Upregulated SPAG6 correlates with increased STAT1 and is associated with reduced sensitivity of interferon-α response in BCR::ABL1 negative myeloproliferative neoplasms

2.1 Patient samples

Bone marrow mononuclear cells were collected from newly diagnosed patients with MPN (ET = 34; PV = 11) and controls (normal/healthy donor = 6; IDA = 20) at the Hematology Department of the First Affiliated Hospital of Chongqing Medical University between 2021 and 2022. The demographic information and clinical data of each patient and control were collected by clinical specialists and summarized in Tables 1 and S1, respectively. The BMMCs were extracted by density gradient centrifugation using Lymphoprep (BD Biosciences) and cultured in DMEM medium containing 10% FBS (PAN Biotech).

2.2 Cell lines and cell culture

Human leukemia cell lines THP-1, CCRF, SKM-1, HL-60, K562, BALL, and HEL.92.1.7 were cryopreserved by our team. Specifically, THP-1, CCRF, SKM-1, BALL, and HEL.92.1.7 cells were cultured in RPMI-1640 medium containing 10% FBS (PAN Biotech), HL-60 cells were cultured in IMDM medium containing 20% FBS (PAN Biotech), and K562 cells were cultured in IMDM medium containing 10% FBS (PAN Biotech). Human MPN-derived leukemia cell lines SET-2 and UKE-1 were donated by Professor Zhijian Xiao, Chinese Academy of Medical Sciences and Peking Union Medical College. SET-2 cells were cultured in RPMI-1640 medium containing 20% FBS (Gibco), and UKE-1 cells were cultured in IMDM supplemented with 10% FBS (Gibco), 10% horse serum (Procell), and 1 mM hydrocortisone (STEMCELL Technologies). Cells were all cultured at 37°C in a humidified incubator with 5% CO₂.

2.3 Knockdown, forced expression of target genes

The lentiviral vectors were purchased from Gene Pharma. The coding sequences of overexpressed human SPAG6 were cloned into LV5 (EF-1a/GFP and Puro) vector. The shRNA of lentivirus targeting the human SPAG6 mRNA sequence (5′-GCCATAAAGAATATCCTGCAA-3′) and its negative control (5′-TTCTCCGAACGTGTCACGT-3′) were cloned into LV3 (H1/GFP and Puro) vector. The sequences of siRNA for STAT1 were synthesized by Gene Pharma, provided in Table S2. Transfection was carried out according to the manufacturer's instructions.

2.4 Plasmid construction and dual luciferase reporter assay

The SPAG6 promoter-luciferase reporter plasmid (psiCheck2-SPAG6-promoter) containing the SPAG6 promoter region was constructed in the psiCheck2 plasmid. Gene STAT1 (NM_001384880.1) was inserted into plasmid pcDNA3.1 to construct the STAT1 overexpression plasmid (pcDNA3.1-STAT1). Luciferase reporter plasmids carrying WT SPAG6 promoter were cotransfected with either STAT1 overexpression plasmids or vector control plasmids in 293T cells. A dual luciferase reporter assay (Promega) was undertaken according to the manufacturer's instructions.

2.5 Chromatin immunoprecipitation assay

SET-2 cells were cross-linked with 1% formaldehyde for 10 min at room temperature and terminated by adding glycine (1.25 M). Fixed cells were harvested in an SDS buffer with a protease inhibitor and then sonicated to generate DNA fragments of 200–1000 bp in length. The sheared chromatin-lysed extracts were incubated overnight at 4°C with the anti-STAT1 Ab or controls (anti-RNA polymerase II Ab/rabbit IgG) with rotation. After IP of the cross-linked protein/DNA complexes, the cross-linking was reversed to release the DNA. Polymerase chain reaction was carried out with the input DNA or the immune precipitate products, and the primers used are listed in Table S1. The PCR products were then separated by agarose gel electrophoresis.

