BACKGROUND

Acute myeloid leukemia (acute myeloid leukemia, AML) is a type of malignant hematological tumor characterized by high incidence, high mortality, and high recurrence rates. Although chemotherapy, targeted therapy, and immunotherapy have achieved certain therapeutic effects in recent years, immune evasion by tumor cells remains a key factor in the relapse and refractoriness of AML. Current research suggests that arsenic trioxide (Arsenic Trioxide, ATO) can regulate autophagy and upregulate NKG2D ligands (NKG2D-L) on the surface of AML cells, while the NKG2D-NKG2D-L axis is negatively associated with AML immune evasion. Our preliminary experimental studies have shown that the combination of curcumin (curcumin, CUR) and ATO can inhibit the proliferation of acute myeloid leukemia KG-1a cells, induce apoptosis, and suppress the malignant progression and metastasis in AML mouse models. However, the mechanisms underlying AML immune evasion have not yet been fully elucidated. Therefore, the main objective of this study is to investigate the effects and mechanisms of CUR combined with ATO in regulating autophagy in AML cells, activating the NKG2D-NKG2D-L axis, and inhibiting AML immune evasion both in vitro and in vivo.

METHODS

First, an AML mouse model was established by intravenous injection of KG-1a cells into the tail vein of mice. The antitumor effects of CUR and ATO, either alone or in combination, on AML mice were observed. The correlation between drug regulation of autophagy, the NKG2D-NKG2D-L axis, and AML immune evasion was elucidated using ELISA and Western Blot techniques. Second, an in vitro AML cell model was established using KG-1a cells. The inhibitory and pro-apoptotic effects of CUR and ATO, either alone or in combination, on KG-1a cells were observed using the CCK8 assay and flow cytometry. The regulatory effects of the drugs on autophagy in KG-1a cells were observed using monodansylcadaverine (Monodansylcadaverine, MDC) staining and transmission electron microscopy. The specific molecular mechanisms by which CUR and ATO promote apoptosis in KG-1a cells were elucidated using Western Blot technology. Finally, an in vitro co-culture model of NK cells and KG-1a cells was established. After 24 h of intervention with CUR and ATO in combination, 3-MA to block autophagy, and 2-PCPA to induce NKG2D-L expression, the immunotoxic effects of NK cells on KG-1a cells were observed using the CCK8 assay and flow cytometry. The mechanisms by which CUR combined with ATO regulate autophagy in KG-1a cells, activate the NKG2D-NKG2D-L axis, and enhance NK cell immunosurveillance to inhibit AML immune evasion were elucidated and verified using ELISA and Western Blot techniques.

