Identification of novel PIEZO1::CBFA2T3 and INO80C::SETBP1 fusion genes in an acute myeloid leukemia patient by RNA-seq

Fusion genes are recurrent molecular aberrations in acute myeloid leukemia, with significant diagnostic and therapeutic value. The identification of novel fusion genes provides advanced biomarkers for diagnosis and facilitates the discovery of drug targets. Bone marrow sample was extracted from an acute myeloid leukemia patient and RNA-sequencing was performed. Several bioinformatic methods, including differential analysis and Gene Set Enrichment Analysis (GSEA) pathway analyses were conducted based on the expression data. Two novel fusion genes, PIEZO1::CBFA2T3 and INO80C::SETBP1, were identified by RNA-seq. Differential analysis found that SETBP1 and CBFA2T3 were overexpressed, and GSEA analysis showed the activation of immune-related pathways. These findings indicate dysfunction of the fusion related- genes and possible pathogenic effect of the fusion genes. We reported a male AML patient with presence of PIEZO1::CBFA2T3 and INO80C::SETBP1 fusion genes.


Introduction
Fusion genes are recurrent molecular aberrations in acute myeloid leukemia (AML), with significant diagnostic and therapeutic value. A large-cohort-based research screened 36 common fusion genes, and reported that only 0.3% of leukemia patients were positive for two concurrent pathogenic fusion genes [1]. In recent years, the application of high-throughput sequencing technologies makes identifying novel fusion genes more convenient and has revealed numerous rare pathogenic fusion genes in leukemia.
The CBFA2T3 gene, located on chromosome 16q24.3, is a member of the ETO gene family and an important regulator in hematopoietic progenitor cell proliferation, erythropoiesis, and leukemia stem cell transformation [2][3][4]. CBFA2T3 encodes a protein that contains four evolutionary conserved nervy homology regions (NHR 1-4), among which the NHR2 domain can interact with hematopoiesisrelated transcription factor RUNX1, promoting acute lymphoblastic leukemia proliferation [5].
The SETBP1 gene is mapped to chromosome 18q12.3, and encodes a protein that contains a SKI-homology domain, a SET-binding domain (SETBD), and 3 A-T hooks. SETBP1 protein binds to oncoprotein SET and forms a heterodimer, inhibiting PTPA and enhancing leukemia cell proliferation. SETBP1 also modulates target genes at the transcriptional level by interacting with DNA via its A-T hooks [6][7][8]. Somatic gain of function mutations of the SETBP1 homology domain has been extensively reported as a common driver mutation in myeloid malignancies [2].
Here we report an AML case identified with PIEZO1::CBFA2T3 and INO80C::SETBP1 fusion genes. To the best of our knowledge, it is the first time these two novel driver fusion genes were reported in AML patients.

Case report
The patient, male, 64 years old, was transferred to our department on September 9th, 2019, with fatigue and consistent fever. Peripheral blood examination showed hemoglobin level of 67 g/L, red blood cell (RBC) count of 2.2 × 10 12 /L, white blood cell (WBC) count of 8.4 × 10 9 /L, and platelet count of 7 × 10 9 /L. Morphological examination of peripheral blood reported a blast cell ratio of 45% and the presence of abnormal erythrocytes. Ultrasound found an enlarged spleen. To further clarify the diagnosis, bone marrow (BM) aspiration was performed and revealed 81.5% blast cells. The blast cells were positive for CD117, CD34, HLA-DR, CD36, and CD17; negative for CD19, CD56, CD22, and MPO. Metaphase analysis was unsuccessful. DNA sequencing of 58 frequently mutated genes in hematologic malignancies [3] identified KRAS c.35G > A/p. G12D and PTPN11 c.1508G > T/p.G503V mutations in this patient. He was diagnosed with AML FAB-M2 based on the patient's symptoms, clinical findings, and immunophenotyping results. Hemostasis and anti-inflammatory treatments were administered since the day of admission. However, the patient refused chemotherapy for financial reasons and was discharged from the hospital on September 19th ,2019, and passed away ten days later.

RNA sequencing and fusion validation
RNA-Seq was performed using RNA extracted from the BM sample by HiSeq 2,500 (Illumina, San Diego, CA, USA) to search for the potential fusion gene. Reads were mapped and processed by combining Arriba (v1.0.1) [4] and STAR-Fusion (v1.3.1) [9] to analyze the gene fusions. For validation, we obtained the cDNA sample of the diagnostic bone marrow and performed RT-PCR using specific primers followed by Sanger sequencing (The primer sequences were listed in the supplementary materials).

Differential expressed genes
We utilized the DESeq2 R package to compare the expression data (HTSeq-Counts) of the patient with the control group of 50 healthy people. The threshold value was log 2-fold change (FC) > 1 and adjusted p value < 0.05. Normal control samples used for gene expression analysis were bone marrow samples from healthy donors. All of the samples used in this study were approved by the ethics committee at the 2nd Affiliated Hospital of Harbin Medical University.

