomography and 3-point bending to
Disease Models & Mechanisms • DMM • Accepted manuscript

assess skeletal structure and mechanical properties. Male and female Dp1Tyb mice, with the
entire Hsa21 homologous region of Mmu16 in three copies, display specific bone deficits similar
to humans with DS and were used as a baseline comparison for the other strains in the
panel. Bone phenotypes varied based on triplicated gene content, sex, and bone compartment.
Three copies of Dyrk1a played a sex-specific, essential role in trabecular deficits and may
interact with other genes to influence cortical deficits related to DS. Triplicated genes in Dp9Tyb
and Dp2Tyb mice improved some skeletal deficits. As triplicated genes may both improve and
worsen bone deficits, it is important to understand the interaction between and molecular
mechanisms of skeletal alterations affected by these genes.
INTRODUCTION
The genotype-phenotype etiology of Down syndrome (DS), or Trisomy 21 (Ts21),
affecting ~1 in 800 live births (de Graaf et al., 2017), is not well understood. Trisomy of human
chromosome 21 (Hsa21) produces a dosage imbalance of the genes on Hsa21, which may result
in altered protein levels to cause deficits in cognitive, skeletal, cardiac, immune and other
systems seen in DS. Multiple non-mutually exclusive hypotheses describe the potential effect of
three copy Hsa21 genes on DS phenotypes: a single “effector” or “driver” triplicated gene may
play a major role in a particular phenotype and may affect other triplicated or non-triplicated
responder genes that cause particular DS phenotypes; multiple triplicated genes may interact to
cause a DS phenotype; or compensatory effects may mask the effect of three copy Hsa21 genes
and their role in a phenotype (Roper and Reeves, 2006, Antonarakis et al., 2020, Moyer et al.,
2021, Duchon et al., 2021). These hypothesized genetic mechanisms may be different for each
DS phenotype. The possibility that Ts21 disrupts normal gene expression also leads to questions
about the mechanistic manifestation of a particular phenotype or how three copies of ‘dosage-
sensitive’ genes may influence subphenotypes of a DS trait, including whether triplicated genes
may have both positive and negative effects on a phenotype or subphenotype (Moyer et al.,
2021). Insights into these questions will help understand the effects of genetic dosage imbalance
and potential correction of temporally and spatially altered gene expression to improve DS traits.
Significant differences in bone mineral density (BMD) have been observed in individuals
with DS beginning in their second and third decades of life. Individuals with DS attain peak bone
mass 5-10 years earlier than the general population, and it is lower than the general population
Disease Models & Mechanisms • DMM • Accepted manuscript

(Costa et al., 2018). Additionally, people with DS experience bone loss sooner and at a higher
rate than the general population (Carfi et al., 2017). Males with DS begin losing BMD in the
femur much earlier than females with DS, suggesting a protective effect of the Hsa21 trisomy in
the female biological sex in terms of maintaining BMD (Carfi et al., 2017, Costa et al., 2017,
Costa et al., 2018, Tang et al., 2019). Lowered BMD is thought to be the product of dysregulated
bone turnover and may be due to low bone formation, which has been shown in people with DS
through serum biomarkers (McKelvey et al., 2013).
Mouse models are commonly used to study DS phenotypes, and the availability of strains
with different segments of Hsa21 orthologous genes in three copies allows mapping of the
specific dosage-sensitive genes that cause particular phenotypic outcomes (Antonarakis et al.,
2020, Lana-Elola et al., 2011, Duchon et al., 2021, Lana-Elola et al., 2016). Genes orthologous
to Hsa21 are located on mouse chromosome (Mmu) 16, Mmu17, and Mmu10, with the greatest
number being located distally on Mmu16 (Davisson et al., 2001). DS skeletal phenotypes,
reflecting changes in both males and females, have been recapitulated in DS mouse models
including Ts65Dn (~50% of Hsa21 orthologous genes on a freely segregating extra
chromosome) and Dp1Tyb (duplication of 145 protein-coding genes and 23 Mb on Mmu16 that
is orthologous to Hsa21) (Blazek et al., 2011, Lana-Elola et al., 2021, Thomas et al., 2020,
Thomas and Roper, 2021, Lana-Elola et al., 2016).
Dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A), found in three
copies in individuals with DS and some DS mouse models including Ts65Dn and Dp1Tyb, is
hypothesized to affect multiple areas of development, including cognitive and skeletal systems.
Normalizing Dyrk1a copy number in Ts65Dn;Dyrk1a+/+/- male mice at 6 weeks of age improved
trabecular and cortical microarchitecture, mechanical, and cellular properties, indicating that
Dyrk1a copy number is important in skeletal properties of male mice (Blazek et al., 2015).
However, other triplicated genes in addition to Dyrk1a may also be important in skeletal
phenotypes and specifically in female DS mouse models.
Ts1Rhr mice have three copies of 31 protein-coding genes (including Dyrk1a) and 4.2
Mb of Mmu16 that is orthologous to Hsa21 (Olson et al., 2004, Olson et al., 2007). Limited
skeletal measurements found only small changes in bone as compared to control littermates, no
skeletal differences between male and female Ts1Rhr mice, and trabecular, cortical, and
mechanical parameters were never quantified in these mice (Olson and Mohan, 2011).
Disease Models & Mechanisms • DMM • Accepted manuscript

Tg(DYRK1A) mice with extra DYRK1A copies showed trabecular deficits in both male and
female mice but no deficits in cortical thickness (Ct.Th--other cortical properties were not
examined) (Lee et al., 2009). These results led to questions pertaining to the importance of
increased Dyrk1a copy number and sexual dimorphism in DS-related trabecular and cortical
bone deficits.
To facilitate the identification of important dosage-sensitive genes that lead to DS
phenotypes, including DS-related cardiac and locomotor deficits, a high-resolution mapping
panel of seven strains, with contiguous genetic segmental duplications covering various regions
of Mmu16 that correspond to Hsa21, was generated (Lana-Elola et al., 2016, Watson-Scales et
al., 2018). We utilized this mouse mapping panel to identify triplicated genes or regions that
were important in bone phenotypes. We hypothesized that mouse lines containing triplicated
Dyrk1a (Dp3Tyb and Dp5Tyb) would have more severe deficits in skeletal phenotypes as
compared to other triplicated lines, and that sexual dimorphisms would be seen in bone deficits.
Coupled with results from Ts1Rhr and Dp1Tyb;Dyrk1a+/+/- mice, we more accurately defined the
contribution of three copies of Dyrk1a to deficits in various bone compartments in as sex-
specific manner and identified potential interacting genes that lead to skeletal phenotypes
associated with DS.
RESULTS
Different segments of Hsa21 orthologous three copy genes alter trabecular microarchitecture in
diverse ways with three copies of Dyrk1a influencing trabecular bone in a sex-dependent manner
Male Dp1Tyb as compared to control littermate mice displayed reduced BMD, BV/TV,
and Tb.Th at 16 weeks (4 months) (Thomas et al., 2020). Female 16-week-old Dp1Tyb and
control mice had similar skeletal measurements in BMD, BV/TV, Tb.N, Tb.Th, and Tb.Sp.
There were sex effects in skeletal deficits as male 16-week-old Dp1Tyb and control mice
displayed higher BMD, BV/TV, and Tb.N and lower Tb.Sp as compared to female Dp1Tyb and
control mice at the same age.
To better comprehend how three copies of Hsa21 orthologous genes on Mmu16 affect
trabecular bone, 4-month-old Dp9Tyb, Dp2Tyb, and Dp3Tyb mouse lines that divide the
Dp1Tyb region of duplication of Mmu16 into three, non-overlapping segments, were analyzed
(Fig. 1). Male and female, triplicated and control mice were examined together within each line.
Disease Models & Mechanisms • DMM • Accepted manuscript

There was a significant effect of sex, with more “positive” effects on trabecular parameters in
male as compared to female mice (together) for the Dp9Tyb, Dp2Tyb, and Dp3Tyb mouse lines
(Fig. 2A-C, Table 1, and Figs S1-S4 A-C). Unexpectedly, there were significant genotype effects
in Dp9Tyb and Dp2Tyb lines as compared to controls with increased BMD, BV/TV, Tb.N,
Tb.Th, and decreased Tb.Sp for the Dp9Tyb as compared to the control mice, and increased
Tb.N and decreased Tb.Sp for the Dp2Tyb as compared to the control mice. Dp3Tyb mice did
not show any significant genotype effects for trabecular parameters.
Given previous work showing trisomic Dyrk1a’s influence on bone, it was surprising to
find no genotypic effect in the Dp3Tyb line. To better understand the contribution of triplicated
genes from the Dp3Tyb region on trabecular phenotypes associated with DS, we analyzed the
Dp4Tyb, Dp5Tyb, and Dp6Tyb lines (including littermate controls) that split the Dp3Tyb
duplication into three separate regions with only Dp5Tyb mice containing three copies of Dyrk1a
(Fig. 1). At 4 months, only sex effects were found in trabecular parameters for the Dp4Tyb line,
where male mice had increased BMD, BV/TV, Tb.Th, and Tb.N and decreased Tb.Sp compared
to female mice (Fig. 2D, Table 2, and Figs S1-S4 D). There were interactive effects for
trabecular traits between sex and genotype for the Dp5Tyb (BMD and BV/TV) and Dp6Tyb
(BMD, BV/TV, Tb.Th and Tb.N) lines when compared to their respective littermate controls
(Fig. 2E-F, Table 2, Figs S1-S4 E-F). Male Dp5Tyb and Dp6Tyb and littermate control mice
displayed greater BMD and BV/TV as compared to female Dp5Tyb and Dp6Tyb and littermate
control mice, respectively. Only male Dp5Tyb as compared to male control mice showed
decreased BMD in post hoc tests, suggesting that three copies of Dyrk1a (or another gene in the
Dp5Tyb region) is necessary to significantly alter BMD in male but not female mice (Table S1).
Furthermore, an interactive effect was observed for Tb.Th where female Dp6Tyb as
compared to control mice show an increase in Tb.Th in post hoc tests (Tables 2 and S1). There
was also a genotype effect with Dp5Tyb as compared to control mice displaying decreased
Tb.Th (Table 2). Taken together, these data suggest complex and sex-specific influences of
triplicated genes (possibly including Dyrk1a) affecting Tb.Th in the genetic regions covered by
Dp5Tyb and Dp6Tyb lines.
Additionally, Ts1Rhr mice (three copies of 31 protein-coding genes including Dyrk1a
with 8 fewer protein-coding genes than Dp3Tyb mice) displayed only sex differences in
trabecular parameters (Fig. 2G, Table 1, and Figs S1-S4 G). In the Ts1Rhr line, male mice as
Disease Models & Mechanisms • DMM • Accepted manuscript

compared to female mice displayed higher BMD and Tb.N and lower Tb.Sp. Finally, the
Dp1Tyb;Dyrk1a+/+/- line was generated to observe the effects of normalizing Dyrk1a copy
number from conception in an otherwise Dp1Tyb mouse. There were no genotypic effects in any
trabecular parameter in Dp1Tyb;Dyrk1a+/+/- mice compared to control (wildtype) animals (Fig.
2H, Table 2, and Figs S1-S4 H); this normalization of phenotypes in Dp1Tyb mice with only two
functional copes of Dyrk1a shows the importance of triplicated Dyrk1a in trabecular phenotypes.
Sex effects were observed in the Dp1Tyb;Dyrk1a+/+/- line with male mice displaying higher
BV/TV, BMD, and Tb.N and lower Tb.Sp as compared to female mice. Taken together, these
data suggest that three copies of Dyrk1a are necessary but not sufficient to decrease BV/TV,
BMD, and Tb.Th and interact with other triplicated genes from the Dp5Tyb and Dp6Tyb regions
to cause trabecular deficits in a sex-specific manner.
Cortical geometry parameters differ according to differential segments of Hsa21 orthologous
genes in three copies
At 16 weeks of age (4 months), Tt.Ar was larger in control male than all other mice, male
Dp1Tyb as compared to female Dp1Tyb mice, and female control as compared to female
Dp1Tyb mice (Thomas et al., 2020). At 16 weeks of age, male control mice had larger Tt.Ar,
Ct.Ar, Ps.Pm, and Ec.Pm compared to all other mice, with male mice as a group with higher
parameters compared to female mice. Marrow area (Ma.Ar) was greater in male control than all
other mice, and female control greater than female Dp1Tyb mice. Female as compared to male
mice as a group had reduced Ct.Th.
At 4 months of age, sex effects were observed in the Dp9Tyb line, where male as
compared to female mice displayed higher Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, Ec.Pm and lower TMD.
There were genotype effects with Dp9Tyb mice with their triplicated genes unexpectedly
showing increased Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, and Ec.Pm as compared to control mice (Figs 3
and 4A, Table 3, and Figs S5-S10 A). Dp2Tyb and control mice showed significant interactions
between sex and genotype with male Dp2Tyb and control as compared to female Dp2Tyb and
control mice displaying higher Tt.Ar, Ma.Ar, and Ec.Pm (Figs 3 and 4B, Table 3, and Figs S5-S9
B). Male control as compared to male Dp2Tyb mice showed increased Tt.Ar, Ma.Ar, and Ec.Pm
in post hoc analysis, suggesting that triplicated gene(s) in this region act in a sex-specific manner
to decrease these cortical parameters (Table S1). Additionally, sex effects were present in the
Disease Models & Mechanisms • DMM • Accepted manuscript

