A gain‐of‐function mutation in microRNA 142 is sufficient to cause the development of T‐cell leukemia in mice

Abstract MicroRNAs (miRNAs) play a crucial role in regulating gene expression. MicroRNA expression levels fluctuate, and point mutations and methylation occur in cancer cells; however, to date, there have been no reports of carcinogenic point mutations in miRNAs. MicroRNA 142 (miR‐142) is frequently mutated in patients with follicular lymphoma, diffuse large B‐cell lymphoma, chronic lymphocytic leukemia (CLL), and acute myeloid leukemia/myelodysplastic syndrome (AML/MDS). To understand the role of miR‐142 mutation in blood cancers, the CRISPR‐Cas9 system was utilized to successfully generate miR‐142‐55A>G mutant knock‐in (Ki) mice, simulating the most frequent mutation in patients with miR‐142 mutated AML/MDS. Bone marrow cells from miR‐142 mutant heterozygous Ki mice were transplanted, and we found that the miR‐142 mutant/wild‐type cells were sufficient for the development of CD8+ T‐cell leukemia in mice post‐transplantation. RNA‐sequencing analysis in hematopoietic stem/progenitor cells and CD8+ T‐cells revealed that miR‐142‐Ki/+ cells had increased expression of the mTORC1 activator, a potential target of wild‐type miR‐142‐3p. Notably, the expression of genes involved in apoptosis, differentiation, and the inhibition of the Akt–mTOR pathway was suppressed in miR‐142‐55A>G heterozygous cells, indicating that these genes are repressed by the mutant miR‐142‐3p. Thus, in addition to the loss of function due to the halving of wild‐type miR‐142‐3p alleles, mutated miR‐142‐3p gained the function to suppress the expression of distinct target genes, sufficient to cause leukemogenesis in mice.


| INTRODUC TI ON
miRNAs are small noncoding RNAs that consist of 20-25 nucleotides that are not translated into proteins. miRNAs become mature singlestranded RNAs in the cell, promote mRNA degradation, and inhibit protein translation. Thus, they are associated with various regulatory networks. Recent reports have indicated that single nucleotide mutations, 1-3 methylations, 4,5 and the upregulation or downregulation of miRNA expression have been observed in various cancers 6-9 ; however, no studies have reported carcinogenesis caused by single nucleotide mutations in miRNAs.
We focused on one of the miRNAs, microRNA 142 (miR-142), which is a highly conserved miRNA, abundantly expressed in the hematopoietic system and reportedly regulates the differentiation and function of megakaryocytes, 10 erythrocytes, 11 dendritic cells, 12 and most frequently T-cells. [13][14][15][16] Recently, miR-142 has been reported to mutate frequently in hematologic cancers. Single nucleotide mutations within the miR-142 gene, including the miR-142 precursor, have been documented to occur throughout the gene in various forms of leukemia, such as follicular lymphoma, 17,18 chronic lymphocytic leukemia, 19 diffuse large B-cell lymphoma, [20][21][22][23] and AML/MDS. 2, 3 We focused on the miR-142-55A>G mutation, which is frequently found in patients with AML/MDS. 2,3 The miR-142 sequence is 100% identical between humans and mice. In addition, homozygous knockout mice with the entire length of miR-142 deleted show abnormal differentiation of their immune system cells, but there have been no reports on heterozygous mice. 14,24 However, most of the miR-142 abnormalities in AML/MDS patients are single nucleotide substitutions and heterozygous. Therefore, we hypothesized that the heterozygosity of a single nucleotide mutation within the seed sequence of miR-142-3p is responsible for the development of leukemia.
In this study, to understand the role of the miR-142 mutation in leukemia, we utilized the CRISPR-Cas9 system to generate miR-142 mutation Ki mice to simulate the most frequently identified mutation in patients. We analyzed miR-142-Ki heterozygous mice and showed that single nucleotide mutations in miRNAs are associated with carcinogenesis.

