Triplications of human chromosome 21 orthologous regions in mice result in expansion of megakaryocyte-erythroid progenitors and reduction of granulocyte-macrophage progenitors

Individuals with Down syndrome (DS) frequently have hematopoietic abnormalities, including transient myeloproliferative disorder and acute megakaryoblastic leukemia which are often accompanied by acquired GATA1 mutations that produce a truncated protein, GATA1s. The mouse has been used for modeling DS based on the syntenic conservation between human chromosome 21 (Hsa21) and three regions in the mouse genome located on mouse chromosome 10 (Mmu10), Mmu16 and Mmu17. To assess the impact of the dosage increase of Hsa21 gene orthologs on the hematopoietic system, we characterized the related phenotype in the Dp(10)1Yey/+;Dp(16)1Yey/+;Dp(17)1Yey/+ model which carries duplications spanning the entire Hsa21 orthologous regions on Mmu10, Mmu16 and Mmu17, and the Dp(10)1Yey/+;Dp(16)1Yey/+;Dp(17)1Yey/+;Gata1Yeym2 model which carries a Gata1s mutation we engineered. Both models exhibited anemia, macrocytosis, and myeloproliferative disorder. Similar to human DS, the megakaryocyte-erythrocyte progenitors (MEPs) and granulocyte-monocyte progenitors (GMPs) were significantly increased and reduced, respectively, in both models. The subsequent identification of all the aforementioned phenotypes in the Dp(16)1Yey/+ model suggests that the causative dosage sensitive gene(s) are in the Hsa21 orthologous region on Mmu16. Therefore, we reveal here for the first time that the human trisomy 21-associated major segmental chromosomal alterations in mice can lead to expanded MEP and reduced GMP populations, mimicking the dynamics of these myeloid progenitors in DS. These models will provide the critical systems for unraveling the molecular and cellular mechanism of DS-associated myeloproliferative disorder, and particularly for determining how human trisomy 21 leads to expansion of MEPs as well as how such an alteration leads to myeloproliferative disorder.


INTRODUCTION
Human trisomy 21, the most common human aneuploidy compatible with postnatal survival, occurs in approximately one out of 700-1000 live births [1,2]. This chromosomal anomaly causes a constellation of developmental abnormalities, classified as Down syndrome (DS). The major clinical manifestations of DS include cardiovascular malformations, craniofacial abnormalities, developmental cognitive deficits, hematopoietic abnormalities, and Alzheimer-type neurodegeneration with variable penetrance and onset [3,4].
Among hematopoietic abnormalities, transient myeloproliferative disorder (TMD) occurs in a substantial percentage of children with DS, and many of these patients later develop acute megakaryoblastic leukemia (AMKL) [5]. It has been demonstrated that children with DS who develop TMD and AMKL have acquired somatic mutations in exon 2 of the GATA1 gene on the X chromosome, which lead to the generation of a mutant protein, GATA1s [6,7]. The GATA1s germline mutation is associated with impaired erythropoiesis [8,9].
To unravel the mechanism underlying development of myeloproliferative disorder in DS, genetic models with phenotypes mimicking DS at the molecular, cellular and organ levels are necessary. The mouse has been the most important model organism for DS due to evolutionary conservation between the human and mouse genomes. Regions on human chromosome 21 (Hsa21) are syntenically conserved with regions on mouse chromosome 10 (Mmu10), Mmu16 and Mmu17, which contain 41, 115, and 19 Hsa21 gene orthologs, respectively ( Figure 1). DS is a contiguous gene syndrome [10] and evidence indicates that, for many DS phenotypes, more than one triplicated gene contributes through direct actions and/or interactions of the triplicated genes [11][12][13][14][15][16][17]. Therefore, in a mouse model, triplication of more Hsa21 gene orthologs increases the probability that all the direct actions and/or interactions that have a significant effect on a specific DS phenotype are mimicked.