Dp2Tyb line where male mice had increased Ct.Ar, Ct.Th, Ps.Pm, and decreased TMD
compared to female mice (Fig. 3, Table 3, and Figs S6-S8, S10 B). Differing genotypic effects
were seen in Ct.Th and Ps.Pm, where Dp2Tyb mice had increased Ct.Th but decreased Ps.Pm
compared to control mice (Fig. 3, Table 3, and Figs S7-S8 B). In animals from the Dp3Tyb line,
there were sex effects where male mice displayed greater Tt.Ar, Ma.Ar, Ct.Ar, Ct.Th, Ps.Pm, and
Ec.Pm compared to female mice. Additionally, there were genotype effects with Dp3Tyb mice
showing significantly decreased Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, and Ec.Pm compared to control
mice (Figs 3 and 4C, Table 3, and Figs S5-S9 C). There was also an interactive effect of sex and
genotype on TMD with male control and Dp3Tyb mice increased compared to female Dp3Tyb
(Table 3, Fig. S10, and Table S1). These results suggest that different triplicated regions may
increase (Dp9Tyb) or decrease (Dp2Tyb and Dp3Tyb) cortical phenotypes.
To understand the impact of the triplicated genes comprising the Dp3Tyb region on
cortical phenotypes, Dp4Tyb, Dp5Tyb, Dp6Tyb, and control littermate mice were analyzed. Sex
effects were observed in the Dp4Tyb and Dp6Tyb lines with male mice exhibiting greater Tt.Ar,
Ma.Ar, Ct.Ar, Ps.Pm, Ec.Pm and reduced TMD as compared to female mice (Figs 3 and 4D and
4F, Table 4 and Figs S5-S10 D and F). Additionally, there was a genotypic effect in Ma.Ar in the
Dp4Tyb line where control mice exhibited greater Ma.Ar compared to Dp4Tyb. The Dp5Tyb
line (containing three copies of Dyrk1a) showed significant interactions in all cortical parameters
measured except Ct.Th and TMD, where only sex effects and genotype effects were observed.
Similar to trabecular bone, only male Dp5Tyb as compared to male control mice showed
decreased Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, and Ec.Pm in post hoc tests (Figs 3 & 4E, Table 4, Figs
S5-S6,S8-S9 E, and Table S1). For most parameters, both male Dp5Tyb and control mice were
significantly greater than female Dp5Tyb and control mice in post hoc tests, suggesting that three
copies of Dyrk1a or other triplicated genes in this region are necessary to significantly alter
Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, and Ec.Pm in male but not female mice (Table S1).
The Ts1Rhr line displayed similar cortical parameters as the Dp3Tyb line with the
exception of no significance in Ct.Ar nor Ct.Th and an opposite sex effect for TMD (Figs 3 and
4G, Table 3, and Figs S5-S10 G). Normalizing Dyrk1a copy number in Dp1Tyb;Dyrk1a+/+/- mice
did not normalize cortical parameters and still showed significant differences in sex and
genotype in Tt.Ar, Ma.Ar, Ct.Ar, Ps.Pm, and Ec.Pm (Figs 3 & 4H, Table 4, and Figs S5-S9 H).
Also, TMD was higher in female mice as compared to male mice in the Dp1Tyb;Dyrk1a+/+/- line
Disease Models & Mechanisms • DMM • Accepted manuscript

(Table 4 and Fig. S10H). Male mice with three copies of Dyrk1a (Dp3Tyb, Dp5Tyb, and
Ts1Rhr) showed significant percent decreases in Tt.Ar, Ma.Ar, Ps.Pm and Ec.Pm but generally
not to the same magnitude as male Dp1Tyb mice (Table S3). Female mice with the Dp3Tyb
region in three copies had significant percent decreases in Tt.Ar, Ma.Ar, Ps.Pm, and Ec.Pm
(Tables S3). Male mice with three copies of the Dp2Tyb region also show a decrease in the
percent reduction of multiple cortical parameters (Table S3). Taken together, these data indicate
that the effects of three copies of Dyrk1a are important in reducing many cortical parameters, but
there is an interactive effect of sex, with cortical parameters mostly affected in male mice. Other
triplicated genes inside and outside the Dp3Tyb region may interact with Dyrk1a to cause
differences in cortical parameters between triplicated and normal mice, and there may be
different interacting genes involved in male and female mice with three copies of Hsa21
homologs.
Whole bone property differences lessen in presence of a smaller number of Hsa21 orthologous
genes in three copies
Alterations in extrinsic mechanical properties, based on bone mass and cortical geometry,
were observed in 16-week-old (4-month-old) Dp1Tyb mice (Thomas et al., 2020). At 16 weeks
of age, there was a sex × genotype interaction for ultimate force, with control males greater than
all other mice and Dp1Tyb males greater than Dp1Tyb females. Control as compared to Dp1Tyb
mice had a higher ultimate displacement, stiffness, and total work; Dp1Tyb as compared to
control mice displayed higher values of displacement to yield and work to yield. Also, males as
compared to females had higher measurements for ultimate displacement, stiffness, and total
work. At 16 weeks, sex and especially genotype had non-interactive effects on overall bone
properties, and additional gene dosage and female sex were detrimental to whole bone
properties.
Comparisons were made between the Dp9Tyb, Dp2Tyb and Dp3Tyb lines that divide the
triplicated region of Dp1Tyb. In the Dp9Tyb line, there was a genotypic effect on ultimate force,
where Dp9Tyb mice could handle a greater force than control mice (Table 5). Sex effects were
also observed in the Dp9Tyb line, where female mice were increased in yield force, displacement
to yield, and work to yield compared to male mice, suggesting they have a greater elastic region.
In the Dp2Tyb line, there was a genotypic effect on ultimate displacement, with control mice
Disease Models & Mechanisms • DMM • Accepted manuscript

having greater displacement than Dp2Tyb mice (Table 5). Additionally, there were sex effects
with male mice showing greater ultimate force, ultimate displacement but lower yield force,
displacement to yield, and work to yield than female mice. The Dp3Tyb line showed sex effects
only, where male mice were increased in ultimate force and ultimate displacement but decreased
displacement to yield and work to yield compared to female mice (Table 5). Dp3Tyb mice also
have a sex effect in ultimate displacement with male increased compared to female mice.
To understand the potential impact of triplicated genes in the Dp3Tyb region on extrinsic
properties in bone, we examined the Dp4Tyb, Dp5Tyb, and Dp6Tyb lines that split the Dp3Tyb
region (Table 6). The Dp4Tyb and Dp5Tyb lines showed no significant differences in any
extrinsic parameter. The Dp6Tyb line showed only sex effects with female mice displaying
greater displacement to yield and work to yield than male mice, and male mice displaying greater
ultimate displacement, stiffness, and total work than female mice.
Ts1Rhr mice showed sex effects in ultimate displacement and total work, where male
mice were increased compared to female mice (Table 5). Dp1Tyb;Dyrk1a+/+/- mice corrected all
extrinsic differences seen in Dp1Tyb mice except for sex and genotype effects in ultimate force
(Table 6).
Material property improvements lessen with smaller regions of Hsa21 orthologous gene
triplication
Improvements in intrinsic mechanical properties, also called material properties, were
noted in 16-week-old (4-month-old) Dp1Tyb mice (Thomas et al., 2020). At 16 weeks, there was
a sex × genotype interaction in modulus with control males showing a lower modulus than all
other mice and both female and male Dp1Tyb mice had a higher modulus than their control
littermates
Additionally, bone from Dp1Tyb as compared to control mice displayed a higher yield stress,
ultimate stress, and resilience. Control as compared to Dp1Tyb mice had a higher ultimate strain.
The triplicated segments dividing the Dp1Tyb region all showed a strong sex effect for
material properties (Table 7). The Dp9Tyb line displayed sex effects in all intrinsic parameters,
where female mice were greater than male mice in all but ultimate strain. The Dp2Tyb line
showed interactions in yield stress, ultimate stress, and resilience, with male Dp2Tyb and female
Dp2Tyb and control mice significantly greater than male control mice in these parameters as
Disease Models & Mechanisms • DMM • Accepted manuscript

demonstrated by a post hoc test (Table S1). Additionally, there were differing genotypic effects
in ultimate strain and modulus and sex effects in strain to yield, ultimate strain, and modulus
(Table 7). Control mice were greater than Dp2Tyb mice in ultimate strain, but Dp2Tyb mice had
greater modulus. Female mice had greater strain to yield and modulus compared to male mice
but lower ultimate strain. The Dp3Tyb line showed female mice had greater yield stress, ultimate
stress, strain to yield, modulus, and resilience compared to male mice, and male mice had greater
ultimate strain compared to female mice. A genotypic effect was observed in yield stress, where
Dp3Tyb mice were greater than control mice.
To better understand the influence of triplicated genes in Dp3Tyb mice on intrinsic bone
parameters, the Dp4Tyb, Dp5Tyb, and Dp6Tyb lines were investigated (Table 8). There were
only minimal effects in the Dp4Tyb line with female as compared to male mice exhibiting a
higher ultimate stress. In the Dp5Tyb line, female mice showed a greater yield stress, ultimate
stress, and modulus as compared to male mice. A genotypic effect was observed in ultimate
strain, where control mice were greater than Dp5Tyb mice. In the Dp6Tyb line, there was a sex ×
genotype interaction for modulus with female Dp6Tyb mice showing greater modulus than
female control, male control, and Dp6Tyb mice in post hoc tests (Table S1). Additionally, female
mice as compared to male mice showed a greater yield stress, ultimate stress, and resilience
while male mice had a larger ultimate strain than female mice.
Ts1Rhr mice showed a genotypic effect in ultimate strain, where they were lower than
control mice (Table 7). Additionally, the Ts1Rhr line showed sex effects in yield and ultimate
stress, ultimate strain, and modulus, where female mice were greater than male mice in all but
ultimate strain. Dp1Tyb;Dyrk1a+/+/- mice appeared to still have genotype effects in yield stress,
ultimate stress, ultimate strain and modulus, where Dp1Tyb;Dyrk1a+/+/- mice were increased
compared to control in all but ultimate strain (Table 8).
DISCUSSION
Triplicated Hsa21 orthologous genes interact to cause skeletal phenotypes and when
reduced in number may have different effects on bone.
The manifestation of skeletal defects associated with DS appears to have genetic and
sexually dimorphic components. Although three copies of Hsa21 homologous, Mmu16 genes
together produce trabecular, cortical, and mechanical bone deficits in the Dp1Tyb DS mouse
Disease Models & Mechanisms • DMM • Accepted manuscript

model in a sex-specific manner (Thomas et al., 2020), the contribution of triplicated genes other
than Dyrk1a has not been elucidated. The analysis of skeletal phenotypes from the mouse
mapping panel comprising different sets of triplicated genes demonstrated that although Dyrk1a
may influence some bone parameters, other triplicated genes also affect the incidence and
severity of DS-related bone phenotypes. Similar to data from the Dp1Tyb line (Thomas et al.,
2020), male mice generally had increased trabecular and cortical skeletal measurements
compared to female mice. Additionally, there appear to be sex effects that interact with
triplicated genes that may portend the sexually dimorphic appearance and severity of skeletal
deficits in humans with DS.
For trabecular bone, the triplicated genes included in the Dp9Tyb and Dp2Tyb mouse
lines improved many trabecular parameters compared to euploid littermates, indicating a positive
effect of triplicated genes in both regions. Other genotypic effects of three copy Hsa21
orthologous genes on trabecular phenotypes were observed when the Dp5Tyb and Dp6Tyb
segments were isolated from the rest of the Hsa21 homologous genes. In post hoc analyses, only
male and not female Dp5Tyb as compared to control mice had reduced BMD.
For cortical bone, Dp9Tyb males and females exhibited generally increased cortical
parameters in addition to their increased trabecular parameters. In Dp2Tyb mice, there were
genotype × sex interactions in Tt.Ar, Ma.Ar, and Ec.Pm. In contrast to increased trabecular
parameters, Dp2Tyb as compared to control mice showed decreased cortical measurements.
Dp3Tyb, Ts1Rhr, and Dp5Tyb mice all showed generally decreased cortical parameters.
For extrinsic bone properties, some measurements were better in female mice and others
were better in male mice. Female mice showed better intrinsic properties, except for ultimate
strain, in most of the lines. Most interactive effects in intrinsic measurements were seen in the
Dp2Tyb line. The Ts1Rhr line seemed to have extrinsic and intrinsic skeletal phenotypes that
bridged differences between the Dp3Tyb and Dp5Tyb lines, suggesting effects from one or more
of the 19 additional triplicated protein-coding genes in the Ts1Rhr but not the Dp5Tyb regions.
The varying changes in trabecular microarchitecture, cortical geometry, and mechanical
properties between the mouse mapping panel lines suggests the presence of interactive, additive,
and compensatory effects with Hsa21 orthologous genes in three copies. However, this study
only accounts for the contribution of Hsa21 homologs from Mmu16, while the contribution of
trisomic homologs from Mmu10 and Mmu17 are unknown and could alter these results. The
Disease Models & Mechanisms • DMM • Accepted manuscript

interaction of triplicated Dyrk1a with other three copy genes found on Mmu16 has been analyzed
for hippocampus-dependent memory processes (Duchon et al., 2021). Increased dosage of
Dyrk1a was found to be important in working memory deficits and increased activity but was
modified by other genes on Mmu16 to suppress changes in activity. Genetic interactions of at
least two causative and two modifying loci were found to influence recognition memory. Similar
to our skeletal phenotypes, there were also triplicated Mmu16 loci analyzed by a panel of
triplicated DS mouse models that did not seem to be causative of behavioral deficits (Duchon et
al., 2021). In the hippocampal expression analyses, 38-57% of the triplicated genes in a panel of
triplicated DS mouse models exhibited dosage compensation and were not significantly
differentially expressed.
Similar additive and compensatory dosage effects have been seen in triplicated Hsa21
orthologous gene dosage experiments in zebrafish (Edie et al., 2018). When combinations of
mRNA from Hsa21 homologs were injected into zebrafish, some combinations produced
additive effects, where the additional mRNA dosage resulted in a significant increase in
phenotypic penetrance, while others represented a partial compensatory effect, where the
additional mRNA dosage resulted in a significant decrease in phenotypic penetrance. Data from
mice, zebrafish, and our current data suggest that interaction between triplicated Hsa21
orthologous genes and the effect of sex have a complex influence on DS-associated phenotypes.
Three copies of Dyrk1a cause DS-related skeletal differences specific to different bone
compartments
Preclinical genetic studies have hypothesized the influence of three copies of Dyrk1a on
skeletal deficits in DS mouse models. For trabecular measures at 4 months of age, three copies of
Dyrk1a appear to play a significant role, with a major effect in trabecular bone deficits in male
mice. In 16-week-old male Dp1Tyb as compared to control mice, there were significant
reductions in BMD, BV/TV, and Tb.Th. Similar reductions were seen in male Dp5Tyb mice that
also have three copies of Dyrk1a and 11 other genes, though these measurements appeared
reduced in magnitude in Dp5Tyb male mice compared to Dp1Tyb mice (Table S2). Combining
the data from the mouse mapping panel with the Dp1Tyb;Dyrk1a+/+/- line suggested a direct
influence of three copies of Dyrk1a and interaction with other genes in the Dp3Tyb region on
Disease Models & Mechanisms • DMM • Accepted manuscript

