Novel spontaneous myelodysplastic syndrome mouse model

Abstract Background Myelodysplastic syndrome (MDS) is a group of disorders involving hemopoietic dysfunction leading to leukemia. Although recently progress has been made in identifying underlying genetic mutations, many questions still remain. Animal models of MDS have been produced by introduction of specific mutations. However, there is no spontaneous mouse model of MDS, and an animal model to simulate natural MDS pathogenesis is urgently needed. Methods In characterizing the genetically diverse mouse strains of the Collaborative Cross (CC) we observed that one, designated JUN, had abnormal hematological traits. This strain was thus further analyzed for phenotypic and pathological identification, comparing the changes in each cell population in peripheral blood and in bone marrow. Results In a specific‐pathogen free environment, mice of the JUN strain are relatively thin, with healthy appearance. However, in a conventional environment, they become lethargic, develop wrinkled yellow hair, have loose and light stools, and are prone to infections. We found that the mice were cytopenic, which was due to abnormal differentiation of multipotent bone marrow progenitor cells. These are common characteristics of MDS. Conclusions A mouse strain, JUN, was found displaying spontaneous myelodysplastic syndrome. This strain has the advantage over existing models in that it develops MDS spontaneously and is more similar to human MDS than genetically modified mouse models. JUN mice will be an important tool for pathogenesis research of MDS and for evaluation of new drugs and treatments.

Research on the physiological mechanisms and treatment of MDS still faces significant challenges. One of the main reasons is that most of the mutations that give rise to MDS cause disease with a short survival period, so patients usually do not survive to adulthood. 5 Thus, research on the disease usually only focusses on the cellular level. Animal models are powerful tools for modeling and studying human diseases, and are very useful preclinical platforms for studying problems that cannot be easily (or at all) solved in the clinic. 6 Current MDS models are usually produced by genetic modification (eg by introducing mutations as transgenes), by chemical induction, or by xenotransplantation. [6][7][8][9][10] However, some of these models' phenotypes cannot be maintained stably for a long time and some cannot simulate the pathogenesis of MDS from abnormal hematopoietic stem cells. MDS models induced experimentally can be quite different from the actual human clinical condition. 5 A spontaneous MDS model would be more similar in onset to the human MDS process, and thus more helpful to translating research results to humans. Therefore there is an urgent need to establish a spontaneous animal model in order to analyze the pathogenesis of MDS and the process of transformation to AML.
The Collaborative Cross (CC) is a family of mouse strains produced by selective breeding from eight genetically diverse founder strains, 11 and is the result of a project formally initiated in 2004 at The Jackson Laboratory. Each CC line originates from an independently breeding funnel so that every recombination site in the CC population is uniquely generated, and the CC strains harness the common genetic diversity of the mouse species, [11][12][13] and can be used to identify genes mediating complex diseases and traits, 14,15 as well as providing models for human diseases, such as osteoporosis, diabetes complications, viral infections and a variety of spontaneous tumors. [16][17][18][19] We investigated a panel of CC strains 20 and identified one, JUN, that appears to develop MDS spontaneously. As with the other CC strains, JUN was developed from an independently breeding funnel, and had genetic information of eight founder strains.

| Blood samples
JUN mice (6-8 weeks) were taken out of the SPF environment and maintained in a barrier environment for 1 month, and then compared to 10-12 weeks old C57BL/6J mice. The mice were anesthesized by intraperitoneal injection of ketamine, xylazine, and benpiate hydrochloride (0.0375/0.0375/0.000125 mg/g body weight). Blood samples were then taken, stored in pediatric plastic tubes with the anticoagulant BDTM K 2 EDTA and analyzed using automated hematological analysis equipment (BAYER ADVIA 2120).

| Mouse bone marrow cell isolation
Femurs and tibias were taken, and marrow cells were collected by flushing with PBS. The bone marrow cells were then analyzed by flow cytometry and blood smear.

| Bone marrow and blood smear preparation
To prepare the blood smear, a drop of blood was taken from the mouse tail vein, dropped onto a glass slide, and then spread across the slide using another glass slide held at a 30° angle and pushed at a constant speed. The bone marrow smear was taken from the sternum of a mouse. The bone marrow cells in the sternum were squeezed onto a pre-prepared glass slide, dripped with serum, and the droplet was spread across the slide in the same way as the blood smear. All smears were then stained with Wright-Giemsa (Jiancheng Biotech, D010).

| Flow cytometry
After red blood cell lysis, bone marrow cells for analysis by flow cytometry were stained with a cocktail of antibodies for 30 minutes.
Quantitative RT-PCR was carried out using an ABI Prism 7900 Sequence detection system (BioRad CFX Connect). Relative levels were normalized to that of GAPDH.

| Library construction and whole genome sequencing
A 1 μg sample of genomic DNA was randomly fragmented by Covaris.
Fragmented DNA was selected using an Agencourt AMPure XP-Medium kit to yield an average size of 200-400 bp. The selected fragments were, through end-repair, 3′ adenylated, adapter-ligated, and PCR amplified and the products were recovered using the AxyPrep Mag PCR clean up Kit. An aliquote of the PCR products was taken for hybridization with BGI Hybridization and Wash kits. After that, the AxyPrep Mag PCR clean up Kit was used to recover the products as before. The double stranded PCR products were heat denatured and circularized using a splint oligo sequence to form single stranded circular DNA (ssCir DNA) as the final library, which was assessed by QC. The library was amplified to make DNA nanoballs (DNBs) which have more than 300 copies of one molecular. The DNBs were loaded into the patterned nanoarray and pair-end 100 base reads were generated and sequenced by combinatorial Probe-Anchor Synthesis (cPAS) on the BGISEQ-500 platform (BGI-shenzhen, China).