Generation of Gata1 Yeym2 mice to delete exon 2 of the Gata1 gene
To assess the effect of GATA1s in DS-associated hematopoietic abnormalities, we used gene targeting to delete exon 2 of the Gata1 gene in mouse ES cells. First, we generated ES cell lines that carried the Gata1 Yeym1 mutant allele using the replacement vector pTVGata1. Recombination between pTVGata1 and the Gata1 locus in ES cells led to the replacement of a genomic region containing exon 2 of the gene with a neomycin-resistance gene cassette flanked by two loxP sites (Figure 2A). We then expressed the Cre recombinase transiently in two Gata1 Yeym1 ES cell clones and the neomycin-resistance gene cassette was removed by Cre/loxP-mediated recombination. The excised allele was designated as Gata1 Yeym2 (Figure 2A). Germline transmission was established for four ES cell lines that carried Gata1 Yeym2 . Southern blot analysis was performed to confirm the deletion of the exon 2 region ( Figure 2B). Expression of the mutant allele in the Gata1 Yeym2 mice was confirmed by RT-PCR ( Figure 2C). CBCs revealed reduced counts of red blood cells (RBCs) starting at 3 months of age (p < 0.001) and lowered hemoglobin concentrations (HGB) starting at 9 months of age (p < 0.05) in both Dp(10)1;Dp(16)1;Dp(17)1 mice and Dp(10)1;Dp(16)1;Dp(17)1;Gata1 Yeym2 mice ( Figure 3A-3B), which indicated that both mutants developed anemia. Dp(10)1;Dp(16)1;Dp(17)1;Gata1 Yeym2 mice showed much lower RBC counts (p < 0.01) and HGB concentrations (p < 0.01) than Dp(10)1;Dp(16)1;Dp(17)1 mice ( Figure  3A-3B), suggesting that the Gata1s mutation causes more severe anemia by further reducing the RBC counts and HGB concentrations.

Triplication of the Hsa21 orthologous region on Mmu16 results in expansion of MEPs and reduction of GMPs
In contrast to Dp(10)1;Dp(16)1;Dp(17)1 mice, Ts65Dn mice have repeatedly shown reduction of MEPs and expansion of GMPs [19,39]. Ts65Dn and Dp(16)1 carry 3 copies of the identical set of 100 Hsa21 gene orthologs on Mmu16, so we speculated that Dp(16)1 mice would likely also show reduction of MEPs and expansion of GMPs. We further speculated that expansion of MEPs and reduction of GMPs in Dp(10)1;Dp (16)1;Dp (17)1 mice may be caused by Dp(10)1 and/or Dp (17)1. To test these possibilities, we carried out phenotypic analysis of Dp(16)1 mice. To our surprise, the hematopoietic phenotypes of Dp(16)1 mice are almost identical to those of Dp(10)1;Dp (16)1;Dp (17)1 mice, including expansion of MEPs and reduction of GMPs ( Figure 9 and Supplementary Figures 1-4). These results suggest that the causative gene(s) responsible for expansion of MEPs and reduction of GMPs are located in the Hsa21 orthologous region on Mmu16. These results also suggest  (17)1;Gata1 Yeym2 mice (n = 8, 5, 5, 6, 5 at 3, 6, 9, 12, 15 months of age, respectively). www.impactjournals.com/oncotarget that triplicated genes which differ between Dp(16)1 and Ts65Dn mice are responsible for diametrically opposite effect on MEPs and GMPs.