decreasing Tb.Th in triplicated mice, although triplicated gene(s) in the Dp6Tyb region may
result in increased Tb.Th in female Dp6Tyb as compared to control mice.
Both male Dp3Tyb and Ts1Rhr mouse models have three copies of Dyrk1a but did not
show significant differences in trabecular phenotypes at 4 months. It is possible that the 19
protein-coding genes in three copies found on the Ts1Rhr but not the Dp5Tyb line modify
trabecular phenotypes. Additionally, the normalization of trabecular phenotypes in
Dp1Tyb;Dyrk1a+/+/- mice indicated that three copies of Dyrk1a also influence manifestation of
these trabecular deficits in male triplicated mice. However, Dyrk1a is likely to interact with a
gene(s) in the Dp6Tyb region to fully manifest this deficit in DS mice.
Three copies of Dyrk1a may also be involved in some cortical deficits, and may interact
with other triplicated Hsa21 orthologous genes, but are not sufficient to cause these phenotypes
at 4 months of age. Dp1Tyb, Dp3Tyb, Ts1Rhr, and Dp5Tyb (all with three copies of Dyrk1a) as
compared to control mice all have significant reductions in Tt.Ar, Ma.Ar, Ps.Pm, and Ec.Pm.
None of the affected cortical measures, however, were corrected when Dyrk1a copy number was
normalized in Dp1Tyb;Dyrk1a+/+/- mice. Additionally, the percent difference in cortical measures
between mice with triplicated regions and control littermates within each strain, especially in
male mice, decreases from Dp1Tyb as compared to Dp3Tyb, Ts1Rhr, and Dp5Tyb lines (Table
S3). Taken together, these data indicate Dyrk1a or other genes in the Dp5Tyb region may be
necessary but not sufficient for cortical phenotypes, but if Dyrk1a is causal, other triplicated
Hsa21 orthologous genes are involved in DS-associated cortical phenotypes as well. It is likely
that interacting and compensatory triplicated Hsa21 orthologous genes affect cortical
phenotypes.
As to potential gene interactions and mechanisms, DYRK1A phosphorylates APP and
RCAN1, also known as DSCR1 (Ryoo et al., 2008, Jung et al., 2011). App is found in three
copies in Dp1Tyb and Dp9Tyb mice, and previous studies indicate changes in App dosage
negatively impacts bone. Male APP-/- mice displayed decreased BV/TV and Tb.Th and increased
Tb.Sp at 2 months old which was attributed to increased bone resorption and decreased bone
formation (Pan et al., 2018). Tg2567 mice overexpress a mutated human APP gene with the
Swedish mutation (APPswe), resulting in increased levels of amyloid-beta and amyloid plaques
(Hsiao et al., 1996). At two months, male Tg2567 mice displayed decreased BV/TV thought to
be due to increased osteoclast activation (Cui et al., 2011). In further study of APPswe in mature
Disease Models & Mechanisms • DMM • Accepted manuscript

osteoblasts, decreased BV/TV, Tb.N, and Tb.Sp were observed in TgAPPswe-Ocn mice at 5
months old, likely due to increased osteoclast formation and decreased osteoblastogenesis (Xia et
al., 2013). Rcan1 is triplicated in Dp2Tyb mice and in vitro overexpression of RCAN1 in bone
marrow-derived macrophage-like cells resulted in less TRAP+ multinucleated osteoclasts,
indicating RCAN1 may have a negative effect on osteoclastogenesis (Kim et al., 2016). In
primary calvarial osteoblasts, there was significantly increased ALP activity and bone nodule
formation, indicating RCAN1 may have a positive effect on osteoblast function (Kim et al.,
2016). While the effects of Rcan1 overexpression alone have not been observed in relation to
bone phenotypes in vivo, we suspect a similar phenomenon would occur, potentially resulting in
increased BMD and other skeletal parameters, and could influence the skeletal differences seen
in the Dp2Tyb line.
Current study compared to previous bone reports in Ts1Rhr and Tg(DYRK1A) mice
Previous reports measuring BMD in 16-week-old Ts1Rhr and control mice found no
significant differences in BMD, BMC, or bone area in areal measures, and no differences in
regional femur BMD, a small decrease in regional tibia BMD, and a small increase in regional
lumbar spine BMD as quantified by dual-energy X-ray absorptiometry (DXA). These results led
the authors to conclude that 31 protein-coding genes in three copies was insufficient to cause a
bone density phenotype (Olson and Mohan, 2011). There were also no significant differences
between sex seen in the previous study, and no further bone quantification of Ts1Rhr mice was
done. In the present study, we examined male and female Ts1Rhr mice on a similar genetic
background as previously reported (C57BL/6) and also found no differences in femoral BMD in
male or female mice between Ts1Rhr and euploid littermate mice at 4 months, although male
mice had increased BMD than female mice. However, when microCT analyses of geometry and
microarchitecture were performed on the cortical and trabecular bone respectively, we found that
Ts1Rhr male and female compared to control mice at 4 months had significantly reduced cortical
bone measurements similar to those observed in Dp3Tyb mice. We also observed sex effects in
trabecular BMD, BV/TV, Tb.N, and Tb.Sp. Taken together, these data indicate that
measurements of BMD, BMC, and bone area assessed via DXA may not accurately show all the
parameters that affect bone in DS mouse models or individuals with DS, and more intricate
measures of quantification, such as microCT, are necessary to find skeletal differences. By
Disease Models & Mechanisms • DMM • Accepted manuscript

extension, individuals with DS may have skeletal deficits or may be developing bone deficits
that may not be reflected at early ages or by measurements of BMD. Only through precise
examination of bone structure were these developmental differences, and differences between
males and females, teased out.
Previous reports of skeletal abnormalities in Tg(DYRK1A) mice showed differences in
BV/TV, Tb.N and Tb.Sp in male and female mice and Tb.Th in just male mice (Lee et al., 2009).
The addition of an extra copy (copies) of DYRK1A with these results corroborates our finding of
the effect of increased Dyrk1a copy number on trabecular phenotypes for male mice. While we
did not see trabecular differences in female Dp5Tyb mice, it may be that other genes in the
Dp5Tyb region are modulating the effects of three copies of Dyrk1a.This could also be true for
Tb.N and Tb.Sp in male mice, where we also do not see any significant differences in Dp5Tyb.
The only cortical phenotype measured in the aforementioned study was Ct.Th, and no
differences were seen. Dp1Tyb;Dyrk1a+/+/- mice did not correct Ct.Th, and Ct.Th is also the least
well correlated phenotype among other cortical phenotypes (Figure S11). We observed other
cortical phenotypes affected by an extra copy of Dyrk1a or other genes triplicated in the Dp5Tyb
region, and similar cortical deficits may have been found in Tg(DYRK1A) mice had these
analyses been done.
Hsa21 orthologous genes in three copies may cause improvement in skeletal measurements
Three copies of some genes orthologous to Hsa21 may also improve trabecular and
cortical bone phenotypes (Figs 2-4 and Tables 1 and 3). Dp9Tyb as compared to control mice,
with 74 protein-coding genes in three copies, displayed improved trabecular and cortical bone
measures. Dosage imbalance of genes in the Dp9Tyb region that could contribute to these
improved phenotypes include Btg3, Usp16, Bach1, Tiam1, and Sod1. ANA, also known as Btg3,
deficiency was shown to enhance ectopic bone formation (Miyai et al., 2009). Deletion of Usp16
was shown to lead to decreased mature and progenitor hematopoietic stem cell populations (Gu
et al., 2016). Bach1 inhibition suppressed osteoclastogenesis and upregulated ALP activity and
osteoblast mineralization (Wada et al., 2020, Hama et al., 2012, Tian et al., 2019). A knockdown
of Tiam1 was shown to enhance ALP activity (Onishi et al., 2013). Sod1 knockout in mice had
effects on bone formation and showed reduced BMD and stiffness, which may be a result of
excessive reactive oxidative species (ROS) (Zhang et al., 2021, Morikawa et al., 2013, Smietana
Disease Models & Mechanisms • DMM • Accepted manuscript

et al., 2010). The aforementioned studies reduced expression of these genes instead of
triplicating the expression, and results may vary with differences in the genetic dosage
imbalance. For example, overexpression of Sod1 may improve bone phenotypes through
increased elimination of ROS.
Dp2Tyb as compared to control littermate mice display increased Tb.N and Ct.Th and
reduced Tb.Sp, Tt.Ar, Ma.Ar, Ps.Pm, and Ec.Pm. These changes may be due to three copies of
Runx1 and/or Rcan1. Runx1 overexpression has rescued bone loss in ovariectomized mice and
enhances osteogenic differentiation (Ji et al., 2017, Luo et al., 2019, Tang et al., 2021). As
described earlier, increased expression of Rcan1 could increase osteoblast function while
decreasing osteoclastogenesis leading to increased bone mass. Additionally, increased copy
number of four interferon receptor genes Ifnar1, Ifnar2, Ingr2, and Il10rb may disrupt osteoclast
function as interferon signaling components are involved in bone homeostasis (Place et al.,
2021). It will be interesting to examine if the overexpression of these genes improves trabecular
deficits and if associated mechanisms could be used to improve diverse types of bone
abnormalities.
Female Dp6Tyb mice appear to have increased Tb.Th compared to control mice. In the
Dp6Tyb region, Pcp4 is naturally overexpressed during the differentiation of osteoblasts, and its
induced overexpression in vitro was shown to enhance Alizarin red staining intensity, alkaline
phosphatase activity, calcium deposition, and expression of osteocalcin and bone sialoprotein, all
indications that osteoblast function is increased (Meng et al., 2020, Xiao et al., 2008). Triplicated
Dyrk1a could interact with Pcp4 or other triplicated genes in the Dp6Tyb region to exacerbate
trabecular deficits.
Sexual dimorphism in triplicated DS model mice
Almost all trabecular and cortical skeletal parameters were increased in male as
compared to female mice, except for TMD. When there was a sex × genotype interaction, male
mice with a triplicated region were almost always greater than female mice within the same line.
The effects of sex on extrinsic parameters were mixed and female mice had better intrinsic
measurements except for ultimate strain.
As with the comparisons between male and female mice that include three copies of
Dyrk1a, there appears to be certain bone phenotypes where triplicated Hsa21 orthologous genes
Disease Models & Mechanisms • DMM • Accepted manuscript

only affect certain skeletal traits in male or female mice. Overall, it appears that female mice at 4
months of age with any of the Hsa21 orthologs have similar or less affected skeletal phenotypes
as compared to euploid littermate mice, whereas male mice with three copies of genes appear to
be more affected. Using BV/TV and Tt.Ar as a representation of trabecular and cortical
phenotypes, respectively, in males, Dp5Tyb skeletal deficits most resemble Dp1Tyb mice, while
in females, Dp3Tyb skeletal deficits most resemble Dp1Tyb mice at 4 months (Fig. 5). This
sexual dimorphism seen in DS mouse models phenocopies that in humans with DS, where males
seem to have more differences in skeletal phenotypes than females compared to the general
population at an age of peak bone mass (Thomas et al., 2020, Thomas and Roper, 2021). If
analyzed at a later time point (post-peak bone mass), skeletal deficits, like decreased BMD, may
appear in female mice, as seen in female humans with DS (Carfi et al., 2017).
Conclusion
It has been assumed that three copies of genes orthologous to Hsa21 on Mmu16 in mice
play an important role in the manifestation of trabecular, cortical, and mechanical bone property
phenotypes. Our previous work, mostly using male mice, implicated triplicated Dyrk1a in
architectural and mechanical bone phenotypes. By using the mouse mapping panel, we show that
three copies of Dyrk1a are essential for some skeletal deficits in mice but may interact with other
triplicated Hsa21 orthologous genes to cause trabecular, cortical, and mechanical deficits in DS
mouse models. Understanding how these triplicated Hsa21 orthologous genes interact on the
molecular level is an essential next step for understanding and improving skeletal deficits
associated with DS.
MATERIALS AND METHODS
Animals
The following mouse strains (Mus musculus) were used: Dp(16Lipi-Zbtb21)1TybEmcf
(Dp1Tyb), Dp(16Mis18a-Runx1)2TybEmcf (Dp2Tyb), Dp(16Mir802-Zbtb21)3TybEmcf
(Dp3Tyb), Dp(16Mir802-Dscr3)4TybEmcf (Dp4Tyb), Dp(16Dyrk1a-B3galt5)5Tyb (Dp5Tyb),
Dp(16Igsf5-Zbtb21)6TybEmcf (Dp6Tyb), Dp(16Lipi-Hunk)9TybEmcf (Dp9Tyb), and
Dp(16Cbr1-Fam3b)1Rhr (Ts1Rhr or Dp1Rhr), all which have been previously reported (Lana-
Elola et al., 2016, Olson et al., 2004). To generate Dp1Tyb;Dyrk1a+/+/-, Dp1Tyb animals were
Disease Models & Mechanisms • DMM • Accepted manuscript

crossed with mice carrying a loss-of-function allele of Dyrk1a (Dyrk1a+/-) (Fotaki et al., 2002).
Breeding of the resulting double mutant resulted in an occasional recombination which brought
both alleles onto the same chromosome. This Dp1Tyb;Dyrk1a+/+/- double mutant was then bred
against wildtype C57BL/6J mice, giving 50% Dp1Tyb;Dyrk1a+/+/- double mutant and 50%
wildtype offspring. The latter were used as controls for the double mutant mice. All mice were
backcrossed to the C57BL/6J background for at least 10 generations and were maintained on this
background by crossing heterozygous mutants to C56BL/6J wildtype mice. Mice were bred and
maintained in SPF conditions at the MRC Harwell Institute using Rat and Mouse No.3 breeding
chow (Special Diets Services, WUK) and given water ad libitum. Mice were housed in cages of
3-5 animals and randomized at weaning into cages of mixed genotype, so each cage had both
mutants and wildtypes across several litters. Male and female mice were used for all strains at
16-18 weeks (4 months) of age. Genotyping was carried out by Transnetyx (Cordova, TN, USA)
using quantitative real-time PCR (qPCR). For each mutant allele, a custom qPCR assay was
established using a forward and reverse primer and a reporter probe (Table S5). Protein-coding
genes were defined by the GRCm39 mouse genome assembly; for this reason, gene numbers
have changed from previous estimates. Right femurs were dissected from male and female mice
at 4 months of age, wrapped in gauze and frozen in phosphate-buffered saline (PBS) for
shipment and stored at -20ºC or -80ºC until needed for microcomputed tomography analysis or
mechanical testing. All regulated procedures were carried out with approval from a Local Ethical
Review Panel and under authority of a Project Licence granted by the UK Home Office, and in
accordance with EU Directive 2010/63/EU. Sample sizes were based on effect sizes of previous
results (Thomas et al., 2020) and sample numbers are listed in Tables 1-4.
Bone extraction and microcomputed tomography (µCT) analysis
Bone analysis was performed as described in (Stringer et al., 2017). Briefly, femurs were
scanned using a high-resolution µCT system (SkyScan 1172, Bruker microCT, Belgium) that
was calibrated using two cylindrical hydroxyapatite phantoms (0.25 and 0.75 g/cm3 CaHA) prior
to each scanning session. Hydration of the femurs was maintained while scanning by wrapping
them in parafilm. Femurs were scanned from the distal condyle to the third trochanter using
60kV, 12µm resolution, 885ms integration time, Al 0.5mm filter and an angular increment of 0.7
degrees. Post-scan, the bones were wrapped in PBS-soaked gauze and stored at -20ºC or -80ºC
Disease Models & Mechanisms • DMM • Accepted manuscript