2.6 Animal model

Female BALB/ca-nu mice (4 weeks) were purchased from Beijing HFK Bioscience Co., Ltd and raised under specific pathogen-free conditions. SET-2-shSPAG6 and SET-2-shCtrl xenograft mice were established by s.c. injecting approximately 10 million cells into the flanks of 5-week-old mice. After 3 days, mice were randomized into control or drug treatment groups and treated with saline solution or 3 μg/kg peg-interferon α-2b every week for 4 weeks. Tumor size and bodyweight were measured twice a week. Tumors were removed, measured, and weighed after drug treatment for 28 days. Tumor volumes were calculated using the formula: V = 1/2(L × W²), where L represents length and W represents width. Mouse femurs, internal organs, and the xenograft tumors were collected and soaked in 4% paraformaldehyde (femurs were then slowly decalcified) and were embedded in paraffin, sectioned (thickness 3 μm), and dewaxed. Mouse tissues were stained with H&E. The histological examination was carried out on tumor sections after TUNEL, immunohistochemical, and immunofluorescent staining, as described previously.¹⁹ More information on the Abs used is provided in Table S3.

The detailed methods for RT-qPCR, western blot analysis, immunocytochemistry and double immunofluorescence, measurement of cytokines, cell viability assay, apoptosis and cell cycle assay, colony formation assay, and statistical analysis are described in Data S1.

3.1 Downregulation of SPAG6 impairs cell proliferation and cytokine release, promotes apoptosis, and reduces G₁ to S phase transition

Initially, we detected the SPAG6 mRNA expression levels in healthy donors and leukemia cell lines THP-1, CCRF, SKM-1, HL-60, K562, BALL, HEL.92.1.7, SET-2, and UKE-1, and found that SPAG6 mRNA was highly expressed in K562, HEL.92.1.7, SET-2, and UKE-1 cells (Figure 1A). HEL.92.1.7, SET-2, and UKE-1 cells are commonly used for research on the underlying mechanisms of MPN. To support our findings, a western blot experiment was undertaken to confirm the upregulation of SPAG6 protein expression in the above three cell lines (Figure 1B). Then we successfully established stable cell lines of SET-2 and UKE-1 with either SPAG6 overexpression or SPAG6 knockdown. These cell lines were derived from MPN patients with known medical histories, making them more representative of MPN biology. UKE-1 cells with the highest SPAG6 expression were only transfected with lentivirus encoding shRNA against SPAG6 and its control virus. Transfection efficiencies of the established cell lines were assessed by fluorescence microscopy and flow cytometry, revealing that the efficiencies were >90% (Figure 1C). The RT-qPCR and western blot analysis validated the successful overexpression or knockdown of the mRNA (Figure 1D) and protein (Figure 1E) expression levels of SPAG6.

Immunocytochemical and double immunofluorescence staining was used to clarify the subcellular localization patterns of the SPAG6 protein. The findings revealed that SPAG6, a microtubule-associated protein,²² showed nucleic and cytoplasmic subcellular localization patterns but was primarily located in the cytoplasm around the nucleus (Figure 2A) and was positively correlated with β-tubulin expression (Figure 2A, lower panels). To further elucidate whether SPAG6 affects the autocrine function of MPN-derived cells by regulating the microtubule system, inflammatory cytokines in culture supernatants were analyzed (Figure 2B). In SET-2 cells, overexpression of SPAG6 increased the concentration of TNF-α. In contrast, knocking down SPAG6 led to a decrease in the concentration of IL-1β. Interestingly, both overexpression and knockdown of SPAG6 resulted in increased levels of IL-8. In UKE-1 cells, the levels of TNF-α, IL-1β, and IL-2 in the UKE-shSPAG6 group were significantly lower compared to the UKE-shCtrl group. This indicates that knocking down SPAG6 in UKE-1 cells resulted in a decrease in the levels of TNF-α, IL-1β, and IL-2.

To investigate the biological effects of SPAG6 on MPN cells, CCK-8, colony formation, apoptosis, and cell cycle assays were carried out in the aforementioned cells. As shown in Figure 3A, the CCK-8 assay confirmed that the knockdown of SPAG6 in SET-2 and UKE-1 cells could significantly inhibit cell proliferation ability. In contrast, forced expression of SPAG6 had no major impact on SET-2 cell proliferation over time. The methylcellulose-based colony formation assay verified that forced expression of SPAG6 in SET-2 cells resulted in a significant increase in the number of clones, while knockdown of SPAG6 in SET-2 and UKE-1 cells showed an opposite effect (Figure 3B). Furthermore, forced expression of SPAG6 in SET-2 cells showed no significant effect on either percentage of annexin V⁺/DAPI⁻ early apoptotic cells or annexin V⁺/DAPI⁺ late apoptotic cells, but downregulation of SPAG6 in SET-2 and UKE-1 cells significantly increased the percentage of total apoptotic cells (Figure 3C). In addition, flow cytometry showed that forced expression of SPAG6 in SET-2 cells promoted G₁ to S cell cycle progression, while inhibition of SPAG6 promoted G₁/G₀ phase arrest in SET-2 and UKE-1 cells (Figure 3D).