| Animal welfare
Animal experiments were carried out in accordance with the Regulations for Animal Experiments at our university and and CD-1 mice were purchased from Clea Japan (Tokyo, Japan) and Charles River Laboratories Japan (Kanagawa, Japan), respectively.

| Generation of the miR-142-55A>G mouse line
The first trial to produce the point mutation was performed using ES cells. pX330-sgRNA9 was constructed by cloning a set of sgRNA oligos (5′-caccGCACT ACT AAC AGC ACTGGA-3′ and 5′-aaacTCCAG TGC TGT TAG TAGTGC-3′) into a BbsI-digested pX330 vector (#42230; Addgene). The 6NK-7 ES cells, which were established from C57BL/6N mice in our laboratory, and were transfected with 2 μg of pX330-sgRNA9 and 3 μg of ssODN with the mutated sequence (5′-CAGAC AGA CAG TGC AGT CAC CCA TA A AGT AGA A AG CAC TAC   TAA CAG CAC TGG AGG GTG TGG TGT TTC CTA CTT TAT GGA TGA GTG CAC TGT GGGCTTCGGAGACCACGCCACGCCGCGGC-3′) using the Xfect mESC Transfection Reagent (Clontech Laboratories), following the manufacturer's protocol. After 2 days, the transfected ES cells were plated at clonal density (3000 pieces/dish) into a 10-cm dish and cultured for 7 days. The culture conditions of the CO 2 incubator were set at 6.2% (100% humidity). On day 8, colonies were picked and stocked. Subsequently, PCR was performed with primers for miR-142-S (5′-GGGAA GAA GGT TAC AAA GAGG-3′) and miR-142-R (5′-TGAGA GAT GCT CAC CTG TTTC-3′) and the resulting product was sequenced. We could not obtain clones with the targeted point mutation but obtained clone #30, which had an 8-bp deletion. Clone #30 ES cells were aggregated with morulae from ICR mice. Chimeric mice were mated with C57BL/6N females.

| Quantification of miRNA expression
Total RNA was extracted from mouse bone marrow (BM) samples using ISOGEN (Nippon Gene). cDNA synthesis was performed from 10 ng of total RNA using a TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems), according to the manufacturer's instructions. Thereafter, miRNA expression levels were quantified using a TaqMan MicroRNA Assay (Applied Biosystems) for miR-142-5p in a Thermal Cycler Dice TP800 Real-Time PCR system (TaKaRa). Each assay was performed in technical triplicates. miR-142 expression levels were normalized to Actb/β-actin as an internal control. The average threshold cycle (C t ) for three replicates per sample was used to calculate ΔC t . Finally, the relative quantification for gene expression was calculated using the 2 −ΔΔCt method.

| Bone marrow transplantation
Whole BM (5 × 10 6 cells) from CD45.2 + donors was injected into the tail veins of irradiated CD45.1 + recipient mice. Irradiation was performed using an X-ray irradiator delivering 8.5 Gy. For leukemia modeling, whole BM (1 × 10 6 cells) from moribund donors was injected into irradiated (5.5 Gy) CD45.1 + recipient mice. Animals that showed no engraftment of donor cells were excluded from further analysis. Mice exhibiting declining health status were sacrificed, and tissues were taken for analysis.

| Flow cytometry and antibodies
Flow cytometry and cell sorting were performed using antihuman or antimurine antibodies, as indicated in Table S1. The lineage mixture solution contained biotin-conjugated anti-Gr1, B220, CD4, CD8α, Ter119, and IL-7Rα antibodies. All flow cytometric analyses were performed on a FACSVerse or FACSCantoII cytometer (BD Biosciences), and cell sorting was performed on a FACSAria II cytometer (BD Biosciences).