until mechanical testing. Scans were reconstructed and rotated using NRecon and Dataviewer
(SkyScan, Bruker microCT, Belgium). Reconstructed and rotated bones were then analyzed
using CT analyzer (CTan) (SkyScan, Bruker microCT, Belgium) and MatLab (MathWorks, Inc.
Natick, MA). Trabecular region of interest (ROI) was defined beginning at the end of the distal
growth plate, extending 10% of the total bone length, and isolated from the cortical bone using a
custom MatLab code that creates an irregular anatomic ROI ten pixels away from the
endocortical perimeter. CTan’s batch analyzer was used to obtain bone mineral density (BMD),
bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and
trabecular number (Tb.N) for trabecular ROI. Cortical ROI was calculated as a region of seven
transverse slices at 60% of total bone length away from the end of the distal growth plate.
Geometric properties, including total (cortical) cross-sectional area (Tt.Ar), marrow area
(Ma.Ar), cortical area (Ct.Ar), cortical thickness (Ct.Th), periosteal perimeter (Ps.Pm),
endocortical perimeter (Ec.Pm) and cortical tissue mineral density (TMD), were obtained using a
custom MatLab code (Berman et al., 2015, Stringer et al., 2017). The threshold used for
segmentation of the trabecular bone was 55 to 255 and 70 to 255 for cortical bone. In addition,
one analyzer performed rotation and ROI determination for the entirety of one mouse strain to
limit variability between analyzers.
Mechanical testing
Mechanical properties were determined as described previously (Thomas et al., 2020).
Briefly, 3-point bending was performed on a mechanical testing machine while the bones were
fully hydrated with PBS (TA ElectroForce 3200, Eden Prairie, MN, USA). Bones were tested
with a 7 mm support span in anterior-posterior direction with the posterior surface in
compression. The loading point was placed directly at the midshaft location, the bone was
preloaded to establish contact, and then testing occurred at 0.025 mm/sec to failure. The yield
point was determined using the slope of the stress-strain curve, then implementing the 0.2%
offset method. The ultimate point was determined as the maximum force recorded. The failure
point was determined as when the bone broke. Whole bone (extrinsic) properties included yield
and ultimate force and displacement, yield and total work, and stiffness based on the force-
displacement curve. Material (intrinsic) properties included yield and ultimate stress and strain,
Disease Models & Mechanisms • DMM • Accepted manuscript

modulus, resilience, and toughness based on the stress-strain curve. Cortical geometry was used
to normalize the stress-strain curve from the force-displacement curve.
Correlational analysis
Data Analytics Computing measured the correlation between cortical features in each
category of data: male wildtype, male triplicated, female wildtype, and female triplicated, in each
test group. The graphs show the correlation values with a range of -1 to 1 and visually represent
these numbers with circles so that the relationship between values is immediately apparent.
Larger circles represent greater correlation between the features. Smaller dots represent
decreasing linear relationship. The darker blue color in both circles and values indicated a greater
degree of linear dependence. While circles and values of darker red correspond with decreasing
linear correlation. The smallest dots and values closest to 0 with a hue of grey approach feature
independence.
In addition, we looked at the bidirectional correlation between male wildtype and male
triplicated features, as well as female wildtype and female triplicated features. The correlation
was plotted with only circles (size representing magnitude and color indicating the increasing,
decreasing, or independent correlation) to get a clear understanding of directional correlation—
positive or negative—or independence of the feature of one genotype to the features in the other
genotype.
Statistical Analysis
Normality of the datasets were assessed via Shapiro-Wilk test using an alpha of 0.05.
When datasets violated Gaussian distribution, the dataset was transformed to their logarithmic
form and normality was assessed again via Shapiro-Wilk test. The transformed datasets are
annotated with a superscript “a” in the tables. A total of three outliers, determined by the ROUT
method (Q = 1%; GraphPad Prism), were excluded from the data analysis: one male Ts1Rhr was
excluded from the trabecular and cortical datasets, a different male Ts1Rhr was excluded from
the intrinsic and extrinsic (mechanical) datasets, and one male Dp5Tyb was excluded from the
mechanical datasets. Two-way ANOVA was performed for all parameters (trabecular, cortical,
and mechanical) to examine the potential effects of sex and genotype and their potential
interaction separately in each independently generated strain. If a significant sex effect occurred,
Disease Models & Mechanisms • DMM • Accepted manuscript

the average for all one sex (e.g., male), regardless of genotype, was compared to the average for
all the other sex (e.g., female). If a significant genotype effect occurred, the average for all one
genotype (e.g., control mice), regardless of sex, was compared to the average for all the other
genotype (e.g., triplicated mice). If a significant interaction occurred, Tukey’s post hoc analysis
was performed for the two-way ANOVA, comparing each of the four groups using a family-wise
alpha threshold of 0.05 (displayed in Table S1). The assumption of equal variance was assessed
using the Levene’s test for equality of sample error variances. In cases where the groups did not
have equal variances, main effects were confirmed by one-way Welch’s F statistic, and the
Games-Howell test for multiple comparisons was used as a confirmation when a significant
interaction occurred (Table S6). For percent difference comparisons, two-tailed Student’s t test
was utilized, comparing either male control to male Dp/Ts or female control to female Dp/Ts
(Tables S2-S4). Given the number of tests run, p values for each category of parameters
(trabecular cortical, extrinsic, and intrinsic) were adjusted using the Benjamini-Hochberg method
(An et al., 2013). An adjusted p value < 0.05 was considered significant, and adjusted p values >
1.000 were reported as 1.000.
Competing Interests
The authors have no competing interests to declare.
Acknowledgement
We thank Dr. Charles R. Goodlett for his help and expertise with the statistical analysis.
Funding
This research in this manuscript was supported by funds from Eunice Kennedy Shriver National
Institute for Child Health and Human Development HD090603 (RJR). Additional funding was
provided by Undergraduate Research Opportunities Program grant from the Center for Research
and Learning at IUPUI (KS, MB, and DV). VLJT and EMCF were supported by the Wellcome
Trust (grants 098327 and 098328) and VLJT was supported by the Francis Crick Institute which
receives its core funding from Cancer Research UK (FC001194), the UK Medical Research
Council (FC001194), and the Wellcome Trust (FC001194). For the purpose of Open Access, the
authors have applied a CC-BY public copyright license to any Author Accepted Manuscript
version arising from this submission.
Disease Models & Mechanisms • DMM • Accepted manuscript

Data Availability
Raw data has been uploaded to DRYAD
CRediT Authorship Contribution Statement
Kourtney Sloan: Conceptualization, Formal analysis, Investigation, Writing – original draft
preparation, visualization, Funding acquisition, Project administration
Jared Thomas: Formal analysis, Investigation, Writing – review and editing
Matthew Blackwell: Conceptualization, Formal analysis, Investigation, Writing – review and
editing, Funding acquisition
Deanna Voisard: Formal analysis, Investigation, Writing – review and editing, Funding
acquisition
Eva Lana-Elola: Resources, Writing – review and editing
Sheona Watson-Scales: Resources, Writing – review and editing
Daniel L. Roper: Software, Formal analysis, Investigation, visualization
Joseph M. Wallace: Software, Resources, Writing – review and editing
Elizabeth M.C. Fisher: Resources, Writing – review and editing, Funding acquisition
Victor L.J. Tybulewicz: Resources, Writing – review and editing, Funding acquisition
Randall J. Roper: Conceptualization, Resources, Writing – original draft preparation, Funding
acquisition, Supervision, Project administration
REFERENCES
AN, Q., XU, D. & BROOKS, G. P. 2013. Type I Error Rates and Power of Multiple Hypothesis
Testing Procedures in Factorial ANOVA. Multiple Linear Regression Viewpoints, 39, 1-
16.
ANTONARAKIS, S. E., SKOTKO, B. G., RAFII, M. S., STRYDOM, A., PAPE, S. E.,
BIANCHI, D. W., SHERMAN, S. L. & REEVES, R. H. 2020. Down syndrome. Nat Rev
Dis Primers, 6, 9.
BERMAN, A. G., CLAUSER, C. A., WUNDERLIN, C., HAMMOND, M. A. & WALLACE, J.
M. 2015. Structural and Mechanical Improvements to Bone Are Strain Dependent with
Axial Compression of the Tibia in Female C57BL/6 Mice. PLoS One, 10, e0130504.
Disease Models & Mechanisms • DMM • Accepted manuscript

BLAZEK, J. D., ABEYSEKERA, I., LI, J. & ROPER, R. J. 2015. Rescue of the abnormal
skeletal phenotype in Ts65Dn Down syndrome mice using genetic and therapeutic
modulation of trisomic Dyrk1a. Hum Mol Genet, 24, 5687-96.
BLAZEK, J. D., GADDY, A., MEYER, R., ROPER, R. J. & LI, J. 2011. Disruption of bone
development and homeostasis by trisomy in Ts65Dn Down syndrome mice. Bone, 48,
275-80.
CARFI, A., LIPEROTI, R., FUSCO, D., GIOVANNINI, S., BRANDI, V., VETRANO, D. L.,
MELONI, E., MASCIA, D., VILLANI, E. R., MANES GRAVINA, E., BERNABEI, R.
& ONDER, G. 2017. Bone mineral density in adults with Down syndrome. Osteoporos
Int, 28, 2929-2934.
COSTA, R., DE MIGUEL, R., GARCIA, C., DE ASUA, D. R., CASTANEDA, S.,
MOLDENHAUER, F. & SUAREZ, C. 2017. Bone Mass Assessment in a Cohort of
Adults With Down Syndrome: A Cross-Sectional Study. Intellect Dev Disabil, 55, 315-
324.
COSTA, R., GULLON, A., DE MIGUEL, R., DE ASUA, D. R., BAUTISTA, A., GARCIA, C.,
SUAREZ, C., CASTANEDA, S. & MOLDENHAUER, F. 2018. Bone Mineral Density
Distribution Curves in Spanish Adults With Down Syndrome. J Clin Densitom, 21, 493-
500.
CUI, S., XIONG, F., HONG, Y., JUNG, J. U., LI, X. S., LIU, J. Z., YAN, R., MEI, L., FENG,
X. & XIONG, W. C. 2011. APPswe/Abeta regulation of osteoclast activation and RAGE
expression in an age-dependent manner. J Bone Miner Res, 26, 1084-98.
DAVISSON, M. T., BECHTEL, L. J., AKESON, E. C., FORTNA, A., SLAVOV, D. &
GARDINER, K. 2001. Evolutionary breakpoints on human chromosome 21. Genomics,
78, 99-106.
DE GRAAF, G., BUCKLEY, F., DEVER, J. & SKOTKO, B. G. 2017. Estimation of live birth
and population prevalence of Down syndrome in nine U.S. states. Am J Med Genet A,
173, 2710-2719.
DUCHON, A., DEL MAR MUNIZ MORENO, M., MARTIN LORENZO, S., SILVA DE
SOUZA, M. P., CHEVALIER, C., NALESSO, V., MEZIANE, H., LOUREIRO DE
SOUSA, P., NOBLET, V., ARMSPACH, J. P., BRAULT, V. & HERAULT, Y. 2021.
Disease Models & Mechanisms • DMM • Accepted manuscript

Multi-influential genetic interactions alter behaviour and cognition through six main
biological cascades in Down syndrome mouse models. Hum Mol Genet, 30, 771-788.
EDIE, S., ZAGHLOUL, N. A., LEITCH, C. C., KLINEDINST, D. K., LEBRON, J., THOLE, J.
F., MCCALLION, A. S., KATSANIS, N. & REEVES, R. H. 2018. Survey of Human
Chromosome 21 Gene Expression Effects on Early Development in Danio rerio. G3
(Bethesda), 8, 2215-2223.
FOTAKI, V., DIERSSEN, M., ALCANTARA, S., MARTINEZ, S., MARTI, E., CASAS, C.,
VISA, J., SORIANO, E., ESTIVILL, X. & ARBONES, M. L. 2002. Dyrk1A
haploinsufficiency affects viability and causes developmental delay and abnormal brain
morphology in mice. Mol Cell Biol, 22, 6636-47.
GU, Y., JONES, A. E., YANG, W., LIU, S., DAI, Q., LIU, Y., SWINDLE, C. S., ZHOU, D.,
ZHANG, Z., RYAN, T. M., TOWNES, T. M., KLUG, C. A., CHEN, D. & WANG, H.
2016. The histone H2A deubiquitinase Usp16 regulates hematopoiesis and hematopoietic
stem cell function. Proc Natl Acad Sci U S A, 113, E51-60.
HAMA, M., KIRINO, Y., TAKENO, M., TAKASE, K., MIYAZAKI, T., YOSHIMI, R., UEDA,
A., ITOH-NAKADAI, A., MUTO, A., IGARASHI, K. & ISHIGATSUBO, Y. 2012.
Bach1 regulates osteoclastogenesis in a mouse model via both heme oxygenase 1-
dependent and heme oxygenase 1-independent pathways. Arthritis Rheum, 64, 1518-28.
HSIAO, K., CHAPMAN, P., NILSEN, S., ECKMAN, C., HARIGAYA, Y., YOUNKIN, S.,
YANG, F. & COLE, G. 1996. Correlative memory deficits, Abeta elevation, and amyloid
plaques in transgenic mice. Science, 274, 99-102.
JI, C., LIU, X., XU, L., YU, T., DONG, C. & LUO, J. 2017. RUNX1 Plays an Important Role in
Mediating BMP9-Induced Osteogenic Differentiation of Mesenchymal Stem Cells Line
C3H10T1/2, Murine Multi-Lineage Cells Lines C2C12 and MEFs. Int J Mol Sci, 18.
JUNG, M. S., PARK, J. H., RYU, Y. S., CHOI, S. H., YOON, S. H., KWEN, M. Y., OH, J. Y.,
SONG, W. J. & CHUNG, S. H. 2011. Regulation of RCAN1 protein activity by Dyrk1A
protein-mediated phosphorylation. J Biol Chem, 286, 40401-12.
KIM, J. H., KIM, K., KIM, I., SEONG, S., JEONG, B. C., NAM, K. I., KIM, K. K.,
MOLKENTIN, J. D. & KIM, N. 2016. RCANs regulate the convergent roles of NFATc1
in bone homeostasis. Sci Rep, 6, 38526.
Disease Models & Mechanisms • DMM • Accepted manuscript

