CHIP‐associated mutant ASXL1 in blood cells promotes solid tumor progression

Abstract Clonal hematopoiesis of indeterminate potential (CHIP) is an age‐associated phenomenon characterized by clonal expansion of blood cells harboring somatic mutations in hematopoietic genes, including DNMT3A, TET2, and ASXL1. Clinical evidence suggests that CHIP is highly prevalent and associated with poor prognosis in solid‐tumor patients. However, whether blood cells with CHIP mutations play a causal role in promoting the development of solid tumors remained unclear. Using conditional knock‐in mice that express CHIP‐associated mutant Asxl1 (Asxl1‐MT), we showed that expression of Asxl1‐MT in T cells, but not in myeloid cells, promoted solid‐tumor progression in syngeneic transplantation models. We also demonstrated that Asxl1‐MT–expressing blood cells accelerated the development of spontaneous mammary tumors induced by MMTV‐PyMT. Intratumor analysis of the mammary tumors revealed the reduced T‐cell infiltration at tumor sites and programmed death receptor‐1 (PD‐1) upregulation in CD8+ T cells in MMTV‐PyMT/Asxl1‐MT mice. In addition, we found that Asxl1‐MT induced T‐cell dysregulation, including aberrant intrathymic T‐cell development, decreased CD4/CD8 ratio, and naïve‐memory imbalance in peripheral T cells. These results indicate that Asxl1‐MT perturbs T‐cell development and function, which contributes to creating a protumor microenvironment for solid tumors. Thus, our findings raise the possibility that ASXL1‐mutated blood cells exacerbate solid‐tumor progression in ASXL1‐CHIP carriers.


| INTRODUC TI ON
During human aging, somatic mutations are accumulated in many types of cells. Recent whole-genome sequencing studies revealed that clonal expansion of blood cells with acquired somatic mutations is unexpectedly common in healthy aged individuals. 1,2 This phenomenon is called clonal hematopoiesis of indeterminate potential (CHIP). Clonal hematopoiesis of indeterminate potential carriers are at increased risk for all-cause mortality, blood cancers, and cardiovascular diseases. [3][4][5][6] Clonal hematopoiesis of indeterminate potential-associated mutations frequently occurred in genes encoding epigenetic regulators, including DNMT3A, TET2, and ASXL1.
Additional sex combs-like 1 (ASXL1) is a member of the mammalian ASXL family. ASXL1 regulates gene expression and signal transduction through interactions with multiple proteins, such as BAP1, 7 EZH2, 8 BMI1, 9 BRD4, 10 AKT, 11,12 and NONO. 13 In addition to CHIP, ASXL1 is frequently mutated in myeloid malignancies and associated with poor prognosis. ASXL1 mutations are detected in the last exon, resulting in the translation of C-terminally truncated ASXL1 proteins.
Clinical evidence suggests that CHIP is particularly prevalent in solid-tumor patients, and its presence has an adverse impact on their overall survival. 18 The high frequency of CHIP in solid-tumor patients is correlated with primal exposure to anticancer therapies and smoking habits. 19 Clonal hematopoiesis of indeterminate potential mutations in DNA damage repair genes, such as TP53, PPM1D, CHEK2, are frequently found in patients with prior exposure to cytotoxic chemotherapy. Smoking is strongly associated with ASXL1 mutations. 19 Whether the blood cells with CHIP-associated mutations have causal effects in solid-tumor progression, as they do in cardiovascular diseases, 20,21 has been unclear. A previous study using Tet2-deficient mice showed that myeloid cell-specific Tet2 deficiency inhibits melanoma progression, 22 while other studies showed that Tet2 deficiency in immune cells promotes the growth of hepatoma and lung cancer cells. 23,24 Thus, it appears that Tet2-deficient immune cells create protumor or antitumor microenvironment in a tumor type-dependent manner. The role of other CHIP-associated mutations in the development of solid tumors has not been investigated experimentally.
In this study, we assessed the role of blood cells with the ASXL1 mutation in various mouse solid-tumor models using the Asxl1-MT fl/fl mice crossed with Vav-Cre, LysM-Cre, and Lck-Cre mice. Our data indicate that Asxl1-MT perturbs T-cell development and function, which contributes to creating a protumor microenvironment for solid tumors.