Gene set enrichment analysis
Gene Set Enrichment Analysis (GSEA) of differentially expressed genes in RNA-Seq data were implemented by the ClusterProfiler R package [10].

Novel fusion gene PIEZO1::CBFA2T3 is associated with dysregulation of CBFA2T3
Whole transcriptome RNA sequencing (RNA-Seq) performed on the patient's bone marrow sample detected a highly likely pathogenic fusion gene, PIEZO1 exon1::CBFA2T3 exon2, which was confirmed by RT-PCR followed by Sanger sequencing (Fig. 1A). The novel PIEZO1::CBFA2T3 fusion in this elderly male patient preserved the coding region of all functional domains of CBFA2T3, possibly acts as a driver in oncogenesis (Fig. 1A). To elucidate the pathogenic significance of PIEZO1::CBFA2T3, we screened four CBFA2T3-coexpressing genes in AML based on a previous study (upregulating genes include CDKN1A, JUP, KAT2A, and the down-regulating gene TYROBP [11]) and compared the expression levels with a control set of 50 normal bone marrow (BM) samples ( Fig. 2A). Overall, these data support the PIEZO1::CBFA2T3 fusion resulted in overexpression and dysregulation of CBFA2T3 and played an essential role in leukemogenesis.

Identification of INO80C::SETBP1 fusion gene and overexpression of SETBP1
RNA-Seq identified an in-frame INO80C exon1 -SETBP1 exon3 fusion in this patient. Notably, the INO80C::SETBP1 fusion protein comprises the intact functional domains of SETBP1, suggesting the dysregulated SETBP1 function possibly contributes to oncogenesis (Fig. 1B). To confirm our speculation, we examined the patient's expression levels of a selected list of genes (SETBP1 along with SETBP1-coexpress genes HOXA9, HOXA10, function-related genes SET, PTPA, and hematopoiesis-associated gene MECOM as well as oncogene MYB, which can be upregulated by SETBP1 and mutated SETBP1, respectively [6,12,13] ) with the expression levels of 50 healthy donors as a control set. Our findings indicated the aberrant upregulation of SETBP1( Fig. 2A) . Other significantly dysregulated pathways include TGFβ, IFNγ, TNFα, and IL2_STAT5, suggesting a broad dysregulation of the immune microenvironment (Fig. 2B).

Discussion
Despite CBFA2T3 fusions being rare, CBFA2T3::GLIS2 has been reported in pediatric AML. RUNX1::CBFA2T3, NFIA::CBFA2T3, and CTCF::CBFA2T3 have been reported in pediatric and adult AML cases. The splicing site of CBFA2T3 in CBFA2T3::GLIS2 is located at exon 10 or exon 11, and the splicing sites in other CBFA2T3 fusions are commonly located at exon 3, exon 4, or exon 5 [17][18][19]. Notably, the main functional domains are generally retained in previously reported CBFA2T3 fusion genes, regardless of whether it acts as a 5' or 3' partner gene, which is in accordance with our findings and strongly suggests the promiscuous oncogenesis function of CBFA2T3 in these fusion scenarios. Fusion genes often serve as prognostic determinant biomarkers and guide risk-adapted treatment in AML. CBFA2T3 protein has almost the same functional structure as RUNX1T1, a frequent fusion partner of RUNX1. RUNX1::CBFA2T3 fusion mimics the characteristics of RUNX1::RUNX1T1, leading to a favorable prognosis [20]. However, CBFA2T3::GLIS2, a recurrent fusion in pediatric AML, correlates with a poor outcome [21]. Steinauer et al. found that the expression level of CBFA2T3 can predict patient-specific outcomes in AML, which is congruent with our findings [11]. Somatic SETBP1 mutations are related to inferior outcomes in myeloid malignancies [22][23][24]. Moreover, INO80C, the SETBP1 fusion partner, also potentially impacts prognosis. INO80C, which regulates chromatin remodeling by interacting with transcription factor YY1, is part of the YY1-EGR1-INO80C network that can act as a prognostic biomarker of AML [25].

Conclusion
In summary, we report a case of AML bearing two previously undescribed fusion genes, PIEZO1::CBFA2T3 and INO80C::SETBP1. RNA-Seq data and in silico analysis Data availability Data included in the present article will not be made publicly available due to local control policies for human genetic resource data. Nonetheless, data access requests from academic investigators can be directed to the corresponding author.

Conflict of interest
The authors declare no conflict of interest.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Consent to participate Informed consent was obtained from the patient's family member and the control subjects in this study.

Consent to publish
Informed consent was obtained from the patient's family member and the control subjects for publication of this study.