Bone marrow mononuclear cells (BM-MNCs) were extracted from 60 newly diagnosed and untreated CML patients between 2016 and 2020 in the Department of Hematology of the Second Hospital of Hebei Medical University, Shijiazhuang, China. Full detailed information of the patient characteristics is presented in Table 1 . Furthermore, BM-MNCs of 30 healthy donors were used as normal controls (NC). Chronic myeloid leukemia was diagnosed by molecular biology, bone marrow morphology, immunology, and cytogenetics examination (19, 20). Patients with severe cardiopulmonary, renal or liver failure or coagulation abnormality or pregnant were excluded. The BM-MNCs were extracted by lymphocyte isolation fluid following the instructions. After centrifugation, the cell layer was collected for analysis. Red blood cells were lysed using RBC lysis solution, and samples were washed twice with PBS. The ethics committee of the Second Hospital of Hebei Medical University approved this experiment.

K562 and KCL22 cells were chosen as representatives of the CML cell lines. HL-60, THP-1, and U937 cells were acute leukemia cell lines. All cell types were maintained in our laboratory. All above cells were cultured in RPMI 1640-based culture medium (Gibco) or Iscoves-modified Dulbecco’s medium (IMDM; Gibco) culture medium supplemented with 10% fetal bovine serum (FBS; Gibco). The aforementioned conditioned media contained 100 units/ml penicillin and 100 ug/ml streptomycin. The cells were cultivated in an incubator at 37°C, 95% air and 5% CO₂ saturated humidity (Thermo, Waltham, MA, USA).

Lentiviruses containing shRNA-HNRNPH1 or overexpression of PTPN6 plasmid were constructed by the Shanghai Genechem Co., Ltd. (Shanghai, China). The multiplicity of infection (MOI) refers to the proportion of infectious viruses per cells. K562 cells were infected with each virus at an MOI of 30. The MOI of KCL22 was 40. The virus was added into cells from the logarithmic growth stage according to the infection conditions and cultured for 12–16 h. Then the cells were incubated with culture medium containing 10% FBS. The cells were treated with puromycin at 2 μg/ml to screen for stably transfected cells lines on conditions.

The cell viability was measured using the CCK-8 (Beibo Biological Reagent Co., Shanghai, China) assay. Approximately 100 ul of mixed suspension cells (1 × 10⁵ cells/ml) required for different experiments were added to the 96-well plate. Following cell culture, 10ul CCK-8 solution was added into each plate at various time points and incubated at 37°C in 5% CO₂ saturated humidity for 2 h. The optical density was read at 450 nm in a microplate reader (BioTek, Winooski, VT, USA) in different time points.

Cell apoptosis was assayed by an Annexin V/FITC/PI Apoptosis Detection Kit (BD Biosciences, Franklin Lakes, NJ, USA). The cells required for different experiments were mixed with 5 µl Annexin V/FITC and 10 µl propidium iodide (PI) based on the manufacturer’s instructions. They were analyzed with a FC500 flow cytometer (Beckman Coulter). The data was performed using Kaluza software (Beckman Coulter).

qRT-PCR analysis was performed with standard procedure as described previously (21). Briefly, total RNA of cells lines and BM-MNCs were extracted using Trizol (Invitrogen, Carlsad, CA, USA). The cDNA was synthesized by a SureScriptTM First-Strand cDNA Synthesis Kit (Funeng, Guangzhou, China). Quantitative real-time reverse transcriptase-polymerase chain reaction (qRT-PCR) was performed by an All-in-OneTM qPCR Mix (Funeng, Guangzhou, China). Each reaction was dependently repeated thrice. qRT-PCR was performed at 95°C for 10 min, followed by 40 cycles of 95°C (15 s), 60°C (30 s), and 72°C (30 s). Moreover, GAPDH was used as an internal reference.