Giemsa staining
BM cells were cytospun using a Shandon Cytospin 3 Cytocentrifuge (Block Scientific) at 100 × g for 3 min. Next, air-dried slides were stained in May-Grünwald solution (Sigma-Aldrich, Tokyo, Japan) for 5 min. After a brief wash with water, the slides were stained in Giemsa solution (Sigma-Aldrich) for 20 min. The slides were finally washed in distilled water and allowed to air dry before microscopic analysis.

| Quantitative RT-PCR
Total RNA was extracted from mouse BM samples and hematopoietic cells using ISOGEN (Nippon Gene), and 500 ng of RNA was converted to cDNA using a PrimeScript 1st strand cDNA Synthesis Kit (6110A; TaKaRa). cDNA aliquots were used for quantitative RT-PCR (RT-qPCR).
For RT-qPCR, TB Green Premix Ex Taq II (RR820A; TaKaRa) was used as the Taq polymerase, and reactions were performed using the intercalator method in a Thermal Cycler Dice TP800 Real-Time PCR system (TaKaRa). In experiments using relative quantification, expression levels were normalized to Actb/β-actin expression. The sequences of all primers used in this study are listed in Table S2.

| Statistical analysis
All statistical tests were performed using GraphPad Prism version 9 (GraphPad Software). The significance of differences was measured using an unpaired two-tailed Student's t-test, Mann-Whitney nonparametric test, or chi-squared test. A p-value <0.05 was considered significant.

| Generation of miR-142-55A>G mice using the CRISPR-Cas9 system
To determine whether a single nucleotide mutation in the miR-142-3p seed sequence (miR-142-3p-55A>G), 2,3 which has been recurrently detected in patients with AML ( Figure 1A), drives leukemogenesis, the same mutation was introduced into the murine  Figure 1B). MicroRNAFold, a web server that predicts the hairpin structure of miRNAs in the genome using the ab initio method, revealed that the insertion of the miR-142-55A>G mutation forms the hairpin structure with equal efficiency ( Figure S1a,b). 26,27 A TaqMan miRNA assay was performed to determine the actual miRNA expression levels. In this assay, the miR-142-5p kit was applied to determine the miR-142 expression level because the miR-142-55A>G mutation prevents precise assessments with the miR-142-3p kit. miRNA expression levels were confirmed to be equal to those of the WT, even when replaced by miR-142-55A>G ( Figure S1c). We designed forward primers for the position where the single nucleotide mutation was inserted. Genotyping confirmed that the mutated mice only had the designated single nucleotide mutation, which was confirmed by Sanger sequencing ( Figure 1C,D).

| miR-142-55A>G mutant mice show abnormal lymphocyte cell differentiation
According to Mendelian ratios, the number of miR-142-55A>G mutant Ki mice generated was significantly lower for homozygous mice Taken together, T-cell differentiation in PB was affected by the miR-142-55A>G mutation, even in heterozygous individuals.

| miR-142-55A>G mutant mice develop CD8 + T-cell leukemia
To determine whether the miR-142-55A>G mutation in blood cells drives the development of leukemia, we transplanted BM cells iso- which showed impaired hematopoiesis phenotypically that was comparable with that previously reported for miR-142-KO mice. 24,28 Moribund Ki/+ and Ki/Ki mice showed increased WBC counts and decreased hemoglobin levels and platelet counts 5 months posttransplantation ( Figure 3C, Figure S2a

| miR-142-55A>G induces the transplantable capacity of T-cell leukemia
The miR-142-55A>G heterozygous cells developed T-cell leukemia in recipient mice. Thus, to assess the transplantable capacity of the leukemic cells, we performed secondary transplantation of the BM cells isolated from two Ki/+ leukemic mice into sublethally irradiated Ly5.1 + recipient mice ( Figure 4A). All recipient mice in two cohorts died 60 days post-transplantation ( Figure 4B). Secondarily  Figure 4C, Figure S4d-g). In addition, we observed that secondary recipients showed mild increases in thymic weight but significantly increased spleen and liver weights ( Figure 4D-F).
Thus, the miR-142 mutant was able to induce the transplantable capacity of T-cell leukemia, resulting in progressive leukemia in serial transplantation.  HSPCs showed that the majority of downregulated genes were associated with metabolic processes, cell development, and differentiation ( Figure S5d, left panel).