LANA-ELOLA, E., CATER, H., WATSON-SCALES, S., GREENAWAY, S., MULLER-
WINKLER, J., GIBBINS, D., NEMES, M., SLENDER, A., HOUGH, T., KESKIVALI-
BOND, P., SCUDAMORE, C. L., HERBERT, E., BANKS, G. T., MOBBS, H.,
CANONICA, T., TOSH, J., NOY, S., LLORIAN, M., NOLAN, P. M., GRIFFIN, J. L.,
GOOD, M., SIMON, M., MALLON, A. M., WELLS, S., FISHER, E. M. C. &
TYBULEWICZ, V. L. J. 2021. Comprehensive phenotypic analysis of the Dp1Tyb
mouse strain reveals a broad range of Down syndrome-related phenotypes. Dis Model
Mech, 14.
LANA-ELOLA, E., WATSON-SCALES, S., SLENDER, A., GIBBINS, D., MARTINEAU, A.,
DOUGLAS, C., MOHUN, T., FISHER, E. M. & TYBULEWICZ, V. L. 2016. Genetic
dissection of Down syndrome-associated congenital heart defects using a new mouse
mapping panel. Elife, 5.
LANA-ELOLA, E., WATSON-SCALES, S. D., FISHER, E. M. & TYBULEWICZ, V. L. 2011.
Down syndrome: searching for the genetic culprits. Dis Model Mech, 4, 586-95.
LEE, Y., HA, J., KIM, H. J., KIM, Y. S., CHANG, E. J., SONG, W. J. & KIM, H. H. 2009.
Negative feedback Inhibition of NFATc1 by DYRK1A regulates bone homeostasis. J
Biol Chem, 284, 33343-51.
LUO, Y., ZHANG, Y., MIAO, G., ZHANG, Y., LIU, Y. & HUANG, Y. 2019. Runx1 regulates
osteogenic differentiation of BMSCs by inhibiting adipogenesis through Wnt/beta-
catenin pathway. Arch Oral Biol, 97, 176-184.
MCKELVEY, K. D., FOWLER, T. W., AKEL, N. S., KELSAY, J. A., GADDY, D., WENGER,
G. R. & SUVA, L. J. 2013. Low bone turnover and low bone density in a cohort of adults
with Down syndrome. Osteoporos Int, 24, 1333-8.
MENG, Y. C., LIN, T., JIANG, H., ZHANG, Z., SHU, L., YIN, J., MA, X., WANG, C., GAO,
R. & ZHOU, X. H. 2020. miR-122 Exerts Inhibitory Effects on Osteoblast
Proliferation/Differentiation in Osteoporosis by Activating the PCP4-Mediated JNK
Pathway. Mol Ther Nucleic Acids, 20, 345-358.
MIYAI, K., YONEDA, M., HASEGAWA, U., TOITA, S., IZU, Y., HEMMI, H., HAYATA, T.,
EZURA, Y., MIZUTANI, S., MIYAZONO, K., AKIYOSHI, K., YAMAMOTO, T. &
NODA, M. 2009. ANA deficiency enhances bone morphogenetic protein-induced ectopic
bone formation via transcriptional events. J Biol Chem, 284, 10593-600.
Disease Models & Mechanisms • DMM • Accepted manuscript

MORIKAWA, D., NOJIRI, H., SAITA, Y., KOBAYASHI, K., WATANABE, K., OZAWA, Y.,
KOIKE, M., ASOU, Y., TAKAKU, T., KANEKO, K. & SHIMIZU, T. 2013.
Cytoplasmic reactive oxygen species and SOD1 regulate bone mass during mechanical
unloading. J Bone Miner Res, 28, 2368-80.
MOYER, A. J., GARDINER, K. & REEVES, R. H. 2021. All Creatures Great and Small: New
Approaches for Understanding Down Syndrome Genetics. Trends Genet, 37, 444-459.
OLSON, L. E. & MOHAN, S. 2011. Bone density phenotypes in mice aneuploid for the Down
syndrome critical region. Am J Med Genet A, 155A, 2436-45.
OLSON, L. E., RICHTSMEIER, J. T., LESZL, J. & REEVES, R. H. 2004. A chromosome 21
critical region does not cause specific Down syndrome phenotypes. Science, 306, 687-90.
OLSON, L. E., ROPER, R. J., SENGSTAKEN, C. L., PETERSON, E. A., AQUINO, V.,
GALDZICKI, Z., SIAREY, R., PLETNIKOV, M., MORAN, T. H. & REEVES, R. H.
2007. Trisomy for the Down syndrome 'critical region' is necessary but not sufficient for
brain phenotypes of trisomic mice. Hum Mol Genet, 16, 774-82.
ONISHI, M., FUJITA, Y., YOSHIKAWA, H. & YAMASHITA, T. 2013. Inhibition of Rac1
promotes BMP-2-induced osteoblastic differentiation. Cell Death Dis, 4, e698.
PAN, J. X., TANG, F., XIONG, F., XIONG, L., ZENG, P., WANG, B., ZHAO, K., GUO, H.,
SHUN, C., XIA, W. F., MEI, L. & XIONG, W. C. 2018. APP promotes osteoblast
survival and bone formation by regulating mitochondrial function and preventing
oxidative stress. Cell Death Dis, 9, 1077.
PLACE, D. E., MALIREDDI, R. K. S., KIM, J., VOGEL, P., YAMAMOTO, M. &
KANNEGANTI, T. D. 2021. Osteoclast fusion and bone loss are restricted by interferon
inducible guanylate binding proteins. Nat Commun, 12, 496.
ROPER, R. J. & REEVES, R. H. 2006. Understanding the basis for Down syndrome phenotypes.
PLoS Genet, 2, e50.
RYOO, S. R., CHO, H. J., LEE, H. W., JEONG, H. K., RADNAABAZAR, C., KIM, Y. S.,
KIM, M. J., SON, M. Y., SEO, H., CHUNG, S. H. & SONG, W. J. 2008. Dual-
specificity tyrosine(Y)-phosphorylation regulated kinase 1A-mediated phosphorylation of
amyloid precursor protein: evidence for a functional link between Down syndrome and
Alzheimer's disease. J Neurochem, 104, 1333-44.
Disease Models & Mechanisms • DMM • Accepted manuscript

SMIETANA, M. J., ARRUDA, E. M., FAULKNER, J. A., BROOKS, S. V. & LARKIN, L. M.
2010. Reactive oxygen species on bone mineral density and mechanics in Cu,Zn
superoxide dismutase (Sod1) knockout mice. Biochem Biophys Res Commun, 403, 149-
53.
STRINGER, M., ABEYSEKERA, I., THOMAS, J., LACOMBE, J., STANCOMBE, K.,
STEWART, R. J., DRIA, K. J., WALLACE, J. M., GOODLETT, C. R. & ROPER, R. J.
2017. Epigallocatechin-3-gallate (EGCG) consumption in the Ts65Dn model of down
syndrome fails to improve behavioral deficits and is detrimental to skeletal phenotypes.
Physiol Behav, 177, 230-241.
TANG, C. Y., WU, M., ZHAO, D., EDWARDS, D., MCVICAR, A., LUO, Y., ZHU, G.,
WANG, Y., ZHOU, H. D., CHEN, W. & LI, Y. P. 2021. Runx1 is a central regulator of
osteogenesis for bone homeostasis by orchestrating BMP and WNT signaling pathways.
PLoS Genet, 17, e1009233.
TANG, J. Y. M., LUO, H., WONG, G. H. Y., LAU, M. M. Y., JOE, G. M., TSE, M. A., IP, P.,
WONG, I. C. K. & LUM, T. Y. S. 2019. Bone mineral density from early to middle
adulthood in persons with Down syndrome. J Intellect Disabil Res, 63, 936-946.
THOMAS, J. R., LACOMBE, J., LONG, R., LANA-ELOLA, E., WATSON-SCALES, S.,
WALLACE, J. M., FISHER, E. M. C., TYBULEWICZ, V. L. J. & ROPER, R. J. 2020.
Interaction of sexual dimorphism and gene dosage imbalance in skeletal deficits
associated with Down syndrome. Bone, 136, 115367.
THOMAS, J. R. & ROPER, R. J. 2021. Current Analysis of Skeletal Phenotypes in Down
Syndrome. Curr Osteoporos Rep, 19, 338-346.
TIAN, X., CONG, F., GUO, H., FAN, J., CHAO, G. & SONG, T. 2019. Downregulation of
Bach1 protects osteoblasts against hydrogen peroxide-induced oxidative damage in vitro
by enhancing the activation of Nrf2/ARE signaling. Chem Biol Interact, 309, 108706.
WADA, S., KANZAKI, H., KATSUMATA, Y., YAMAGUCHI, Y., NARIMIYA, T.,
ATTUCKS, O. C., NAKAMURA, Y. & TOMONARI, H. 2020. Bach1 Inhibition
Suppresses Osteoclastogenesis via Reduction of the Signaling via Reactive Oxygen
Species by Reinforced Antioxidation. Front Cell Dev Biol, 8, 740.
WATSON-SCALES, S., KALMAR, B., LANA-ELOLA, E., GIBBINS, D., LA RUSSA, F.,
WISEMAN, F., WILLIAMSON, M., SACCON, R., SLENDER, A., OLERINYOVA, A.,
Disease Models & Mechanisms • DMM • Accepted manuscript

MAHMOOD, R., NYE, E., CATER, H., WELLS, S., YU, Y. E., BENNETT, D. L. H.,
GREENSMITH, L., FISHER, E. M. C. & TYBULEWICZ, V. L. J. 2018. Analysis of
motor dysfunction in Down Syndrome reveals motor neuron degeneration. PLoS Genet,
14, e1007383.
XIA, W. F., JUNG, J. U., SHUN, C., XIONG, S., XIONG, L., SHI, X. M., MEI, L. & XIONG,
W. C. 2013. Swedish mutant APP suppresses osteoblast differentiation and causes
osteoporotic deficit, which are ameliorated by N-acetyl-L-cysteine. J Bone Miner Res,
28, 2122-35.
XIAO, J., WU, Y., CHEN, R., LIN, Y., WU, L., TIAN, W. & LIU, L. 2008. Expression of Pcp4
gene during osteogenic differentiation of bone marrow mesenchymal stem cells in vitro.
Mol Cell Biochem, 309, 143-50.
ZHANG, Y., YANG, Y., XU, M., ZHENG, J., XU, Y., CHEN, G., GUO, Q., TIAN, W. & GUO,
W. 2021. The Dual Effects of Reactive Oxygen Species on the Mandibular Alveolar
Bone Formation in SOD1 Knockout Mice: Promotion or Inhibition. Oxid Med Cell
Longev, 2021, 8847140.
Disease Models & Mechanisms • DMM • Accepted manuscript

Figures and Tables
Fig. 1. Mouse mapping panel of triplicated regions of the Hsa21 orthologous genes on
mouse chromosome 16 (Mmu16). The solid black lines indicate the extent of the triplicated
regions with the first and last genes included in each triplication on the right. The numbers of
coding genes are based on the GRCm39 mouse genome assembly and have changed slightly
from the numbers reported in Lana-Elola et al. (2016) due to changes in genome annotation.
Disease Models & Mechanisms • DMM • Accepted manuscript

Fig. 2. Bone volume fraction (BV/TV) measurements in triplicated (Dp) mouse models and
control mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb
data from Thomas et al. (2020).
Disease Models & Mechanisms • DMM • Accepted manuscript

Fig. 3. Cortical models of triplicated (Dp) mouse models and control mice. Animal numbers
are as listed in Tables 1 and 2. Dp1Tyb data comes from Thomas et al. (2020). These radar
graphs were made using average periosteal and endocortical perimeter measurements taken
every 0.5 degrees for each cortical slice (total of 7) of each animal. The blue, filled-in image
represents the average cross-section of male or female control animals and the yellow outline
represents the average cross-section of male or female triplicated animals.
Disease Models & Mechanisms • DMM • Accepted manuscript

Fig. 4. Total cross-sectional area (Tt.Ar) measurements in triplicated (Dp) mouse models
and control mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM.
Dp1Tyb data comes from Thomas et al. (2020).
Disease Models & Mechanisms • DMM • Accepted manuscript