DISCUSSION
In this study, we analyzed the hematopoietic phenotype of the Dp(10)1;Dp(16)1;Dp(17)1 mouse model of DS, which contains duplications spanning the entire Hsa21 orthologous regions on Mmu10, Mmu16, and Mmu17 [15,28]. These mice exhibited anemia, macrocytosis, thrombocytosis, megakaryocytosis, splenomegaly, and perturbed hematopoietic stem cells and progenitor cell compartments including expanded MEP and reduced GMP populations ( Figure 10). The subsequent analysis of Dp(16)1 mice indicated that these phenotypes are the consequences of the triplication of Hsa21 gene orthologs on Mmu16.
Hematopoietic phenotypes have been characterized in two other major mouse models of DS, Ts65Dn and Tc1 [18,19]. Ts65Dn mice carry a triplication of a genomic fragment of approximately 13.5 Mb extending from miR-155 to the telomere on Mmu16 and encompassing 100 Hsa21 gene orthologs, which accounts for approximately 57% of the Hsa21 gene orthologs ( Figure 10) [22,40,41]. Ts65Dn mice also carry a triplicated genomic fragment of approximately 10 Mb extending from the centromere to 1700010I14Rik on Mmu17 which contains 36 Hsa6 gene orthologs [40,41]. Ts65Dn mice, Dp(10)1;Dp(16)1;Dp(17)1 mice and Dp(16)1 mice share some major features of the hematopoietic phenotype, such as anemia, macrocytosis and myeloproliferative disease characterized by thrombocytosis, megakaryocytic hyperplasia and a distorted hematopoietic stem cell compartment [19]. However, these three models also exhibit important differences in the hematopoietic features. Compared with the wild-type controls, the percentages of MEPs and GMPs in the myeloid progenitor compartments are increased and decreased, respectively, in both Dp(10)1;Dp(16)1;Dp(17)1 and Dp (16)1 mice ( Figures 8B, 9B). In contrast, the percentages of MEPs and GMPs are decreased and increased, respectively, in  Ts65Dn mice (Figure 10) [19]. Because MEPs are the progenitors of megakaryocytes, this specific observation in Dp(10)1;Dp(16)1;Dp(17)1 and Dp(16)1 mice may better reflect the abnormal hematopoiesis underlying DSassociated myeloproliferative disorder. Such reasoning is also supported by the observation that the percentages of MEPs and GMPs are increased and decreased, respectively, in the livers of fetuses carrying human trisomy 21 [42][43][44]. Tc1 mice carry a Hsa21 with three major deletions and other genomic alterations, including duplications and gene mutations, which are likely caused by irradiation of Hsa21 during microcell-mediated chromosome transfer [26,45]. The random loss of the transferred human chromosome during mouse development resulted in variable levels of mosaicism of the extra chromosome in different tissues [26]. Tc1 mice, Dp(10)1;Dp(16)1;Dp(17)1 and Dp(16)1 mice share some important features of the hematopoietic phenotype, such as anemia, macrocytosis, splenomegaly, and an increase in the numbers of megakaryocytes in the spleen [18]. However, unlike Dp(10)1;Dp(16)1;Dp(17)1 and Dp(16)1 mice, Tc1 mice did not exhibit significant abnormalities in hematopoietic stem cell and myeloid  progenitor compartments, such as MEPs and GMPs [18]. The differences in these phenotypic features could be caused by the Hsa21 genes deleted or mutated on the transchromosome, by transchromosome mosaicism, and/ or by species-specific differences between the proteins encoded by human and mouse chromosomes. Therefore, Dp(10)1;Dp(16)1;Dp(17)1 mice exhibit more important DS-patient-associated hematopoietic abnormalities than other major mouse models of DS carrying an additional copy of Hsa21 orthologous regions, particularly expansion of MEPs and reduction of GMPs. These mice will provide a more desirable system for a better understanding of disease processes and the underlying pathogenic mechanism.
It is a surprise that Dp(16)1 and Ts65Dn exhibit opposing MEP and GMP phenotypes since they share the same 100 triplicated genes. However, there are additional triplicated genes which differ between the two models: Dp(16)1 carries a triplication of 15 additional Hsa21 gene orthologs on Mmu16, while Ts65Dn carries a triplication of 36 additional Hsa6 gene orthologs on Mmu17. Therefore, these genetic differences may underlie the opposite phenotypes related to MEPs and GMPs.
MEPs are the myeloid progenitors of megakaryocytes, and therefore the expansion of MEPs observed in human DS and Dp(10)1;Dp(16)1;Dp(17)1 and Dp (16)1 mice could be a major logical cellular event leading to myeloproliferative disorder in humans and mouse models ( Figure 10). However, Ts65Dn mice also exhibit myeloproliferative disorder ( Figure  10) [19] despite exhibiting a reduction of MEPs. One possible reason for this is that differentiation of MEPs to megakaryoblasts may be further accelerated in Ts65Dn, which may then lead to a relative reduction in the number of MEPs. Regardless of the possible reasons, reduction of MEPs detected in Ts65Dn mice deviates from the hematopoietic abnormality observed in human DS. Thus, unraveling why Dp (16)1 and Ts65Dn are different in MEP-and GMP-related phenotypes may help to improve our understanding of the mechanism underlying myeloproliferative disorder in DS.
Expansion of MEPs and reduction of GMPs were also observed in compound transgenic mice which carry a human ERG transgene and a germline Gata1s mutation [46]. The ERG transgene was driven by regulatory elements from the mouse vav gene. There are several potential reasons why this compound mutant may not accurately model the related pathogenic process in human DS. First, the cell-and temporal-specific expression patterns of vav and Erg may be different, so the expression pattern of the ERG transgene may deviate from that of the endogenous Erg. Second, the levels of expression between the wild-type allele of Erg and the transgenic ERG may be also different. A more desirable approach to assess the contribution of an individual Hsa21 gene ortholog to a phenotype is to use a "subtractive strategy", which involves normalizing the dosage of a Hsa21 gene ortholog by compounding its nonfunctional allele in a segmental chromosomal alteration mouse model of DS. When this strategy was applied to Erg in Ts65Dn mice, reducing Erg from 3 copies to 2 copies actually led to a significant expansion of MEPs [39], which indicates that the increased Erg copy number in Ts65Dn mice is responsible for the reduction of MEPs. This result suggests the ERG/Gata1s transgenic mice described above may not be appropriate to model the MEP expansion in human DS.
In this study, we demonstrated expansion of MEPs and reduction of GMPs in major mouse models of DS with human trisomy 21-associated segmental genomic alterations for the first time, mirroring the corresponding changes of these myeloid progenitors in DS. Thus, further study of these models may lead to identification of the causative gene(s) for expansion of MEPs, which will be the key to a better understanding of the pathogenic mechanism of DS-associated myeloproliferative disorder.
In particular, these models may provide further insight into how trisomy 21 leads to expansion of MEPs and how the quantitative alteration of MEPs contributes to the development of myeloproliferative disorder in DS.

