Ginsenoside Rk3 Inhibits the Extramedullary Infiltration of Acute Monocytic Leukemia Cell via miR-3677-5p/CXCL12 Axis

Background Acute monocytic leukemia belongs to type M5 of acute myeloid leukemia (AML) classified by FAB, which appears a high incidence of extramedullary infiltration (EMI) and poor prognosis. In this study, we observed the inhibitory effect of ginsenoside Rk3 on the EMI of monocytic leukemia cells and initially explored its related mechanism of targeting the miR-3677-5p/CXCL12 axis. Methods The MTT assay and colony formation assay were used to detect the inhibitory effect of Rk3 on proliferation. Both cellular migration and invasion were observed by the Transwell assay. The expression levels of miR-3677-5p, CXCL12, and CXCR4 were detected by RT-qPCR and Western blot, as well as overexpression of miR-3677-5p by transfected with lentivirus and detection of a dual luciferase reporter gene. The expression of MMP2 and TIMP2 was detected by immunofluorescence. Results Rk3 effectively inhibits the proliferation, migration, and invasion associated with EMI of leukemia. The leukemia cells of M5 patients with EMI showed low expression of miR-3677-5p but high expression of the mRNA of CXCL12 and CXCR4. Overexpression of miR-3677-5p or intervention of CXCL12 effectively inhibited the proliferation, migration, and invasion of SHI-1 cells. The luciferase assay showed that CXCL12 was the downstream target gene of miR-3677-5p. After overexpression of miR-3677-5p or intervention of CXCL12 in combination with Rk3, the inhibitory effect on the proliferation, migration, and invasion of SHI-1 cells was more obvious. Importantly, Rk3 significantly regulated the expression levels of miR-3677-5p, CXCL12, CXCR4, and EMI-related functional proteins including MMP2 and TIMP2. Overexpression of miR-3677-5p or intervention of CXCL12 also regulated the expression of MMP2 and TIMP2. Conclusions The leukemia cells of M5 patients with EMI appeared to have low expression of miR-3677-5p and high expression of the mRNA of CXCL12 and CXCR4, which may be used as indicators of EMI and poor prognosis. Rk3 is effective in inhibiting the EMI of SHI-1 cells by targeting the miR-3677-5p/CXCL12 axis.


Introduction
Acute myeloid leukemia (AML) is a clonal, malignant disease of the hematopoietic tissues that is characterized by the accumulation of abnormal leukemic blast cells, principally in the marrow, and impaired production of normal blood cells. Tus, the leukemic cell infltration in the marrow is accompanied, nearly invariably, by anemia and thrombocytopenia. Te absolute neutrophil count may be low or normal, depending on the total white cell count [1]. Acute monocytic leukemia belongs to type M5 of AML, as classifed by FAB. Te incidence of extramedullary infltration (EMI) in clinical patients with M5 leukemia is as high as 30% to 40%. Infltration of the central nervous system is a serious complication, leading to recurrence and death, with a poor prognosis and no specifc efective medicine. Te treatment of patients is still dominated by chemotherapy, and the main problems of chemotherapy are treatment-related toxicity, mortality, and the occurrence of multidrug resistance (MDR) [2][3][4]. Existing targeted therapies and immunotherapy is also not ideal for prevention and radical cure of EMI. Furthermore, they are expensive.
Terefore, it is of important clinical value to research and develop new Chinese medicine or natural compounds that are safe and efective for anti-EMI so as to improve the prognosis of M5 leukemia patients.
Ginseng is a Chinese medicine herb, and its main ingredient is ginsenosides, which are characteristics of anti-infammatory properties, antitumor and other pharmacological efects [5][6][7][8][9]. Our previous work has identifed a biological active compound extracted and isolated from ginseng herb by using biological activities assay as indicators, we termed the purifed polarity ginsenosides as ALK compound, the invention patent was authorized by Patent Ofce of China, which possesses clear efects on inhibiting the proliferation and inducing diferentiation in AML cells, as well as anti-EMI of central nervous system and other organs, which caused by monocytic leukemia SHI-1 cells in mice models with EMI. In order to explore the material basis of the ALK compound for antileukemia, the purifed polarity ginsenoside monomers of F4, Rh4, and Rk3 were isolated from ALK, and the results showed that F4 signifcantly inhibited the proliferation, and also induced the diferentiation of K562 and SHI-1 leukemia cells [10]. Ginsenoside Rh4 was efective to inhibit the proliferation, arrest cell cycle, and induce diferentiation of K562 cells [11]. Since Rk3, nothing has been reported about its action and mechanism in antileukemia.
MicroRNA (miRNA) is a single-stranded RNA molecule of about 22 nucleotides, located at the core of the ncRNA regulatory network. Its function is to participate in the maintenance of cellular proliferation, diferentiation, and apoptosis. Abnormal expression or regulation of miRNAs is closely related to the occurrence and progression of leukemia/tumor [12][13][14][15][16]. Studies have shown that miR-3677 is low expressed in serum samples of osteosarcoma patients and is associated with a poor prognosis of osteosarcoma [17]. Overexpression of miR-3677 can inhibit the proliferation and migration of osteosarcoma cells [18]. But the expression level of miR-3677 and its specifc mechanism of actin have not yet been elucidated in acute monocytic leukemia. Te miR-3677-5p is processed from the 5′ end arm of the miR-3677 precursor. CXCL12 is a ligand of the G-protein-coupled receptor CXCR4, and CXCR4 and its ligand-induced signal cascade are usually defned as the CXCL12/CXCR4 axis [19]. Te CXCL12/ CXCR4 axis is involved in the survival and homing of hematopoietic stem cells in bone marrow and is closely related to the proliferation and invasion of malignant cells [20][21][22].
Terefore, we studied the efects of miR-3677-5p and CXCL12 expression on the proliferation, migration, and invasion of SHI-1 cells and explored the active mechanism of Rk3 for anti-EMI in monocytic leukemia cells.