| Mice
Asxl1-MT fl/fl mice were generated as described previously 17 and were crossed with Vav-Cre mice, 25 LysM-Cre mice, 26 and Lck-Cre mice. 27 LysM-Cre mice were purchased from The Jackson Laboratory (stock no. 004781). Lck-Cre mice were provided by the Laboratory Animal Resource Bank, National Institutes of Biomedical Innovation, Health, and Nutrition. MMTV-PyMT mice (FVB/N background) were also purchased from The Jackson Laboratory (stock no. 002374) and backcrossed into the C57BL/6J (CLEA Japan) background for at least 12 generations. 28

| Genotyping
Vav-Cre; Asxl1-MT fl/fl mice carry the 3xFLAG-Asxl1-MT-flox/flox-IRES-EGFP cDNA. Therefore, we used GFP expression as a marker of Asxl1-MT expression. We collected peripheral blood from the mice's tails and assessed GFP expression in blood cells by flow cytometry. Asxl1-MT heterozygous (fl/wt) and Asxl1-MT homozygous (fl/fl) mice were genotyped by genomic PCR. LysM-Cre, Lck-Cre, and MMTV-PyMT mice were also genotyped by genomic PCR. DNA was extracted from the mice's tails by the NaOH method. Primers used for the genotyping are provided in Table S1.

| Spontaneous mammary tumor model
Vav-Cre; MMTV-PyMT-Asxl1-MT fl/fl and MMTV-PyMT-Asxl1-MT fl/fl mice were palpated weekly from the age of 12 weeks, and the time of the tumor onset was recorded. Note that C57BL/6 MMTV-PyMT mice develop tumors slower than the commonly used FVB background. 29 At the endpoint, mammary tumors/glands were collected.
The numbers of developed tumors (visible and palpable) in the cervical, thoracic, and abdominal-inguinal mammary gland were counted, extirpated, and weighed to measure the whole-body tumor weights.
The samples were then fixed and preserved in formalin for H&E staining. Blood was collected at the endpoint day and was used to measure complete blood count. The tumor-infiltrated immune cells were examined by flow cytometry.

| Flow cytometry
For T-cell analysis, the spleen and thymus were isolated from the mice and were crushed above the 100μm cell strainer (Greiner Bio-One, # 54200) with the head of the sterilized 2.5-ml syringe plunger.
For the analysis of tumor-infiltrated blood cells, tumors were dissociated (see Section 2.6), collected, and put into FACS buffer. Red blood cells were lysed with 1×RBC Lysis Buffer. A total of 1 × 10 6 cells per sample were stained with antibodies listed in Table S2.

| Histological tumor examination
The murine mammary gland containing adenocarcinomas or lumps from control (MMTV-PyMT-Asxl1-MT fl/fl ) and Asxl1-MT (Vav-Cre; MMTV-PyMT-Asxl1-MT fl/fl ) mice were fixed and stocked in 10% formalin-PBS solution. The tissues were embedded in paraffin and processed for sectioning and H&E staining by the Pathology Core Laboratory of the IMSUT. Multiple unpaired t test with Welch correction was used for the tumor volumes. Unpaired t test was used for the comparison of the endpoint weights. Two or three independent experiments using littermate mice were performed in a blind manner CD4 + or CD8 + T cells were sorted by AriaIII. Total RNA was extracted using RNeasy Mini Kit (Qiagen). mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. Pair-end sequencing FASTQ files were aligned to the mouse reference genome (mm10) using HISAT2 30

| Statistical analysis
GraphPad Prism 9 was used to perform statistical analyses. Twotailed unpaired t test or multiple t test was used for pairwise comparisons, and one-way ANOVA was used for multiple comparisons of significance. The log-rank test (Mantel-Cox test) was used for the tumor-free survival curve. FlowJo was used for FCM data analysis.

| T cells expressing Asxl1-MT promote solidtumor progression in syngeneic transplantation models
We first investigated the influence of blood cells with the ASXL1 mutation on the growth of solid tumors using the syngeneic mouse tumor models. We crossed Asxl1-MT fl/fl mice with Vav-Cre, LysM-Cre, or Lck-Cre mice to generate the mice expressing Asxl1-MT in all blood cells, myeloid cells, or T cells, respectively. Homozygous (fl/ fl) Asxl1-MT mice were identified using PCR and used for experiments ( Figure S1A). Expression of GFP (Asxl1-MT) in the myeloid or lymphoid fraction in LysM-Cre or Lck-Cre mice, respectively, was confirmed by FACS ( Figure S1B-E).
Next, we subcutaneously injected the C57BL/6 mouse-derived solid-tumor cell lines (B16F10 melanoma cells, LLC cells, and MC38 colon cancer cells) into these mice. The tumor size was measured every other day, and the tumor weight was measured at the endpoint day ( Figure 1A). We did not observe substantial changes in the growth of all these tumor cells between control and Vav-Cre; Asxl1-