Generation of the Gata1 Yeym2 mutant mice
The mutant mouse strain was generated using genetargeting in ES cell to delete exon 2 of the Gata1 gene which resulted in the expression of a truncated mutant protein, GATA1s. The targeting vector, pTVGata1, was constructed by inserting two homologous regions into a Bluescript KS plasmid that contained a PGKneobpA (loxP) cassette in the polylinker (a gift from Dr. Richard Behringer). The first homology region was amplified by primers N1033 (5′-TGCGGCCGCTGGCCCTGATCTC AGCTCAGAATA-3′) and B3614 (5′-TGGATCCTACC GCCCCATTTGTACCAATCCT-3′) to produce a 2598-bp fragment in between intron 2 and intron 5. After cloning the first fragment into pCR-XL-TOPO ® using TOPO ® XL PCR Cloning Kit (Invitrogen), the homology arm was excised out by digesting with NotI and BamHI. The second homology region was amplified by primers H4204 (5′-TTAAGCTTCCACCACCGCCTGGCTCTTC-3′) and K7468 (5′-TGGTACCAGTGCTGGGTTTAAAGGTG TGC-3′) to produce a 3286-bp fragment inside the first intron. After cloning the second fragment into pCR-XL-TOPO ® , the homologous arm was excised out by digesting with HindIII and KpnI. The excised fragments were then sequentially inserted between NotI and BamHI site as well as between HindIII and KpnI sites of the ploylinkers on either side of the PGKneobpA(loxP) cassette. The resulting targeting vector, pTVGata1, was linerized with KpnI and electroporated into AB2.2 ES cells. The targeted allele, Gata1 Yeym1 , was obtained by G418 selection as described previously [52,53]. Cre/loxP-mediated marker Solid bar, the wild-type controls (n = 6); horizontal-lined bar, Dp(16)1 mice (n = 5). * p < 0.05; ** p < 0.01. excision [54] was carried out and the final excised allele, Gata1 Yeym2 , was identified by Southern blot analysis with BamHI digested DNA. The probe used in the Southern blot hybridization was amplified by primers Probe-U (5′-GGG GGACATTGGGACAGAATAGTT-3′) and Probe-L (5′-AT TCCCCCACGCTCCTATCTCAT-3′), which is mapped to the genomic region downstream of the homologous region in Intron 1 (Figure 2).