Cell Culture, Lentivirus Transfection, and Treatment.
Human monocytic leukemia SHI-1 cells were donated by Professor Chen Zixing, Institute of Hematology, and the First Afliated Hospital of SooChow University. Te cells were maintained in Iscove's Modifed Dulbecco's Medium (IMDM, Gibco, USA) supplemented with 15% new bovine serum (NBS, Every Green, China) in a humidifed 5% CO 2 atmosphere at 37°C. Te structure of ginsenoside Rk3 (WINHERB, Shanghai, China) was identifed by nuclear magnetic resonance, mass, infrared spectrum, and ultraviolet absorption spectra. Te purity was >98%, and the batch No. 140606. Polybrene transfection reagent was used for the transfection of miR-NC (GenePharma, Shanghai, China) and miR-3677-5p mimic (GenePharma, Shanghai, China) according to the manufacturer's instructions. Te AMD3100 (the inhibitor of the CXCL12/CXCR4 axis) was purchased from Selleck Chemicals (Shanghai, China), and the batch No. S803002.

Collection of Bone Marrow Specimens and Clinical Data.
Te bone marrow samples from 20 patients diagnosed with AML-M5 were collected in our hospital from April 2021 to March 2022; the diagnostic criteria were based on the 2016 World Health Organization Classifcation of AML [23]. Te bone marrow samples were aspirated before chemotherapy, and anticoagulated by heparin, afterwards, leukemia cells were isolated from these bone marrow samples. Te patient clinical manifestation and data were recorded, including extramedullary infltration (EMI) or nonextramedullary infltration (N-EMI), the percentage of bone marrow leukemia blasts, chemotherapy regimens, and response to chemotherapy (Table 1). 20 cases of AML patients were divided into the EMI group (10 cases) and the N-EMI group (10 cases) according to their clinical manifestations. Patients in the EMI group showed infltration of brain tissue (according to the cerebrospinal fuid examination), enlargement of the liver, spleen, lymph nodes, and tenderness of the sternum, which suggested that leukemia cells were infltrated in these organs. Te normal control group consisted of 10 normal bone marrow samples, which required bone marrow examination for clinical diagnosis, but the bone marrow smears showed normal hematopoietic morphology. Tis study was approved by the Ethics Committee of the First Afliated Hospital of Zhejiang Chinese Medicine University.