3.2 Downregulation of SPAG6 sensitizes MPN cells for INF-α response in vitro

There is an increasing consensus suggesting that INF-α is important in MPN treatment.^10-12 Accordingly, we sought to investigate whether SPAG6 expression affected the response of MPN to INF-α. First, we determined the inhibitory effect of INF-α on MPN-derived leukemia cells. As shown in Figure 4A, downregulation of SPAG6 could increase the inhibitory effect of INF-α on SET-2 and UKE-1 cells, resulting in a significant decrease in the IC₅₀ value in shSPAG6 groups (SET-2, IC₅₀ = 59.89; UKE-1, IC₅₀ = 26.57) compared with the shCtrl groups (SET-2, IC₅₀ = 98.04; UKE-1, IC₅₀ = 61.75). Nevertheless, overexpression of SPAG6 in SET-2 cells had no significant impact on the inhibitory effect of INF-α. To standardize the experimental conditions, we used the IC₅₀ concentration of INF-α in each group for subsequent experiments. The peak inhibition effects of INF-α were reached at 48 h. Meanwhile, the peak inhibition effects were maintained for 72 h in the shSPAG6 groups, but the inhibitory effects of the SPAG6 overexpression and control groups decreased at 72 h.

Cells were then treated with the IC₅₀ concentration of INF-α in each group for 48 h; the apoptosis rate and cell cycle were detected by flow cytometry. The results showed a significant increase in apoptotic rate (Figure 4B) and percentage of cells in the G₁ phase (Figure 4C) treated with INF-α. Importantly, the INF-treated SPAG6-knockdown groups displayed the highest levels of G₁/G₀ phase arrest and apoptosis, while forced expression of SPAG6 showed the opposite effect. As various STAT proteins are involved in INF-α signaling and crucial to determine the effects of INF-α response,^{8, 9} we examined the primary STAT proteins and their phosphorylation levels (Figure 4D). Western blot analysis showed that the expressions of STAT1 and p-STAT1 were decreased following SPAG6 knockdown but yielded no significant change in the SPAG6 overexpression cells. Interestingly, after treatment with INF-α, STAT1 and p-STAT1 were induced to similar levels in both SPAG6-knockdown and its negative control groups but were lower in SPAG6 overexpression group compared with its empty vector control. As a result, STAT1 induction levels were significantly higher in shSPAG6 cells, while forced expression of SPAG6 showed the opposite effect. At the same time, lower STAT5 phosphorylation was observed in the SPAG6-knockdown groups treated with INF-α. Furthermore, with increased expression of STAT1, the expression of SPAG6 was upregulated in INF-treated cells.

In the present study, the inhibitory effects of INF-α were observed in primary MPN cells. Given that PV and ET share similar pathobiological and clinical features, while PMF presents with different clinical characteristics and poses challenges in bone marrow aspiration, only ET and PV patient samples were included in this study. The BMMCs collected from 34 patients of newly diagnosed ET and 11 patients of newly diagnosed PV were cultured and treated with INF-α of the same concentrations of 50 ng.mL⁻¹ for 48 h. At the same time, RT-qPCR was carried out using RNA extracted from whole bone marrow cells of the above 45 patients and 20 IDA cases (Figure 4E). The SPAG6 mRNA expression in IDA cases was comparable to healthy donors and used as the experimental control for this study.²¹ The results revealed that SPAG6 mRNA expression levels were significantly upregulated in patients with MPN compared to controls (p < 0.01). Based on the SPAG6 mRNA expression levels, a ROC curve was plotted, and the results yielded an area under the ROC value was 0.8922, with a sensitivity of 75.56% and a specificity of 95.00% (Figure 4F), indicating the significant value of SPAG6 in distinguishing patients with MPNs from cancer-free controls, consistent with our previous research.¹⁹ In addition, according to the median value of SPAG6 mRNA levels (9.39), patients were divided into SPAG6-low or SPAG6-high groups. As shown in Figure 4G and Table 1, the inhibition rates of INF-α on primary MPN cells were lower in the SPAG6-high group compared to the SPAG6-low group (p < 0.05). Nevertheless, no significant differences were observed in sex, age, blood corpuscle parameters, and the three driver gene mutations JAK2, CALR, and MPL (Table 1; p > 0.05).