| Statistical analysis
For each set of assays, three independent experiments were performed. Results are expressed as the means ± SEM. Statistical significance was calculated using the unpaired Student's t test. P values of less than .05 were considered significant. All tests were carried out using Prism software version 5 (GraphPad Software), and the gene sequencing data was analyzed using Integrative Genomics Viewer (IGV).

| JUN mice have a hematological disease-like phenotype
Most mice of the albino JUN strain grew well in the SPF environ- The unfavorable phenotype of the JUN mice is similar to the susceptibility to infection of patients with blood diseases, 22 so we initially thought that JUN mice had spontaneous myelocytic leukemia.

| Lesions in multiple organs of JUN mice
Histological analysis was conducted on various organs of JUN mice that had died naturally after living in a non-SPF environment.
Lesions were identified in the kidney, liver, spleen, and bone mar- Furthermore, the liver also exhibited extra-medullary hematopoiesis with random infiltrating basophilic mononuclear lymphoid cells ( Figure 2C,D). Vessels in the pulmonary parenchyma ( Figure 2G,H) and other organs also showed basophilic premature mononuclear cell infiltration ( Figure 2B-D). Here, high levels of myeloperoxidase (MPO), a marker of both benign and neoplastic myeloid cells, were observed in many organs ( Figure 2I,J). 23,24 Individual cells scattered throughout the liver and spleen were labeled; many were within areas that appeared to be myeloid hematopoietic. This was consistent with our initial suspicions that these JUN mice had myeloid leukemia.

| Abnormal development of white blood cells in JUN mice
To determine whether the JUN mouse strain progresses to myeloid leukemia, samples were collected for hematological analysis from 14-week-old mice that had lived in the non-SPF environment for 2 months. WBC counts were lower in JUN mice compared with C57/ BL6 mice, manifesting as cytopenia. The lower counts were generally due to lower numbers of neutrophils, monocytes, and lymphocytes ( Figure 3A). Although the loss of WBCs does not affect the diagnosis of myeloid leukemia, we also found Pseudo-Pelger-Hüet cells in JUN blood ( Figure 3B); these cells are often seen in MDS. 25 Furthermore, in peripheral blood smears, we also found the abnormal development of neutrophils ( Figure 3C), which is a major criterion in the assessment of dysplasia in MDS. 25 Evaluation of the number and morphology of blasts in peripheral blood is the only peripheral blood parameter incorporated into the WHO classification of MDS, and is therefore critical. 26 Firstly, we found that immature leukocytes accounted for an increased proportion of leukocytes ( Figure 3A); the next most It is well known that MDS can transform into acute myeloid leukemia. In the initial period, these mice showed no any abnormalities, so, combined with phenotype and pathological analysis, it was concluded that JUN mice are likely to have MDS, with later progression to acute myeloid leukemia. 25,27

| Nuclear abnormalities were found in JUN bone marrow cells
To obtain more evidence that JUN mice suffer from MDS similar to human diseases, an appropriate morphologic assessment of dysplasia in bone marrow smears was essential, as this is a diagnostic criterion of MDS. 32 The diagnosis of MDS can also be established in a markedly cytopenic mouse if a typical (MDS-related) nuclear anomaly is found. 33,34 Therefore, we compared JUN cells with human bone marrow smears collected from the hematology of Peking University People's Hospital. It was found that the characteristic abnormal nuclear abnormalities observed in human MDS bone marrow, including Pseudo-Pelger-Hüet cells ( Figure 5A), stab granulocytes ( Figure 5B), segmented form granulocytes ( Figure 5C), megaloblastic changes ( Figure 5D,E), and micromegakaryocytes ( Figure 5F), were also found in JUN.

| A high CD34 + cell count has been found in JUN mice
Currently, there is little consensus as to which cell type should be defined as an MDS "stem cell". At present, CD34 + cells possessing a clonal marker comprise the best prototype. 35 was also reduced in JUN compared with C57BL/6J ( Figure 7B). The severe macrocytic anemia in del(5q) MDS patients has been linked to haploinsufficiency of (Rps14), 38,39 and it has been described in CD34 + cells. 40 Based on this, some conditional Rps14 knockout MDS models have been generated, that have developed a progressive anemia. 39 Our research found that, unlike the conditional Rps14 knockout MDS models whose Rps14 was excised in hematopoietic cells, Rps14 was not missing in JUN mice but the expression level was significantly decreased. Moreover, these two models showed different MDS phenotypes: JUN mice mainly showed leukopenia, while the reticulocyte count of Rps14 haploinsufficient mice decreased precipitously.
Therefore, it is likely that the mutation of Rps14 is one of the factors contributing to the spontaneous MDS exhibited by the JUN strain.  Mouse models can provide a platform for pre-clinical testing, including testing novel small molecules (singly or in combination) and can even be used to study how HSCT works for patients with MDS. CC mice whose phenotypic diversity is on a par with human populations have proved very useful for providing novel animal models of human disease. 16  can be tested in the spontaneous model to ensure efficacy before recruiting patients. 45,46 By testing in JUN mice, the mechanism of drugs currently used can be clarified, to provide a deeper understanding of the underlying mechanisms in heterogeneous diseases, and hopefully improve our ability to treat MDS patients.

| D ISCUSS I ON
Due to the high spontaneous incidence of MDS in JUN mice and the high frequency of conversion to AML, studies using these mice would be more stable and reliable than other models.
In addition, further study of the Rps14 gene to explore its role in the hematopoietic system would be of great interest.

CO N FLI C T O F I NTE R E S T
The authors declared no conflict of interest.

AUTH O R CO NTR I B UTI O N S
The