Fig. 5. Representative line images summarizing the significant (via one-way ANOVA;
adjusted p value < 0.05) results of bone volume fraction (BV/TV) and total cross-sectional
area (Tt.Ar) for all Dp mouse mapping strains. Solid black line indicates normal compared to
control, dotted black line indicates a deficit compared to control, and a double black line
indicates improvement compared to control.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 1. Significance of trabecular bone measurements from triplicated (Dp) mouse models and control mice.
Dp9Tyb Dp2Tyb Dp3Tyb Ts1Rhr
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Trabecular bone
BMD F(1,48)=16.20 F(1,48)=53.28 F(1,48)=0.143 F(1,54)=0.911 F(1,54)=187.7 F(1,54)=1.448 F(1,55)=0.108 F(1,55)=81.93 F(1,55)=0.306 F(1,56)=2.250 F(1,56)=44.51 F(1,56)=0.416
<0.001 <0.001 0.707 0.430 <0.001 0.319 0.930 <0.001 0.874 0.261 <0.001 0.783
BV/TV F(1,48)=16.84 F(1,48)=54.76 F(1,48)=0.404 F(1,54)=4.439 F(1,54)=252.6 F(1,54)=1.865 F(1,55)=0.148 F(1,55)=406.4 F(1,55)=0.775 F(1,56)=0.079 F(1,56)=157.3 F(1,56)=4.705
<0.001 <0.001 0.660 0.075 <0.001 0.267 0.957 <0.001 0.820 0.835 <0.001 0.086
Tb.Th F(1,48)=9.733 F(1,48)=11.26 F(1,48)=0.335 F(1,54)=0.024 F(1,54)=33.45 F(1,54)=0.187 F(1,55)=0.396 F(1,55)=49.77 F(1,55)=0.0001 F(1,56)=0.329 F(1,56)=1.803 F(1,56)=0.214
0.005 0.003 0.653 0.877 <0.001 0.715 0.886 <0.001 0.991 0.776 0.308 0.807
Tb.Sp F(1,48)=7.132 F(1,48)=46.17 F(1,48)=0.251 F(1,54)=6.705 F(1,54)=206.3 F(1,54)=0.429 F(1,55)=0.020 F(1,55)=302.4 F(1,55)=0.021 F(1,56)=0.202 F(1,56)=193.1 F(1,56)=2.311
0.016 <0.001 0.663 0.026 <0.001 0.595 0.951 <0.001 1.000 0.767 <0.001 0.287
Tb.N F(1,48)=14.93 F(1,48)=57.67 F(1,48)=0.409 F(1,54)=8.256 F(1,54)=331.0 F(1,54)=2.016 F(1,55)=0.469 F(1,55)=431.8 F(1,55)=1.281 F(1,56)=0.006 F(1,56)=196.6 F(1,56)=5.239
<0.001 <0.001 0.717 0.015 <0.001 0.269 0.931 <0.001 0.657 0.940 <0.001 0.078
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance was determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. Dp9Tyb: 13 male control, 12 male Dp, 13 female control, 14 female
Dp; Dp2Tyb: 15 male control, 15 male Dp, 14 female control, 14 female Dp; Dp3Tyb: 15 male control, 15 male Dp, 15 female control, 14 female Dp; Ts1Rhr:
15 male control, 14 male Ts, 15 female control, 16 female Ts.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 2. Significance of trabecular bone measurements from triplicated (Dp) mouse models and control mice.
Dp4Tyb Dp5Tyb Dp6Tyb Dp1Tyb,Dyrk1a+/+/-
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Trabecular Bone
BMD F(1,54)=0.152 F(1,54)=111.4 F(1,54)=0.0008 F(1,57)=1.969 F(1,57)=205.8 F(1,57)=9.921 F(1,47)=0.173 F(1,47)=183.0 F(1,47)=7.516 F(1,35)=0.918 F(1,35)=106.6 F(1,35)=0.982
0.873 <0.001 1.000 0.249 <0.001 0.008 0.784 <0.001 0.016 0.470 <0.001 0.493
BV/TV F(1,54)=0.667 F(1,54)=178.5 F(1,54)=0.005 F(1,57)=2.184 F(1,57)=219.9 F(1,57)=7.199 F(1,47)=0.004 F(1,47)=246.0 F(1,47)=9.572
0.783 <0.001 1.000 0.242 <0.001 0.020 0.949 <0.001 0.008
F(1,35)=2.747a F(1,35)=104.3a F(1,35)=2.175a
0.228a <0.001a 0.280a
Tb.Th F(1,54)=0.0007 F(1,54)=17.76 F(1,54)=1.648 F(1,57)=9.098 F(1,57)=0.643 F(1,57)=1.370 F(1,47)=5.182 F(1,47)=19.73 F(1,47)=5.820 F(1,35)=0.115 F(1,35)=6.014 F(1,35)=1.518
0.980 <0.001 0.512 0.010 0.491 0.308 0.041 <0.001 0.033 0.736 0.058 0.377
Tb.Sp F(1,54)=0.245 F(1,54)=206.9 F(1,54)=0.291 F(1,57)=0.069 F(1,57)=283.0 F(1,57)=1.557 F(1,47)=0.080 F(1,47)=349.1 F(1,47)=3.601 F(1,35)=0.639 F(1,35)=130.8 F(1,35)=0.476
0.934 <0.001 0.986 0.793 <0.001 0.296 0.835 <0.001 0.087 0.537 <0.001 0.530
Tb.N F(1,54)=1.273 F(1,54)=179.2 F(1,54)=0.161 F(1,57)=0.078 F(1,57)=251.0 F(1,57)=4.502 F(1,47)=0.554 F(1,47)=298.8 F(1,47)=8.180 F(1,35)=2.907 F(1,35)=137.1 F(1,35)=0.483
0.566 <0.001 0.940 0.837 <0.001 0.072 0.575 <0.001 0.014 0.243 <0.001 0.568
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance was determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. a logarithmic transformation of data. Dp4Tyb: 14 male control, 15
male Dp, 15 female control, 14 female Dp; Dp5Tyb: 14 male control, 15 male Dp, 17 female control, 15 female Dp; Dp6Tyb: 16 male control, 15 male Dp, 11
female control, 9 female Dp; Dp1Tyb,Dyrk1a+/+/-: 7 male control, 6 male Dp, 14 female control, 12 female Dp.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 3.: Significance of cortical bone measurements from triplicated (Dp) mouse models and control mice.
Dp9Tyb Dp2Tyb Dp3Tyb Ts1Rhr
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Cortical bone
Tt.Ar F(1,48)=11.94 F(1,48)=56.46 F(1,48)=0.331 F(1,54)=15.00 F(1,54)=105.6 F(1,54)=5.221 F(1,55)=26.74 F(1,55)=146.5 F(1,55)=1.915 F(1,56)=13.93 F(1,56)=47.84 F(1,56)=2.973
0.003 <0.001 0.745 <0.001 <0.001 0.037 <0.001 <0.001 0.241 0.002 <0.001 0.158
Ma.Ar F(1,48)=9.063 F(1,48)=67.63 F(1,48)=0.248 F(1,54)=28.37 F(1,54)=102.0 F(1,54)=11.95 F(1,55)=33.08 F(1,55)=102.8 F(1,55)=1.909 F(1,56)=13.14 F(1,56)=44.39 F(1,56)=2.779
0.009 <0.001 0.724 <0.001 <0.001 0.002 <0.001 <0.001 0.227 0.002 <0.001 0.163
Ct.Ar F(1,48)=10.03 F(1,48)=17.40 F(1,48)=0.284 F(1,54)=0.397 F(1,54)=49.33 F(1,54)=0.002 F(1,55)=6.099 F(1,55)=116.9 F(1,55)=0.809 F(1,56)=4.570 F(1,56)=3.055 F(1,56)=2.029
0.006 <0.001 0.737 0.558 <0.001 0.966 0.027 <0.001 0.412 0.077 0.164 0.240
Ct.Th F(1,48)=1.960 F(1,48)=1.528 F(1,48)=0.078 F(1,54)=5.845 F(1,54)=5.458 F(1,54)=3.545 F(1,55)=3.843 F(1,55)=20.34 F(1,55)=0.151 F(1,56)=0.271 F(1,56)=0.223 F(1,56)=0.012
0.294 0.359 0.821 0.031 0.035 0.080 0.083 <0.001 0.699 0.706 0.706 0.912
Ps.Pm F(1,48)=13.69 F(1,48)=68.84 F(1,48)=0.142 F(1,54)=15.91 F(1,54)=119.3 F(1,54)=4.421 F(1,55)=30.42 F(1,55)=185.0 F(1,55)=1.553 F(1,56)=12.52 F(1,56)=52.80 F(1,56)=2.016
0.002 <0.001 0.782 <0.001 <0.001 0.053 <0.001 <0.001 0.269 0.002 <0.001 0.226
Ec.Pm F(1,48)=8.681 F(1,48)=62.01 F(1,48)=0.0001 F(1,54)=21.29 F(1,54)=152.7 F(1,54)=8.015 F(1,55)=25.77 F(1,55)=116.8 F(1,55)=0.196 F(1,56)=10.18 F(1,56)=48.26 F(1,56)=0.699
0.009 <0.001 0.991 <0.001 <0.001 0.011 <0.001 <0.001 0.692 0.005 <0.001 0.502
TMD F(1,48)=0.754 F(1,48)=12.63 F(1,48)=0.403 F(1,54)=3.018 F(1,54)=10.43 F(1,54)=2.216 F(1,55)=1.421 F(1,55)=12.25 F(1,55)=6.320 F(1,56)=0.884 F(1,56)=10.46 F(1,56)=0.020
0.584 0.003 0.740 0.103 0.004 0.157 0.278 0.002 0.026 0.461 0.006 0.931
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance was determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. Dp9Tyb: 13 male control, 12 male Dp, 13 female control, 14 female
Dp; Dp2Tyb: 15 male control, 15 male Dp, 14 female control, 14 female Dp; Dp3Tyb: 15 male control, 15 male Dp, 15 female control, 14 female Dp; Ts1Rhr:
15 male control, 14 male Ts, 15 female control, 16 female Ts.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 4. Significance of cortical bone measurements from triplicated (Dp) mouse models and control mice.
Dp4Tyb
Dp5Tyb
Dp6Tyb
Dp1Tyb,Dyrk1a+/+/-
Genotype
Sex
Interaction
Genotype
Sex
Interaction
Genotype
Sex
Interaction
Genotype
Sex
Interaction
Cortical Bone
Tt.Ar
F(1,54)=3.392
F(1,54)=77.79
F(1,54)=2.387
F(1,57)=25.52
F(1,57)=73.06
F(1,57)=8.322
F(1,47)=0.004
F(1,47)=167.0
F(1,47)=1.169
F(1,35)=63.41
F(1,35)=53.26
F(1,35)=0.011
0.166
<0.001
0.224
<0.001
<0.001
0.010
0.998
<0.001
0.499
<0.001
<0.001
0.963
Ma.Ar
F(1,54)=7.169
F(1,54)=111.4
F(1,54)=5.681
F(1,57)=19.77a
F(1,57)=139.5a
F(1,57)=5.607a
F(1,47)=0.609
F(1,47)=198.5
F(1,47)=0.056
F(1,35)=62.19
F(1,35)=53.40
F(1,35)=0.013
0.029
<0.001
0.054
<0.001a
<0.001a
0.026a
0.659
<0.001
0.950
<0.001
<0.001
1.000
Ct.Ar
F(1,54)=0.1641
F(1,54)=20.43
F(1,54)=0.034
F(1,57)=24.02
F(1,57)=2.936
F(1,57)=4.910
F(1,47)=0.515
F(1,47)=23.41
F(1,47)=1.921
F(1,35)=23.21
F(1,35)=18.68
F(1,35)=0.102
0.802
<0.001
0.855
<0.001
0.102
0.036
0.667
<0.001
0.452
<0.001
<0.001
0.928
Ct.Th
F(1,54)=1.021
F(1,54)=0.066
F(1,54)=1.178
F(1,57)=7.897
F(1,57)=53.22
F(1,57)=0.776
F(1,47)=1.470
F(1,47)=1.326
F(1,47)=2.197
F(1,35)=1.955
F(1,35)=1.385
F(1,35)=0.350
0.416
0.882
0.396
0.011
<0.001
0.401
0.540
0.536
0.435
0.276
0.371
0.733
Ps.Pm
F(1,54)=2.741
F(1,54)=91.65
F(1,54)=1.660
F(1,57)=20.74
F(1,57)=89.98
(1,57)=6.106
F(1,47)=0.039
F(1,47)=182.0
F(1,47)=0.510
F(1,35)=54.68
F(1,35)=55.42
F(1,35)=0.016
0.198
<0.001
0.305
<0.001
<0.001
0.023
0.933
<0.001
0.628
<0.001
<0.001
1.000
Ec.Pm
F(1,54)=3.214a
F(1,54)=151.1a
F(1,54)=2.365a
F(1,57)=16.67b
F(1,57)=174.5
F(1,57)=6.026
F(1,47)=0.065
F(1,47)=286.5
F(1,47)=0.929
F(1,35)=27.29
F(1,35)=39.06
F(1,35)=0.001
0.165a
<0.001a
0.210a
<0.001b
<0.001
0.023
0.989
<0.001
0.549
<0.001
<0.001
0.978
TMD
F(1,54)=0.053
F(1,54)=11.32
F(1,54)=0.460
F(1,57)=7.614
F(1,57)=29.86
F(1,57)=0.131
F(1,47)=1.212
F(1,47)=57.09
F(1,47)=0.001
F(1,35)=2.016
F(1,35)=12.77
F(1,35)=1.315
0.859
0.005
0.619
0.012
<0.001
0.719
0.528
<0.001
0.972
0.288
0.002
0.363
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance is determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. a logarithmic transformation of data. b significant Levene’s test
(unequal variances) and subsequent analysis indicated no significant difference. Dp4Tyb: 14 male control, 15 male Dp, 15 female control, 14 female Dp;
Dp5Tyb: 14 male control, 15 male Dp, 17 female control, 15 female Dp; Dp6Tyb: 16 male control, 15 male Dp, 11 female control, 9 female Dp;
Dp1Tyb,Dyrk1a+/+/-: 7 male control, 6 male Dp, 14 female control, 12 female Dp.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 5. Significance of extrinsic mechanical measurements from triplicated (Dp) mouse models and control mice.
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance is determined by an adjusted p value < 0.05.
Orange text signifies control mice are significantly greater than triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than
control mice. Blue text signifies male mice are significantly greater than female mice; purple text signifies female mice are significantly greater than male mice. a
logarithmic transformation of data. Dp9Tyb: 13 male control, 12 male Dp, 11 female control, 12 female Dp; Dp2Tyb: 15 male control, 15 male Dp, 11 female
control, 13 female Dp; Dp3Tyb: 12 male control, 14 male Dp, 12 female control, 13 female Dp; Ts1Rhr: 15 male control, 11 male Ts, 15 female control, 15
female Ts.
Dp9Tyb Dp2Tyb Dp3Tyb Ts1Rhr
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Extrinsic
Yield Force F(1,44)=0.020 F(1,44)=23.50 F(1,44)=0.230 F(1,50)=3.251 F(1,50)=9.765 F(1,50)=4.360 F(1,47)=0.576 F(1,47)=5.928 F(1,47)=2.930 F(1,52)=0.963 F(1,52)=0.981 F(1,52)=1.728
0.889 0.001 0.887 0.148 0.011 0.088 0.593 0.079 0.179 0.463 0.490 0.340
Ultimate Force F(1,44)=9.878 F(1,44)=1.042 F(1,44)=0.097 F(1,50)=0.092 F(1,50)=15.46 F(1,50)=0.710 F(1,47)=3.341 F(1,47)=18.01 F(1,47)=0.540 F(1,52)=3.164 F(1,52)=4.430 F(1,52)=0.789
0.016 0.598 0.795 0.843 0.006 0.498 0.155 0.002 0.576 0.213 0.169 0.497
Displacement
to Yield
F(1,44)=0.808 F(1,44)=42.25 F(1,44)=0.996 F(1,50)=0.947 F(1,50)=12.47 F(1,50)=4.729 F(1,47)=5.517 F(1,47)=15.41 F(1,47)=2.720 F(1,52)=5.098 F(1,52)=0.607 F(1,52)=1.761
0.604 0.001 0.567 0.469 0.005 0.080 0.081 0.003 0.185 0.148 0.543 0.363
Ultimate
Displacement
F(1,44)=1.402a F(1,44)=6.688a F(1,44)=1.323a F(1,50)=12.53 F(1,50)=13.50 F(1,50)=1.300 F(1,47)=3.428 F(1,47)=8.426 F(1,47)=0.200 F(1,52)=7.228 F(1,52)=14.24 F(1,52)=0.487
0.728a 0.055a 0.673a 0.005 0.006 0.389 0.164 0.039 0.690 0.067 0.008 0.570
Stiffness F(1,44)=3.659 F(1,44)=0.120 F(1,44)=1.309 F(1,50)=0.048 F(1,50)=1.996 F(1,50)=0.134 F(1,47)=4.810 F(1,47)=1.278 F(1,47)=0.360 F(1,52)=3.694 F(1,52)=0.234 F(1,52)=0.175
0.218 0.852 0.604 0.868 0.265 0.884 0.100 0.427 0.643 0.210 0.662 0.678
Work to Yield F(1,44)=0.131 F(1,44)=40.58 F(1,44)=1.123 F(1,50)=2.381 F(1,50)=11.77 F(1,50)=5.173 F(1,47)=0.982 F(1,47)=7.865 F(1,47)=3.552 F(1,52)=2.590a F(1,52)=1.260a F(1,52)=1.909a
0.888 0.001 0.620 0.226 0.005 0.072 0.457 0.038 0.173 0.265a 0.431a 0.363a
Total Work F(1,44)=0.118 F(1,44)=0.169 F(1,44)=0.308 F(1,50)=0.019a F(1,50)=6.13a F(1,50)=0.762a F(1,47)=1.103 F(1,47)=0.041 F(1,47)=0.241 F(1,52)=3.308 F(1,52)=10.96 F(1,52)=0.436
0.810 0.896 0.873 0.890a 0.0501a 0.508a 0.449 0.840 0.692 0.224 0.018 0.566
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 6. Significance of extrinsic mechanical measurements from triplicated (Dp) mouse models and control mice.
Dp4Tyb Dp5Tyb Dp6Tyb Dp1Tyb,Dyrk1a+/+/-
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Extrinsic
Yield Force F(1,49)=0.177 F(1,49)=0.005 F(1,49)=1.028 F(1,55)=0.208 F(1,55)=6.894 F(1,55)=0.011 F(1,46)=2.334 F(1,46)=5.903 F(1,46)=3.463 F(1,34)=0.229 F(1,34)=0.010 F(1,34)=0.115
1.000 0.994 0.829 0.975 0.078 1.000 0.255 0.067 0.162 0.834 0.921 0.814
Ultimate Force F(1,49)=0.708 F(1,49)=7.856 F(1,49)=0.145 F(1,55)=3.177 F(1,55)=2.372 F(1,55)=0.402 F(1,46)=1.203 F(1,46)=5.465 F(1,46)=1.603 F(1,34)=13.41 F(1,34)=19.72 F(1,34)=0.937
0.943 0.076 0.925 0.281 0.388 0.925 0.418 0.071 0.371 0.008 0.002 0.893
Displacement
to Yield
F(1,49)=0.275 F(1,49)=1.037 F(1,49)=0.087 F(1,55)=4.357 F(1,55)=0.0009 F(1,55)=0.163 F(1,46)=0.029 F(1,46)=12.84 F(1,46)=0.012 F(1,34)=0.350a F(1,34)=0.216a F(1,34)=0.402a
1.000 1.000 0.951 0.174 1.000 0.964 0.956 0.008 0.960 0.902a 0.797a 0.928a
Ultimate
Displacement
F(1,49)=0.194 F(1,49)=2.397 F(1,49)=1.542 F(1,55)=5.851 F(1,55)=0.2722 F(1,55)=0.009 F(1,46)=1.279 F(1,46)=8.479 F(1,46)=1.077 F(1,34)=4.424 F(1,34)=4.533 F(1,34)=0.187
1.000 0.896 1.000 0.099 0.976 1.000 0.426 0.039 0.427 0.225 0.284 0.780
Stiffness F(1,49)=0.145 F(1,49)=8.586 F(1,49)=0.017 F(1,55)=8.758 F(1,55)=7.196 F(1,55)=0.424 F(1,46)=0.892 F(1,46)=8.343 F(1,46)=4.860 F(1,34)=0.683 F(1,34)=0.656 F(1,34)=0.876
0.987 0.107 0.993 0.095 0.101 0.988 0.459 0.031 0.085 0.870 0.809 0.830
Work to Yield F(1,49)=0.208 F(1,49)=1.034 F(1,49)=0.658 F(1,55)=1.807 F(1,55)=0.812 F(1,55)=0.001 F(1,46)=0.091 F(1,46)=15.24 F(1,46)=0.319 F(1,34)=0.294a F(1,34)=0.071a F(1,34)=0.336a
1.000 0.943 0.885 0.388 0.780 0.975 0.892 0.006 0.710 0.827a 0.831a 0.849a
Total Work F(1,49)=0.074 F(1,49)=1.399 F(1,49)=0.00008 F(1,55)=0.003 F(1,55)=1.431 F(1,55)=0.017 F(1,46)=0.007 F(1,46)=6.879 F(1,46)=2.527 F(1,34)=2.393a F(1,34)=3.932a F(1,34)=3.096a
0.917 1.000 0.993 1.000 0.552 1.000 0.936 0.0496 0.250 0.394a 0.233a 0.306a
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance is determined by an adjusted p value < 0.05.
Blue text signifies male mice are significantly greater than female mice; purple text signifies female mice are significantly greater than male mice. a logarithmic
transformation of data. Dp4Tyb: 13 male control, 13 male Dp, 14 female control, 13 female Dp; Dp5Tyb: 14 male control, 14 male Dp, 16 female control, 15
female Dp; Dp6Tyb: 16 male control, 15 male Dp, 11 female control, 9 female Dp; Dp1Tyb,Dyrk1a+/+/-: 7 male control, 5 male Dp, 14 female control, 12 female
Dp.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 7. Significance of intrinsic mechanical measurements from triplicated (Dp) mouse models and control mice.
Dp9Tyb Dp2Tyb Dp3Tyb Ts1Rhr
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Intrinsic
Yield
Stress
F(1,44)=2.56 F(1,44)=60.14 F(1,44)=0.058 F(1,50)=11.43 F(1,50)=60.76 F(1,50)=4.808 F(1,47)=10.92 F(1,47)=54.51 F(1,47)=0.910 F(1,52)=5.922 F(1,52)=13.84 F(1,52)=1.271
0.263 <0.001 0.911 0.003 <0.001 0.0495 0.007 <0.001 0.414 0.055 0.003 0.341
Ultimate
Stress
F(1,44)=0.192 F(1,44)=16.62 F(1,44)=0.033 F(1,50)=8.052 F(1,50)=12.27 F(1,50)=5.574 F(1,47)=3.686 F(1,47)=7.095 F(1,47)=0.779 F(1,52)=1.794 F(1,52)=19.18 F(1,52)=2.975
0.796 <0.001 0.906 0.012 0.003 0.036 0.110 0.027 0.430 0.305 0.002 0.181
Strain to
Yield
F(1,44)=0.003a F(1,44)=35.38a F(1,44)=0.854a F(1,50)=0.135 F(1,50)=8.481 F(1,50)=4.003 F(1,47)=2.105 F(1,47)=8.525 F(1,47)=3.077 F(1,52)=2.410a F(1,52)=0.630a F(1,52)=0.792a
0.958a <0.001a 0.649a 0.715 0.011 0.065 0.197 0.016 0.141 0.228a 0.485a 0.453a
Ultimate
Strain
F(1,44)=3.74 F(1,44)=7.005 F(1,44)=0.687 F(1,50)=14.51 F(1,50)=16.46 F(1,50)=0.436 F(1,47)=5.260 F(1,47)=12.04 F(1,47)=0.327 F(1,52)=10.95 F(1,52)=15.74 F(1,52)=1.750
0.153 0.034 0.674 0.001 <0.001 0.576 0.059 0.005 0.570 0.006 0.002 0.288
Modulus F(1,44)=2.38 F(1,44)=17.96 F(1,44)=0.399 F(1,50)=15.43 F(1,50)=32.55 F(1,50)=0.272 F(1,47)=4.077 F(1,47)=30.98 F(1,47)=0.749 F(1,52)=0.564 F(1,52)=11.29 F(1,52)=0.372
0.260 <0.001 0.683 0.001 <0.001 0.640 0.098 <0.001 0.414 0.483 0.007 0.545
Resilience F(1,44)=0.597 F(1,44)=55.93 F(1,44)=0.647 F(1,50)=3.833 F(1,50)=31.01 F(1,50)=4.764 F(1,47)=2.795 F(1,47)=22.11 F(1,47)=2.246 F(1,52)=4.191 F(1,52)=3.589 F(1,52)=1.381
0.614 <0.001 0.638 0.067 <0.001 0.047 0.152 <0.001 0.195 0.118 0.143 0.340
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance is determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. a logarithmic transformation of data. Dp9Tyb: 13 male control, 12
male Dp, 11 female control, 12 female Dp; Dp2Tyb: 15 male control, 15 male Dp, 11 female control, 13 female Dp; Dp3Tyb: 12 male control, 14 male Dp, 12
female control, 13 female Dp; Ts1Rhr: 15 male control, 11 male Ts, 15 female control, 15 female Ts.
Disease Models & Mechanisms • DMM • Accepted manuscript