3.3 Downregulation of SPAG6 inhibits tumor growth and sensitizes INF-α response in vivo

To explore the biological significance of downregulated SPAG6 on the antitumor activity of INF-α in vivo, xenografts mice were established. After 3 days, mice bearing established xenografts were randomized into control or treatment groups and were treated with placebo or peg-INF-α-2b as indicated (Figure 5A). Weight loss of mice began on day 14 in placebo groups, whereas sustained weight loss began on day 21 in INF groups (Figure 5B). All mice survived until the end of the treatment period. Mice were then killed, and tumorigenesis-related phenotypes were measured. The tumor size (Figure 5C), volume (Figure 5D), and weight (Figure 5E) were reduced in shSPAG6-transduced xenograft mice compared to shCtrl-transduced controls. Significantly, the shSPAG6-transduced xenograft mice treated with INF showed a marked reduction in tumor volume and weight compared to shSPAG6-transduced placebo controls. Moreover, tumor tissue sections underwent immunofluorescence staining of p-STAT1 and p-STAT5 (Figure 5F), and the results showed a significant decrease in p-STAT1 staining intensity in shSPAG6 tumors which recovered to the same level as the shCtrl group after treatment with INF. In addition, downregulation of p-STAT5 was observed upon INF-α treatment in shSPAG6 groups. Furthermore, infiltration of leukemic cells and organ toxicity of INF were assessed by H&E staining. As shown in the upper panels of Figure 6, no signs of leukemic cell dissemination, myelosuppression, or morphological and structural abnormality of internal organs were observed. Finally, immunohistochemistry analysis was used to detect the expression levels of markers related to apoptosis (TUNEL), proliferation (Ki-67 and PCNA), and cell cycle (CDK4, cyclin D1, and P21) in tumor tissues. The results suggested that INF treatment and SPAG6-knockdown significantly decreased Ki-67, PCNA, CDK4, and cyclin D1 and significantly increased TUNEL and P21 expression (Figures 5G and 6, lower panels).

3.4 STAT1 binds to SPAG6 promoter and promotes its expression

In order to ascertain the role of STAT1 signaling in regulating SPAG6 expression, we chemically inhibited STAT1 activation in HEL.92.1.7, SET-2 and UKE-1 cells by pre-incubating them with fludarabine (5 μM, 48 h). Interestingly, SPAG6 protein expression was downregulated when STAT1 expression was inhibited and not activated (Figure 7A). Next, we performed transient silencing of the STAT1 transcription factor. In line with our previous results, a decrease in SPAG6 protein expression was noted when STAT1 signaling was blocked (Figure 7B).

To gain insights into whether STAT1 could bind to the SPAG6 promoter and affect SPAG6 expression, promoter analysis in the JASPAR database revealed multiple STAT1-binding promoter regions of the SPAG6 gene (Table S4). We then undertook a Ch-IP assay in MPN-derived leukemia cells SET-2 with high expression of STAT1. Five pairs of PCR primers (Table S2) were used to amplify five separate segments of the SPAG6 promoter. As shown in Figure 7C, amplification products were observed in two short DNA fragments corresponding to 520–507 nt (SPAG-2, IP/IGG = 7.7) and 1612-1602 nt (SPAG-5, IP/IGG = 18) of the SPAG6 promoter. To further confirm that SPAG6 is a functional target of STAT1, we constructed luciferase reporter plasmids carrying WT SPAG6 promoter (Figure 7D, left panel) and the STAT1 overexpression plasmids (Figure 7D, right panel). A dual luciferase reporter assay was undertaken and revealed that overexpressing STAT1 did not affect the luciferase activity of the vector control plasmid itself, while cotransfection with STAT1 overexpression plasmid significantly increased WT SPAG6 luciferase reporter activity compared to the plasmid carrying WT SPAG6 promoter only (Figure 7F).