Proteins were obtained using radioimmunoprecipitation assay (RIPA) buffer to dissolve the cells. The quantification was tested using bicinchoninic acid Protein Assay Kit (Boster Biological Company, Ltd., Wuhan, China). All protein samples were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The target strip was transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore, Burlington, MA, USA) via electrophoresis. The PVDF membrane was blocked in 5% nonfat milk. The protein bands were incubated with specific primary antibodies as follows: HNRNPH1 (1:2,000) (ab 154894, Abcam, CA, USA), PTPN6 (1:1,000) (ab 32559, Abcam, CA, USA), AKT (1:1,000) (ab 38449, Abcam, CA, USA), p-AKT (1:1,000) (ab 8805, Abcam, CA, USA), and β-ACTIN (1:8,000; Abways Technology, New York, NY, USA; AB0035). Moreover, the PVDF membrane was incubated in a goat-anti-rabbit secondary antibody (1:10,000, Boster Biological Company, Ltd., Wuhan, China) overnight. Images of protein quantification were captured by BioSpectrum Imaging System (UVP, LLC, Upland, CA, USA).

Cell samples were centrifuged, fixed, and smeared onto coverslips. The cells were fixed using 4% formaldehyde and preincubated with 10% normal goat serum (710,027, KPL, Gaithersburg, MD, USA). The cells were rinsed with PBS and incubated with anti-HNRNPH1 antibody (Abcam, ab 154894) at 37°C for 1 h. Immunofluorescence staining was enhanced using a rabbit anti-red fluorescent-labeled antibody (Rockland Immunochemicals Inc., Gilbertsville, PA, USA; 1:100). The smears were then incubated with DAPI. Furthermore, nuclear staining was observed by confocal microscopy (Zeiss LSM 700, Germany) and digitized with confocal software.

The colony formation was performed with standard procedure in accordance with the method described in previous study (2, 22). The logarithmic growth phase of treated cells was added to the 1% methylcellulose for 14–18 days at 37°C with 5% CO₂ saturated humidity. The methylcellulose medium was mixed with powder culture medium with ultrapure water, 40% FBS, and 1% penicillin/streptomycin. The colonies were quantified as aggregates with greater than 50 cells by microscope (Axio-observer D1, Zeiss, Germany). Then the colonies were counted manually and photographed.

The cells were treated with cell lysis buffer. The 10% lysis sample (named input) was stored, and 80% (named IP) was used in immunoprecipitation reactions with HNRNPH1 antibody (Abcam No 154894), and 10% (named IgG) was incubated with rabbit IgG (Cell Signaling Technology, Danvers, MA, USA) as a negative control The RNA of input and IP was extracted using TRIzol reagent (Invitrogen). The purified RNA samples were analyzed with conventional RT-PCR.

Three biological replicates in each control and knockdown HNRNPH1 groups in K562 cells were collected for microarray analysis after 14 days of infection. The total RNA was extracted using TRIzol reagent (Invitrogen) and subject to RNA sequencing. RNA seq was performed by BGI Technology Services Co., Ltd (Shenzhen, China). The differentially expressed genes were screened based on fold change (>0.5) and Student’s t-test (P <0.05). The RNA-sequencing raw data have been deposited into sequence read archive (SRA) database (https://www.ncbi.nlm.nih.gov/sra).

Sixteen female severe combined immunodeficient (SCID) athymic nude mice (16–20 g, 5–6 weeks’ old) were purchased from SPF Biotechnology Co., Ltd. (Beijing, China). All mice were raised at an ambient temperature of 18–22°C and relative humidity of 50–60%. The stably knockdown-HNRNPH1 K562 cells and negative control K562 cells were subcutaneously injected into the left dorsal flanks of each mouse (0.5 × 10⁶ cells per injection). Tumor size and body weight were dynamically observed. The nude mice were sacrificed on day 28 post-inoculation. All animal experiment protocols were approved by the committee on animal experimentation of Hebei Medical University and carried out following the guidelines on animal experimentation.