Table 8. Significance of intrinsic mechanical measurements from triplicated (Dp) mouse models and control mice.
Dp4Tyb Dp5Tyb Dp6Tyb Dp1Tyb,Dyrk1a+/+/-
Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction Genotype Sex Interaction
Intrinsic
Yield Stress F(1,49)=0.007 F(1,49)=3.361 F(1,49)=3.361 F(1,55)=6.380 F(1,55)=19.36 F(1,55)=0.646 F(1,46)=1.392 F(1,46)=34.41 F(1,46)=1.551 F(1,34)=12.93 F(1,34)=2.435 F(1,34)=0.297
0.932 0.262 0.487 0.052 <0.001 0.589 0.366 <0.001 0.395 0.006 0.329 0.816
Ultimate Stress F(1,49)=5.430 F(1,49)=16.12 F(1,49)=3.830 F(1,55)=4.234 F(1,55)=37.74 F(1,55)=2.108 F(1,46)=1.488 F(1,46)=8.924 F(1,46)=0.359 F(1,34)=20.48 F(1,34)=0.0002 F(1,34)=0.548
0.216 0.004 0.252 0.133 <0.001 0.304 0.374 0.014 0.710 0.001 0.987 0.759
Strain to Yield F(1,49)=0.460 F(1,49)=0.328 F(1,49)=0.062 F(1,55)=1.875a F(1,55)=0.142a F(1,55)=0.003a F(1,46)=0.301 F(1,46)=5.207 F(1,46)=0.069 F(1,34)=0.036a F(1,34)=0.012a F(1,34)=0.402a
0.751 0.733 0.852 0.318a 0.780a 0.959a 0.703 0.061 0.841 0.956a 0.969a 0.795a
Ultimate Strain F(1,49)=0.385 F(1,49)=4.780 F(1,49)=1.664 F(1,55)=10.05 F(1,55)=1.281 F(1,55)=0.243 F(1,46)=2.588 F(1,46)=21.50 F(1,46)=0.959 F(1,34)=10.44 F(1,34)=6.559 F(1,34)=0.082
0.745 0.202 0.487 0.011 0.394 0.749 0.229 <0.001 0.461 0.012 0.054 0.932
Modulus F(1,49)=0.225 F(1,49)=0.603 F(1,49)=0.980 F(1,55)=0.635 F(1,55)=34.14 F(1,55)=1.715 F(1,46)=10.58 F(1,46)=46.16 F(1,46)=7.877 F(1,34)=24.60 F(1,34)=3.245 F(1,34)=0.938
0.764 0.722 0.654 0.552 <0.001 0.320 0.008 <0.001 0.019 0.001 0.242 0.679
Resilience F(1,49)=0.175 F(1,49)=2.811 F(1,49)=0.784 F(1,55)=3.996 F(1,55)=3.261 F(1,55)=0.004 F(1,46)=0.073 F(1,46)=21.64 F(1,46)=0.006 F(1,34)=2.202a F(1,34)=0.581a F(1,34)=0.290a
0.751 0.300 0.685 0.130 0.172 1.000 0.886 <0.001 0.939 0.331a 0.812a 0.764a
Top row contains the F value and bottom row contains the adjusted p value from two-way ANOVA. Significance is determined by an adjusted p value < 0.05.
Significant interactions are highlighted in yellow and post hoc analysis can be found in Table S1. Orange text signifies control mice are significantly greater than
triplicated (Dp) mice; green text signifies triplicated (Dp) mice are significantly greater than control mice. Blue text signifies male mice are significantly greater
than female mice; purple text signifies female mice are significantly greater than male mice. a logarithmic transformation of data. Dp4Tyb: 13 male control, 13
male Dp, 14 female control, 13 female Dp; Dp5Tyb: 14 male control, 14 male Dp, 16 female control, 15 female Dp; Dp6Tyb: 16 male control, 15 male Dp, 11
female control, 9 female Dp; Dp1Tyb,Dyrk1a+/+/-: 7 male control, 5 male Dp, 14 female control, 12 female Dp.
Disease Models & Mechanisms • DMM • Accepted manuscript