| Blood cells expressing Asxl1-MT promote the development of spontaneous mammary tumors
In the syngeneic tumor models, tumors are generated by subcuta- Importantly, we observed a significant reduction of infiltrated CD3 + and CD4 + T cells in mammary tumors of Asxl1-MT mice ( Figure 3E).
Although the number of CD8 + T cells was not significantly reduced in tumors of Asxl1-MT mice ( Figure 3E), the expression of an exhaustion marker PD-1 was upregulated in them ( Figure 3F,G). These data indicate that Asxl1-MT-expressing T cells have an exhausted phenotype and impaired ability to infiltrate the tumor site.  Figure 4C,D). However, the proportion of DN1 cells, especially lineage − c-Kit + early T-cell precursor (ETP) population, was significantly increased in Asxl1-MT-expressing thymi ( Figure 4E). In contrast, CD44 − CD25 + DN3 cells were decreased in Asxl1-MT-expressing thymi, indicating a developmental defect from DN1 to later stages ( Figure 4F). We also found that the CD8 + single-positive (SP) cells were increased, while the CD4 + CD8 + double-positive (DP) cells and CD4 + SP cells remained unchanged ( Figure 4G). Thus, the expression of Asxl1-MT alters the intrathymic differentiation of immature T cells.
Next, we assessed T-cell phenotypes in the peripheral organs.  Figure 5A,B). Intriguingly, we also found a dramatic reduction of naïve CD4 + and CD8 + T cells and an increase of memory and effector CD8 + T cells in the spleen of Vav-Cre; Asxl1-MT fl/fl mice ( Figure 5C,D). The reduction of peripheral naïve T cells and a relative increase of memory/effector T cells are typical immunosenescent features that are observed in aged T cells. [36][37][38] The T-cell phenotypes were also confirmed in Asxl1-MT mice inoculated with the Py8119 breast cancer cells ( Figure S3A). Interestingly, we observed the tendency of PD-1 upregulation in peripheral CD8 + T cells of Asxl1-MT mice with Py8119 cells (Figure S3B), indicating that tumor cells may promote the exhaustion of Asxl1-MT T cells. However, this possibility should be examined in spontaneous tumor modes in future.
Taken together, these data suggest that mutant Asxl1 perturbs T-cell development in the thymus and induces naïve-memory imbalance in peripheral organs ( Figure 5E).

| Asxl1-MT induces inflammation and mitochondrial dysregulation in T cells
To examine the molecular changes in T cells expressing Asxl1-MT, we performed RNA-seq using splenocytes from control (Asxl1- We previously showed that Asxl1-MT activates mitochondrial dysregulation as well as overproduction of ROS in HSCs. 11 Consistent with the phenotypes of Asxl1-MT-expressing HSCs, we observed increased mitochondrial membrane potential in CD4 + and CD8 + Asxl1-MT-expressing T cells ( Figure 6E). Asxl1-MT also increased the intracellular ROS level in CD4 + T cells and tended to increase it in CD8 + T cells ( Figure 6F). Collectively, these data suggest that Asxl1-MT provokes inflammation and mitochondrial dysregulation in T cells, thereby accelerating T-cell aging.

| DISCUSS ION
In this study, we demonstrated that T-cell-specific expression of shown. Data are shown as mean ± SEM. Unpaired t test was used for the comparison of T cells, as it does in HSCs. 11 It is also tempting to speculate that the altered T-cell function may contribute to the evolution of ASXL1-CHIP and the development of myeloid neoplasms driven by ASXL1 mutations. How ASXL1 mutations affect T-cell development is still obscure and warrants further investigation.
In summary, we showed that CHIP-associated Asxl1-MT induces T-cell dysregulation and promotes tumor progression in multiple solid-tumor models. Our findings raise the possibility that blood cells with ASXL1 mutations exacerbate solid-tumor progression in ASXL1-CHIP carriers.

ACK N OWLED G EM ENTS
We sincerely thank Shiori Shikata for her expert technical assistance.
We sincerely thank the Flow Cytometry Core and the Mouse Core at IMSUT for their technical help. We also thank Dr. Takuma

D I SCLOS U R E
The authors declare no conflict of interest.