The tumor tissues of mice were fixed with 4% paraformaldehyde and routinely dehydrated, embedded in paraffin sections, and cut into 4-μm thick sections. The antigen repair was then performed with sodium citrate. Moreover, endogenous peroxidase was inactivated by 3% hydrogen peroxide. Tissue sections were incubated with the primary antibodies at 4°C overnight and with secondary antibody at room temperature for 30 min respectively. Sections were incubated with DAB chromogen at room temperature for 3–10 min after washing with phosphate-buffered saline with detergent Tween. Slices were sealed with coverslips after rinsing with running water and hematoxylin counterstain.

The data were presented by means ± SD. Student’s t-test and Chi-square test were applied to detect the significant difference by using the Statistical Package for the Social Sciences, version 13.0 (SPSS Inc., Chicago, IL, USA). P values <0.05 were considered statistically significant.

The difference in the HNRNPH1 expression between the BM-MNCs of CML patients and the healthy donors was first explored to investigate the role of HNRNPH1 in CML. qRT-PCR result showed that mRNA expression of HNRNPH1 in BM-MNCs of CML patients was significantly higher compared with the normal controls ( Figure 1A ). The HNRNPH1 expression in different progressions of CML was then compared. Moreover, the HNRNPH1 expression in the progressive phase was higher compared with the chronic phase ( Figure 1B ). However, the expression level in the blast phase was the highest. HNRNPH1 protein level was then also detected in the BM-MNCs of 10 patients selected in different stages of CML and healthy donors. The Western blot analysis result showed that the HNRNPH1 protein level was significantly increased in CML patients compared with normal controls. Consistent with qRT-PCR results, the HNRNPH1 protein level was increased in the blast and accelerated phase compared with the chronic phase ( Figure 1C ). Furthermore, mRNA and HNRNPH1 protein levels were found to be upregulated in leukemia cell lines compared with BM-MNCs of healthy donors ( Figures 1D, E ). Immunofluorescence also displayed that HNRNPH1 protein was primarily localized in the nucleus and increased in CML patients ( Figure 1F ). These results suggest that HNRNPH1 expression is abnormally elevated in CML, especially in the progressive phase.

HNRNPH1 was knocked by transfection of sh-RNA or empty vector (sh-Con) into CML cell lines to investigate the impact of HNRNPH1 on CML progression. The sh-RNA transfection of HNRNPH1 successfully reduced the HNRNPH1 expression level in both K562 and KCL22 cells compared with the negative controls ( Figure 2A ). Cell viabilities were then measured by CCK-8 assay. The HNRNPH1 downregulation markedly reduced cell proliferation compared with negative controls ( Figure 2B ). The cell apoptosis by flow cytometry using Annexin V-FITC/PI staining was then detected next. The result showed that HNRNPH1 knockdown in CML cells promoted cell apoptosis compared with the negative controls ( Figure 2C ). Consistently, HNRNPH1 suppression was found to inhibit the multiplication of CML cells by mainly arresting their cell cycle at G1/G0 phase(K562 44.34% vs. 52.35%, KCL22 43.61% vs. 53.60%)and caused S phase reduction(K562 48.58% vs. 38.86%, KCL22 52.41% vs. 41.9%), suggesting that HNRNPH1 knockdown played an inhibitory effect on cell arrest phase ( Figure 2D ). Furthermore, colony formation experiments also confirmed that HNRNPH1 knockdown inhibited cell proliferation ( Figure 2E ). Taken together, these results revealed that HNRNPH1 played a curtail role in CML cell survival in vitro.

CML cells were transfected with sh-RNA of HNRNPH1 or sh-Con and then treated with different concentrations of imatinib to investigate the effect of HNRNPH1 on imatinib sensitivity. The IC50 of imatinib was detected by the CCK-8 assay. Moreover, HNRNPH1 silencing significantly increased imatinib sensitivity compared with the negative control ( Figure 3A ). Consequently, the number of apoptotic cells induced by imatinib was increased in the knockdown group compared with the negative control ( Figure 3B ). Moreover, colony formation assay showed that HNRNPH1 knockdown in CML cells could promote cell sensibility to imatinib ( Figure 3C ). These results suggest that HNRNPH1 knockdown promotes imatinib sensitivity in CML cells.