Fig. S1. Bone mineral density measurements in triplicated (Dp) mouse model and control
mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data
comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S2. Trabecular thickness measurements in triplicated (Dp) mouse models and control
mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data
comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S3.Trabecular separation measurements in triplicated (Dp) mouse models and control
mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data
comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S4. Trabecular number measurements in triplicated (Dp) mouse models and control
mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data
from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S5. Marrow area measurements in triplicated (Dp) mouse models and control mice.
Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data comes
from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S6. Cortical area measurements in triplicated (Dp) mouse models and control mice.
Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data comes
from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S7. Cortical thickness measurements in triplicated (Dp) models and control mice.
Animal numbers in Tables 2. Data are mean ± SEM. Dp1Tyb data from Thomas et al.
(2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S8. Periosteal perimeter measurements in triplicated (Dp) mouse models and control
mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb data
comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S9. Endocortical perimeter measurements in triplicated (Dp) mouse models and
control mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb
data comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Fig. S10. Tissue mineral density measurements in triplicated (Dp) mouse models and
control mice. Animal numbers are as listed in Tables 1 and 2. Data are mean ± SEM. Dp1Tyb
data comes from Thomas et al. (2020).
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Male Dp5Tyb
Male Dp5Tyb control
Female Dp5Tyb
Male Dp5Tyb control
Male Ts1Rhr
Female Ts1Rhr Control
Female Ts1Rhr
Male Ts1Rhr Control
Fig. S11. Correlation between bone parameters of Dp5Tyb and Ts1Rhr male and female
mice. Scale of correlation is at the right, with blue as a positive and red as a negative correlation.
The larger the circle, the greater the correlation.
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Table S1. Results of Tukey’s post hoc analysis (multiple comparisons) for two-way ANOVAs.
This was only performed if there was a significant (adjusted p value < 0.05) interaction on the
two-way ANOVA. a logarithmic transformation of data. b significant Levene’s test (unequal
variances) and subsequent analysis indicated no significant difference. M = male, F = female,
WT = control/wildtype, Dp = triplicated Dp/Ts strain
Dp2Tyb
Tt.Ar
Ma.Ar
Ec.Pm
Yield
Stress
Ultimate
Stress
Resilience
M WT vs M Dp
0.0003
<0.0001
<0.0001
0.0006
0.0016
0.0158
M WT vs F WT
<0.0001
<0.0001
<0.0001
<0.0001
0.0010
<0.0001
M WT vs F Dp
<0.0001
<0.0001
<0.0001
<0.0001
0.0002
<0.0001
M Dp vs F WT
0.0002
0.0073
<0.0001
0.0185
0.9673
0.0728
M Dp vs F Dp
<0.0001
0.0001
<0.0001
0.0010
0.8413
0.0804
F WT vs F Dp
0.6889
0.5669
0.6048
0.8560
0.9886
0.9988
Dp3Tyb
TMD
M WT vs M Dp
0.7818
M WT vs F WT
0.8954
M WT vs F Dp
0.0093
M Dp vs F WT
0.3618
M Dp vs F Dp
0.0005
F WT vs F Dp
0.0565
Dp5Tyb
BMD
BV/TV
Tt.Ar
Ma.Ara
Ct.Ar
Ps.Pm
Ec.Pm
M WT vs M Dp
0.0137
0.0283b
<0.0001
<0.0001
<0.0001
<0.0001
0.0002
M WT vs F WT
<0.0001
<0.0001
<0.0001
<0.0001
0.0346
<0.0001
<0.0001
M WT vs F Dp
<0.0001
<0.0001
<0.0001
<0.0001
0.0002
<0.0001
<0.0001
M Dp vs F WT
<0.0001
<0.0001
0.0654
<0.0001
0.1077
0.0039
<0.0001
M Dp vs F Dp
<0.0001
<0.0001
0.0011
<0.0001
0.9847
<0.0001
<0.0001
F WT vs F Dp
0.5883
0.8185
0.4037
0.4407
0.2208
0.4387
0.6417
Dp6Tyb
BMD
BV/TV
Tb.Th
Tb.N
Modulus
M WT vs M Dp
0.0685
0.0684
0.9995
0.029b
0.9836
M WT vs F WT
<0.0001
<0.0001
<0.0001
<0.0001
0.0292
M WT vs F Dp
<0.0001
<0.0001
0.4481
<0.0001
<0.0001
M Dp vs F WT
<0.0001
<0.0001
<0.0001
<0.0001
0.0684
M Dp vs F Dp
<0.0001
<0.0001
0.5147
<0.0001
<0.0001
F WT vs F Dp
0.4520
0.2257
0.0215
0.5337
0.002
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Table S2. Percent differences in trabecular skeletal values of male and female triplicated and control
littermate mice. Percent change ([([mean of Dp-mean of Control]/mean of Control)*100]) reported in
tables. Red indicates significant deficit; blue indicates significant improvement compared to same-sex
control littermates (adjusted p < 0.05 via two-tail t test). Adjusted p values reported in Table S4.
MALE
Dp1Tyb
Dp1Tyb,
Dyrk1a+/+/-
Dp9Tyb
Dp2Tyb
Dp3Tyb
Ts1Rhr
Dp4Tyb
Dp5Tyb
Dp6Tyb
BMD
-15.78%
-0.11%
17.03%
6.85%
-0.82%
-5.27%
1.21%
-10.35%
-8.03%
BV/TV
-15.83%
1.30%
29.48%
13.17%
-3.29%
-7.82%
2.71%
-11.63%
-8.72%
Tb.Th
-6.71%
-2.60%
3.64%
0.52%
0.88%
-1.21%
-2.33%
-5.68%
-0.19%
Tb.Sp
3.76%
-0.46%
-11.25%
-9.35%
0.67%
2.94%
-2.87%
2.27%
4.96%
Tb.N
-10.29%
4.16%
25.04%
12.88%
-4.12%
-6.64%
5.14%
-6.36%
-8.35%
FEMALE
Dp1Tyb
Dp1Tyb,
Dyrk1a+/+/-
Dp9Tyb
Dp2Tyb
Dp3Tyb
Ts1Rhr
Dp4Tyb
Dp5Tyb
Dp6Tyb
BMD
10.84%
10.55%
19.26%
-1.31%
4.99%
-21.00%
2.17%
6.59%
10.09%
BV/TV
11.12%
22.32%
37.26%
6.27%
2.84%
11.34%
6.35%
6.60%
17.36%
Tb.Th
0.73%
1.57%
5.61%
-1.22%
1.02%
-0.13%
2.49%
-2.57%
7.41%
Tb.Sp
-4.37%
-4.14%
-6.29%
-4.00%
-0.004%
-3.80%
0.09%
-2.46%
-2.50%
Tb.N
9.97%
20.15%
29.24%
8.78%
2.01%
11.48%
4.33%
9.32%
9.19%
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Table S3. Percent differences in cortical skeletal values of male and female triplicated and control
littermate mice. Percent change ([([mean of Dp-mean of Control]/mean of Control)*100]) reported in
tables. Red indicates significant deficit; blue indicates significant improvement compared to same-sex
control littermates (adjusted p < 0.05 via two-tail t test). Adjusted p values reported in Table S4.
MALE
Dp1Tyb
Dp1Tyb,
Dyrk1a+/+/-
Dp9Tyb
Dp2Tyb
Dp3Tyb
Ts1Rhr
Dp4Tyb
Dp5Tyb
Dp6Tyb
Tt.Ar
-22.64%
-11.99%
7.81%
-10.19%
-9.12%
-10.22%
-6.23%
-12.88%
-1.95%
Ma.Ar
-28.91%
-14.16%
8.80%
-17.78%
-14.11%
-14.39%
-10.57%
-14.72%
-2.26%
Ct.Ar
-16.18%
-9.44%
6.65%
-1.08%
-4.28%
-5.59%
-1.24%
-10.55%
-1.61%
Ct.Th
-3.48%
-3.14%
2.12%
6.32%
1.60%
0.65%
3.39%
-3.72%
-0.59%
Ps.Pm
-9.89%
-5.71%
3.50%
-4.87%
-4.13%
-4.54%
-2.60%
-5.36%
-0.44%
Ec.Pm
-13.55%
-7.34%
4.17%
-7.85%
-5.68%
-5.88%
-4.33%
-6.91%
-1.35%
FEMALE
Dp1Tyb
Dp1Tyb,
Dyrk1a+/+/-
Dp9Tyb
Dp2Tyb
Dp3Tyb
Ts1Rhr
Dp4Tyb
Dp5Tyb
Dp6Tyb
Tt.Ar
-11.54%
-13.12%
6.44%
-3.32%
-6.49%
-4.49%
-0.67%
-4.32%
2.29%
Ma.Ar
-16.09%
-16.78%
7.82%
-5.22%
-11.19%
-6.88%
-0.84%
-4.41%
-1.77%
Ct.Ar
-7.13%
-9.04%
5.07%
-1.42%
-2.34%
-2.23%
-0.52%
-4.23%
5.96%
Ct.Th
-0.01%
-1.31%
1.39%
0.79%
2.51%
1.00%
-0.12%
-1.82%
5.89%
Ps.Pm
-4.94%
-6.26%
3.06%
-1.68%
-2.87%
-2.11%
-0.36%
-1.75%
0.89%
Ec.Pm
-7.77%
-8.13%
4.66%
-2.23%
-5.39%
-3.88%
-0.38%
-2.06%
0.98%
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Table S4. Results (adjusted p values) of two-tail t tests for trabecular and cortical parameters. Significance was declared with an
adjusted p value < 0.05). M = male, F = female, WT = control/wildtype, Dp = triplicated Dp/Ts strain
Dp1Tyb
Dp1Tyb,
Dyrk1a+/+/-
Dp9Tyb
Dp2Tyb
Dp3Tyb
Ts1Rhr
Dp4Tyb
Dp5Tyb
Dp6Tyb
BMD
M WT vs M Dp
0.014
0.986
0.003
0.249
1.000
0.562
0.836
0.024
0.089
F WT vs F Dp
0.444
0.135
0.063
0.822
1.000
0.330
0.851
0.394
0.188
BV/TV
M WT vs M Dp
0.015
1.000
0.002
0.073
1.000
0.609
0.837
0.034
0.054
F WT vs F Dp
0.178
0.077
0.059
0.834
1.000
0.391
0.973
0.371
0.050
Tb.Th
M WT vs M Dp
0.014
1.000
0.068
0.861
1.000
0.478
1.000
0.019
0.920
F WT vs F Dp
0.689
0.471
0.097
0.791
1.000
0.937
1.000
0.368
0.113
Tb.Sp
M WT vs. M Dp
0.304
1.000
0.018
0.007
0.974
0.578
0.764
0.482
0.100
F WT vs F Dp
0.189
0.299
0.180
0.748
0.999
0.295
0.976
0.301
0.341
Tb.N
M WT vs M Dp
0.059
1.000
0.004
0.015
0.829
0.316
1.000
0.211
0.045
F WT vs F Dp
0.272
0.129
0.068
1.000
0.938
0.538
0.854
0.556
0.113
Tt.Ar
M WT vs M Dp
<0.001
0.007
0.068
0.001
0.002
0.009
0.105
<0.001
1.000
F WT vs F Dp
<0.001
<0.001
0.043
0.226
<0.001
0.191
1.000
0.217
0.746
Ma.Ar
M WT vs M Dp
<0.001
0.004
0.085
<0.001
0.002
0.008
0.030
<0.001
0.986
F WT vs F Dp
<0.001
<0.001
0.035
0.236
<0.001
0.239
1.000
0.178
0.588
Ct.Ar
M WT vs M Dp
<0.001
0.052
0.046
0.723
0.058
0.101
0.734
<0.001
0.986
F WT vs F Dp
0.001
<0.001
0.068
0.651
0.178
0.351
0.997
0.305
0.082
Ct.Th
M WT vs M Dp
0.108
0.297
0.288
0.024
0.284
0.822
0.277
0.019
0.860
F WT vs F Dp
0.997
0.466
0.393
0.556
0.122
0.510
0.943
0.165
0.0158
Ps.Pm
M WT vs M Dp
<0.001
0.007
0.101
0.001
0.001
0.006
0.107
<0.001
0.860
F WT vs F Dp
<0.001
<0.001
0.080
0.231
<0.001
0.343
1.000
0.242
0.667
Ec.Pm
M WT vs M Dp
<0.001
0.048
0.113
<0.001
0.004
0.016
0.131
<0.001
1.000
F WT vs F Dp
<0.001
<0.001
0.043
0.211
<0.001
0.567
1.000
0.221
0.698
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information

Table S5. qPCR probes utilized by Transnetyx (Cordova, TN, USA) for genotyping
Probe 1
Probe 2
Probe 3
Probe 4
Name
Forward
Primer
Reporter
Reverse
Primer
Name
Forward
Primer
Reporter
Reverse
Primer
Name
Forward
Primer
Reporter
Reverse
Primer
Name
Forward
Primer
Reporter
Reverse
Primer
Dp1Tyb
3’i17
CGGGCC
TCTTCG
CTATTA
CG
CTGCAA
ACTCTA
AAAGAT
CCGGC
CTCTCT
CCCTGA
GTGCAT
TCTC
5’i16
CCCTAA
GTCCTT
GTCCCT
CACA
CAGTGC
AGATCC
GGCGCG
GCAGTT
GTTTAA
ACTTCT
AGAGAA
TGAGTT
C
Dp9Tyb
3’i17
CGGGCC
TCTTCG
CTATTA
CG
CTGCAA
ACTCTA
AAAGAT
CCGGC
CTCTCT
CCCTGA
GTGCAT
TCTC
5’h04
CTTCTC
TGGACC
AAAGGG
TTCTTG
ACA
CTAGTG
GATCTC
GAGCC
CTATGG
CTTCTG
AGGCGG
AAAGAA
CCA
Dp2Tyb
3’h04
CGGTGC
GGGCCT
CTT
ATTACG
CCAGGG
CGCG
CCCCAC
CCAATG
TCCAAA
GAC
5’i02
GCCTTG
ACTGAG
GACGTT
GA
CGCGCC
GGATCG
AT
GCAGTT
GTTTAA
ACTTCT
AGAGAA
TGAGTT
C
Dp3Tyb
3’i02
CGTTGG
CCGATT
CATTAA
TGCA
CTTAAC
CACACC
CTTACT
CG
GACACA
CCACAT
CACTGA
AACAG
5’i16
CCCTAA
GTCCTT
GTCCCT
CACA
CAGTGC
AGATCC
GGCGCG
GCAGTT
GTTTAA
ACTTCT
AGAGAA
TGAGTT
C
Dp4Tyb
3’i02
CGTTGG
CCGATT
CATTAA
TGCA
CTTAAC
CACACC
CTTACT
CG
GACACA
CCACAT
CACTGA
AACAG
5’c09
GCGTTA
CACACA
GAGCAT
GAAC
CCGGAT
CACACT
CATGTC
G
GGCTTC
TGAGGC
GGAAAG
A
Dp5Tyb
3’c09
CGGGCC
TCTTCG
CTATTA
CG
CACAGC
TTTGAT
CCGGCG
CG
AGCCAG
GCGGTG
CTG
5’b18
CGAACA
ACTCAA
GGGAGG
AAAGAT
C
CGCGCC
AAGCTT
TA
AGAGCA
GAATAG
CAGTTG
TTTAAA
CTTCT
Dp6Tyb
3’b18
CGTTGG
CCGATT
CATTAA
TGCA
ATTTGA
GCTTTG
ATCCGG
CGCG
CCTTCC
TTCATA
ACTGAG
TGTCGT
A
5’i16
CCCTAA
GTCCTT
GTCCCT
CACA
CAGTGC
AGATCC
GGCGCG
GCAGTT
GTTTAA
ACTTCT
AGAGAA
TGAGTT
C
Ts1Rhr
Ts1Rhr
Tg
CCTGAA
GTCCCG
CACACC
ATATCT
CATCAA
TGTATC
GATGCC
A
GCATCA TTATCA
TGTCTT
TTCCGG
GCT
Dp1Tyb,
Dyrk1a+/+/-
3’i17 CGGGCC
TCTTCG
CTATTA
CG
CTGCAA
ACTCTA
AAAGAT
CCGGC
CTCTCT
CCCTGA
GTGCAT
TCTC
5’i16 CCCTAA
GTCCTT
GTCCCT
CACA
CAGTGC
AGATCC
GGCGCG
GCAGTT
GTTTAA
ACTTCT
AGAGAA
TGAGTT
C
Dyrk1a-
1-KO
GGAAGA
CAATAG
CAGGCA
TGCT
CTATGG
GTCTAG
AGCTCA
TG
GTACTT
CATTTC
AGTGTC
GTGTTT
GTT
Dyrk1a-
1-WT
GCGTTT
CTGAAT
CAAGCC
CAGATA
AAGTGC
GGCTGC
TTGAGC
T
TCATTT
CAGTGT
CGTGTT
TGTTCA
TG
Table S6. Results of significant Levene's test and subsequent Welch's F statistic and Games-Howell test. M = male, F = female, WT =
control/wildtype, Dp = triplicated Dp/Ts strain. Highlighted cells indicate significance was lost with the alternative analysis.
Click here to download Table S6
Disease Models & Mechanisms: doi:10.1242/dmm.049927: Supplementary information
Disease Models & Mechanisms • Supplementary information