RESULTS

After intravenous injection of logarithmic phase KG-1a cells, numerous dysplastic cells were observed in bone marrow smears under the microscope. Hematoxylin and eosin (HE) staining of spleen tissue showed disordered spleen structure and tissue damage. Flow cytometry revealed a significant increase in the proportion of CD34+ cells in the peripheral blood of model mice (P < 0.001), indicating a significant increase in undifferentiated primitive cells in the peripheral blood, which confirmed the successful establishment of the AML mouse model. After CUR and ATO intervention, the model mice showed no clustering, were more active, had no arched backs, and their body weight gradually increased, indicating better survival conditions in the drug treatment groups compared to the model group. The body weight of mice in the CUR combined with ATO treatment group increased significantly (P < 0.05). The disordered spleen structure and tissue damage were significantly alleviated, and the number of dysplastic cells in the bone marrow decreased markedly, suggesting a synergistic anti-AML effect of CUR and ATO. ELISA analysis showed that compared to the control group, the model group had a significant increase in IL-10 (P < 0.0001) and a non-significant decrease in granzyme B expression (P > 0.05), indicating the presence of immune suppression and the risk of immune evasion in AML mice. Compared to the model group, drug treatment resulted in decreased IL-10 expression and increased granzyme B expression, with the most significant changes observed in the CUR combined with ATO treatment group (P < 0.001), indicating a synergistic effect of CUR and ATO in promoting immune system function in AML mice. Western Blot analysis showed that compared to the control group, the model group had increased expression of HIF-1α protein (P < 0.0001) and decreased expression of Beclin-1, LC3, NKG2D, ULBP1, and MICA proteins (P < 0.05). After CUR and ATO treatment, the expression of Beclin-1, LC3, NKG2D, ULBP1, and MICA proteins was upregulated (P < 0.05), while the expression of HIF-1α protein was downregulated (P < 0.01). CUR and ATO exhibited a synergistic effect on the upregulation of Beclin-1, LC3, and NKG2D proteins and the downregulation of HIF-1α protein (P < 0.05). CUR had the most significant effect on the upregulation of ULBP1 protein (P < 0.0001), while CUR and ATO showed a combined promoting effect on the upregulation of MICA protein (P < 0.05). In vitro experiments showed that ATO significantly promoted apoptosis in KG-1a cells (P < 0.0001), while CUR significantly promoted autophagy in KG-1a cells (P < 0.0001). The CUR combined with ATO treatment group had the lowest cell viability and the highest apoptosis rate in KG-1a cells (P < 0.001). CCK8, flow cytometry, and ELISA analyses showed that compared to the KG-1a cell group, co-culture of NK cells with KG-1a cells resulted in decreased KG-1a cell viability (P < 0.0001) and increased apoptosis rate (P > 0.05), decreased IL-10 expression, and increased granzyme B expression (P > 0.05), indicating the cytotoxic effects of NK cells on KG-1a cells. After combined treatment with CUR and ATO, the inhibition and apoptosis rates of KG-1a cells were significantly increased (P < 0.0001), and the changes in IL-10 and granzyme B expression were more significant (P < 0.05), indicating that CUR and ATO enhanced the immunotoxic effects of NK cells on KG-1a cells. Western Blot analysis showed that after CUR and ATO treatment, the anti-apoptotic protein Bcl-2 was significantly downregulated (P < 0.05), with the most significant decrease observed in the CUR combined with ATO treatment group (P < 0.05). The pro-apoptotic protein Caspase3 was upregulated, with the most significant increase observed in the ATO group (P < 0.01) and the CUR combined with ATO treatment group. In the co-culture system, after CUR and ATO treatment, the expression of Beclin1, LC3, ULBP1, and MICA proteins in KG-1a cells was significantly upregulated (P < 0.05), while the expression of HIF-1α protein was significantly downregulated (P < 0.0001). The expression of NKG2D on NK cells was also significantly upregulated (P < 0.05), with no significant difference compared to the 2-PCPA group (P > 0.05). When the autophagy inhibitor 3-MA was added to the co-culture cells treated with CUR and ATO, the expression of Beclin1, LC3, MICA, ULBP1, and NKG2D proteins all showed a downward trend.

CONCLUSION

In the AML mouse model, the combination of CUR and ATO exerted a synergistic anti-tumor effect during the progression of AML. In the in vitro KG-1a cell model, it was demonstrated that CUR combined with ATO promoted apoptosis and inhibited proliferation in KG-1a cells through upregulation of the pro-apoptotic protein Caspase3 and downregulation of the anti-apoptotic protein Bcl-2, which was accompanied by a significant increase in autophagosomes in KG-1a cells. In both the co-culture and in vivo models, CUR combined with ATO enhanced the immunotoxic effects of NK cells on KG-1a cells by improving the immunosuppressive microenvironment. The in vivo and in vitro mechanistic studies revealed that CUR combined with ATO primarily upregulated the expression of LC3 protein to promote the formation of autophagosomes in AML cells, downregulated the expression of the hypoxia-inducible factor HIF-1α protein to ameliorate the immunosuppressive microenvironment of AML, and upregulated the expression of MICA protein to activate the NKG2D-NKG2D-L axis. These actions collectively enhanced the immunotoxic and immunosurveillance capabilities of NK cells, thereby inhibiting AML immune evasion. These findings provide novel insights into the mechanisms underlying the synergistic anti-AML effects of CUR combined with ATO and their potential role in inhibiting AML immune evasion.