HNRNPH1 was knocked down in K562 cells to identify molecular mechanisms of HNRNPH1 in regulating the growth of CML cells. RNA-seq was then performed. In addition, the differential expressed genes were displayed in the volcano plot between the sh-RNA and negative control groups ( Figure 4A ). Among those relative upregulated genes, PTPN6, which had been confirmed to play a crucial role as a cancer suppressor in leukemia, was focused on (21). The HNRNPH1 was knocked down in K562 cells to confirm the results of the RNA-seq and detect the PTPN6 expression. Consistent with the result of RNA-seq, the fold change of PTPN6 expression was upregulated in HNRNPH1 knocked down cells than control cells by qRT-PCR analysis. As shown in Figure 4B , the fold change of PTPN6 mRNA expression in K562 and KCL22 cell were 1.56 and 1.42, respectively. Likewise, the western blot result showed that the protein level of PTPN6 was increased in HNRNPH1-reduced cells compared with the negative control ( Figure 4C ). Additionally, the PTPN6 expression in CML patients was detected by qRT-PCR analysis. The PTPN6 expression which was reduced in CML-CP patients compared with healthy donors was related to illness aggravation ( Figure 4D ). A significant inverse correlation existed between the HNRNPH1 expression and PTPN6 in CML patients ( Figure 4E ). Taken together, these data revealed that PTPN6, which was downregulated in CML, may negatively correlate with HNRNPH1.

As an RNA binding protein, HNRNPH1 is involved in gene regulation by directly assembling on its mRNA (23). Radioimmunoprecipitation (RIP) followed by RT-PCR and qRT-PCR were performed to investigate whether HNRNPH1 could bind to the mRNA of PTPN6 to regulate its expression. RIP-PCR analysis showed that PTPN6 mRNA, but not Actin, was present in the protein–RNA complex pulled down by the antibody HNRNPH1 ( Figures 5A, B ), indicating that HNRNPH1 could physically bind to the mRNA of PTPN6. The screened differential genes were enriched in the PI3K/AKT pathway ( Figure 5C ), just like a previous study reported that PTPN6 could influence the activity of the PI3K/AKT pathway (24). To further verify whether HNRNPH1 contributed to the regulation of the PI3K/AKT pathway, CML cells were transfected with sh-HNRNPH1 or PTPN6 overexpression plasmid or co-transfected them together. The western blot results showed that PTPN6 overexpression could decrease the p-AKT protein level, while this reduction effect of p-AKT could be further reduced by HNRNPH1 knockdown together in CML cells ( Figure 5D ), suggesting that PTPN6 may mediate the relationship between the HNRNPH1 and the activation of PI3K/AKT in CML cells. Interestingly, this study also found that the expression trend of P210 protein that encoded BCR–ABL1 fusion gene was surprisingly consistent which is regulated positively by HNRNPH1 but negatively by PTPN6. Collectively, these results revealed that HNRNPH1 contributed to CML cell progression by moderating PI3K/AKT pathway.

The nude mice xenograft model was applied to confirm whether the reduction of HNRNPH1 expression could inhibit the proliferation of CML cells through the PTPN6–PI3K/AKT pathway in vivo. The shRNA K562 cell lines stably suppressing HNRNPH1 were screened. The HNRNPH1-reduced K562 and control cells were then subcutaneously implanted into the nude mice. The HNRNPH1 reduction could inhibit the tumor size compared with the control in vivo ( Figures 6A, B ). Similarly, this study also found that downregulating the HNRNPH1 expression could decrease the weight of xenograft tumors ( Figure 6C ). In addition, the protein level of HNRNPH1, PTPN6, AKT, p-AKT and BCR-ABL was detected by immunohistochemical ( Figure 6D ) and western blot analysis ( Figure 6E ). The results showed that the PTPN6 protein level increased while the p-AKT and BCR-ABL dramatically decreased in the HNRNPH1 reduction group compared with the negative control. The results proved that HNRNPH1 reduction also inhibited CML cell growth in vivo through the PTPN6-PI3K/AKT pathway.