| miR-142-55A>G suppresses T-cell differentiation regulators and promotes leukemic proliferation
Next, to determine the characteristics of T-cells from miR-142-55A>G mice, we concentrated on the clusters of genes whose expression fluctuated in T-cells before and after leukemia relative to those in WT T-cells. Pre-leukemia had a greater number of downregulated than upregulated genes ( Figure 5C); however, post-leukemia, the manifestation of various gene clusters deviated significantly ( Figure 5D). Evidently, GO analysis revealed that terms related to the immune system and apoptosis were enriched post-leukemia, and these terms were enriched pre-leukemia ( Figure 5E).
Conversely, the upregulated genes were significantly enriched in GO terms associated with the cell cycle and metabolic processes. The gene expressions correlated with these terms were not enriched in pre-leukemia T-cells ( Figure 5F), indicating that they are upregulated during leukemogenesis ( Figure 5F).
Therefore, the miR-142-55A>G mutant induces the activation of a cluster of genes associated with immune function in HSPCs before and after leukemia, indicating that mutated HSPCs preferentially differentiated into lymphoid cells. Conversely, immune function and apoptosis were repressed in T-cells before and after leukemia, and various factors implicated in cellular growth were manifested after leukemia, indicating that miR-142-55A>G suppressed immune function and differentiation in T-cells but activated proliferation. repressing T-cell leukemia development. [35][36][37] Quantitative RT-PCR confirmed that the expression levels of Camk1d, Bcl2l11, and Parp1

| miR
were mildly decreased in Ki/+ (Leu) HSPCs relative to those of WT HSPCs ( Figure 6H). Thus, the miR-142-55A>G mutation appeared to repress its newly acquired target genes and lose the function of WT miR-142 to repress target genes, thereby driving the development of leukemia.

| DISCUSS ION
Single nucleotide mutations in miR-142 are recurrently found in patients with various hematological malignancies. WT target genes and 55A>G mutation target genes vary widely. in which the miR-142-55A>G target gene is newly repressed (proposed model in Figure 6I).
As a single miRNA is thought to be involved in the regulation of more than 100 genes depending on the sequence of its seeded re-  46,47 Additionally, the miRDB-predicted miR-142 target genes showed that 418 and 374 WT miR-142-3p target genes were found in humans and mice, respectively, and ~40%-45% of them were shared. However, the miR-142-55A>G mutation targets 294 genes in humans and 298 genes in mice, sharing only ~14% of the same genes, suggesting that the miR-142 single nucleotide mutation may result in different phenotypes in humans and mice relative to WT miR-142 (Data S5). The fact that the miR-142-55A>G target gene cluster focused on in this study was detected only in mice may have resulted in ALL rather than AML/MDS. As the miR-142-3p seed sequence contains various single nucleotide mutations other than 55A>G, further analysis is required to determine the relationship between single nucleotide mutations in the miRNA seed sequence and its target genes.
Abnormal expression of miRNAs, regulators of gene expression, has been observed in many diseases, including cancer. A miRNA gain-of-function mutation in human disease can be used to identify single miR-140 seed sequence mutations detected in bone system disease. 48 However, single nucleotide mutations in miRNAs have widely been studied as loss-of-function mutations. [49][50][51][52][53] Here, we

ACK N OWLED G M ENTS
We thank Narumi Koga, Takako Keida, Arisa Igarashi, Sho Kubota, Mariko Morii-Kubota, Mayumi Muta, Yoko Mine, and Kumiko Murakami for their technical assistance. We would also like to thank Editage (www.edita ge.com) for English language editing.

FU N D I N G I N FO R M ATI O N
This work was supported by the Japan Society for the Promotion of Science (JSPS KAKENHI, Grant number 21H02391 to K.A. and 21K05999 to M.A.).

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors have no conflict of interest.

E TH I C S S TATEM ENT
Approval of the research protocol by an Institutional Reviewer