Mice
Gata1 Yeym2 mice were first established in a 129SvEvxC57BL/6JF1 strain background and then backcrossed to C57BL/6J mice for five generations. Dp(10)1, Dp(16)1, and Dp(17)1 mice were generated using recombinase-mediated genome engineering, which carry the duplications spanning the entire Hsa21 orthologous regions on Mmu10, Mmu16 and Mmu17, respectively [15,28]. Dp(10)1, Dp(16)1 and Dp (17)1 mice were backcrossed to C57BL/6J mice for five generations before being used to generate Dp(10)1;Dp(16)1;Dp(17)1 mice. The latter was crossed to Gata1 Yeym2 mice to establish Dp(10)1;Dp(16)1;Dp(17)1;Gata1 Yeym2 mice. All mice were maintained at a temperature-and humidity-controlled animal facility. Gata1 is located on X chromosome. To facilitate appropriate interpretation, we have used only male mice with different genotypes in all the experimental procedures of this study, so only single copy of the Gata1 or Gata1s allele is present in all mice used in the procedures. All the experimental procedures performed were approved by the Institutional Animal Care and Use Committee at Roswell Park Cancer Institute.

RT-PCR
Total RNA was extracted from the livers of E14.5 embryos using TRIzol Reagent and PureLink RNA Micro kit (Invitrogen Corp., Carlsbad, CA). One µg of the total RNA from each embryo was used to synthesize cDNA by Superscript version III reverse transcriptase (Invitrogen Corp., Carlsbad, CA). To detect the expression of Gata1 and Gata1s, cDNA was analyzed by PCR amplification with the primers GATA1SRTF1 (TCAGAGGCCAAGGCCAGTGAGGACT) and GATA1 SRTR1 (TTGCCATAGGCCCAGCTAGCATAAGG). The lengths of the PCR products are 410-and 171-bp, reflecting to the wild-type and the Gata1 Yeym2 alleles, respectively ( Figure 2C).

Histology
Bones from femur or sternum and spleen were fixed in 10% buffered formalin and were further processed by the Roswell Park Cancer Institute Pathology Core Facility. Bone marrow samples were decalcified prior to processing. Paraffin sections of spleen and bone marrow were stained with hematoxylin and eosin. Slides were photographed on an Olympus BX41 microscope with a DP70 camera and captured with DP Controller software version 1.2.1.108 (Olympus Optical Co., Japan).

Flow cytometry
Single-cell suspensions from bone marrow or spleen were prepared by passing the tissues through 70 μm cell strainers (BD Pharmingen, San Diego, CA) and red blood cells were lysed in RBC lysis buffer (150 mM NH 4 Cl, 10 mM KHCO 3 , 0.1 mM EDTA). One million cells from each sample were then labeled with the following, anti-CD41-FITC (BD Pharmingen) or anti-TER119-APC-Cy7 (BD Pharmingen) for 30 minutes on ice. Labeled cells were washed in autoMACS Rinsing Solution (Miltenyi Biotec, Aubum, CA) with the addition of 0.5% BSA, fixed with 1% paraformaldehyde in PBS overnight and then analyzed on a LSRII cytometer (BD Biosciences, San Jose, CA). For analysis of Lineage -Sca1 + c-Kit + (LSK) cell and myeloid progenitor cells, ten million bone marrow cells were lineage depleted by using the Lineage Cell Depletion Kit (Miltenyi Biotec) according to the manufacturer's protocol with the addition of the anti-IL-7R-biotin antibody (BD Pharmingen). After magnetic separation of the lineage cells through the MACS MS column, the lineage negative cells were incubated with streptavidin-PE-Cy5.5 (eBioscience, San Diego, CA), anti-Sca1-PE-Cy7 (BD Pharmingen), anti-c-Kit-APC (BD Pharmingen), anti-FcγR-PE (BD Pharmingen) and anti-CD34-FITC (BD Pharmingen) on ice for 30 minutes and analyzed on a LSRII cytometer after washing and fixation as described above.

Statistical analysis
Statistical analysis was performed using Student t-test and the non-parametric Mann-Whitney U-test. The level of statistical significance was set at p = 0.05. Data are presented as Mean ± SEM.

CONFLICTS OF INTEREST
None of the authors has any conflicts of interest.

FUNDING
This study was supported in part by grants from Roswell Park Alliance Foundation, Louis Sklarow Memorial Fund, American Cancer Society, Association for Research of Childhood Cancer, the Children's Guild Foundation, the National Institutes of Health (R01HL91519, R01NS66072, R21GM114645, P30CA16056) and the National Natural Science Foundation of China (81370745).