MTT Assay.
About 2 × 10 4 SHI-1 cells were plated into each well of a 96-well plate, and treated with Rk3 at the fnal concentration of 60, 120, and 180 mol/L for 48 hours at 37°C in a 5% CO 2 incubator, and each group of Rk3 concentration was repeated planting for 5 wells. Ten, 20 μl MTT solutions (5 mg/mL) was added and incubated for 4 hours. After the supernatant was removed by centrifugation, 150 μl DMSO was added, the absorbance of the solution was read by a microplate reader ELx808 (BioTek, USA), using a test wavelength of 490 nm. Te inhibition rate of Rk3 on the proliferation of SHI-1 cells was calculated, the inhibition rate (%) � [(control group A value − experimental group A value)/control group A value] × 100%. Te same experiment was repeated for 5 times.   In addition, the other three groups included miR-3677-5p mimic, miR-NC, and 2.5 μM of AMD3100, which were collected for the Transwell assay. In the migration experiment, 1 × 10 5 cells (200 μl) were seeded into the upper chamber of a Transwell insert (8 mm pore size, Corning Inc., USA), and a medium with 15% NBS (600 μl) was added in the lower chamber. In the invasion experiment, Matrigel (BD Bioscience, Shanghai, China) was coated on the upper chamber seeded with 1 × 10 5 cells, and the lower chamber contained 15% NBS medium. After incubation at 37°C, 5% CO 2 for 24 hours, the Transwell chamber was taken out and washed with PBS. Te cells on the membrane were fxed with 4% paraformaldehyde for 20 mins, and stained with crystal violet for 30 mins. Te upper unmigrated cells were gently wiped of with a cotton swab, and fve felds were randomly selected under the microscope (100 × magnifcation, Olympus, Japan) to take photos and count cell numbers.

Western
Blot. Cells were washed twice with ice PBS, and total cell lysate samples were extracted and subjected to SDS-PAGE gel, then, transferred to a nitrocellulose membrane (Amersham, USA), and blocked with Protein-Free Rapid Blocking Bufer (Epizyme, China). Te membranes were then reacted with primary antibodies including CXCL12 (Abcam9797, 1 : 1000), CXCR4 (SC53534, 1 : 1000), and β-actin (CST13E5, 1 : 1000) overnight at 4°C. After being washed with 1 × TBST bufer for three times, the membranes were incubated with secondary antibody (BIO-RAD, 1 : 3000, USA) for 1 hour at room temperature. Te special bands from the conjugation reaction of protein antigen and antibodies were visualized by using Luminol from the enhanced chemiluminescence (ECL) kit (Santa Cruz, USA). Specifc protein bands were measured with the densitometric intensity (SHST, China). Te experiment above was repeated 3 times.
2.10. Immunofuorescence. SHI-1 cells were collected after being treated with 60, 120, and 180 mol/L of Rk3 for 48 hours,and the other groups included miR-3677-5p mimic, miR-NC, and 2.5 μM of AMD3100. Ten, prepared smear slides by cyto-centrifugation at 800 rpm for 5 mins. Te slides were fxed for 10 mins in 4% paraformaldehyde and treated with 0.5% TritonX-100 for 20 mins. Nonspecifc binding was blocked by incubating with 1% BSA. Te primary antibodies, including MMP2 (CST40994, 1 : 200) and TIMP2 (Abcam180630, 1 : 200) were incubated over night at 4°C, and secondary antibodies were incubated for 1 hour in the dark. Te nuclei were counterstained with DAPI. Te cellular fuorescence intensity on slides was observed under the fuorescence microscope of ZEISS. Te fuorescence intensity was calculated using Image J software.
2.11. Statistical Methods. Te data were analyzed using GraphPad Prism software (La Jolla, CA, USA) and shown as mean ± Standard error of the mean. Te data comparison was conducted by using the Student's t-test or One-way ANOVA, and P < 0.05 was considered as statistically signifcant.

Rk3 Inhibited the Migration and Invasion of SHI-1 Cells.
Te inhibition efect of Rk3 on migration and invasion were observed by the Transwell assay. Te number of cell migrations in Rk3 groups of 60, 120, and 180 mol/L were 83.6 ± 6.4, 55.8 ± 6.3, and 28.4 ± 5.7, respectively, which were signifcantly lower than the 110 ± 6.0 in the control group (P < 0.01). Te inhibition rates of Rk3 on the migration of SHI-1 cells were from 23.7% to 74.4%. Te similar results were also observed in the invasion assay, the number of cell invasions in the 60, 120, and 180 mol/L groups were 98.7 ± 7.4, 69.5 ± 4.3, and 42.2 ± 3.7, respectively, which was lower than the 115.5 ± 6.2 in the control group (P < 0.01).
Te inhibition rates of Rk3 on the invasion of SHI-1 cells were from 14.5% to 63.4%, as shown in Figure 2.

Te Correlation between the Expression of miR-3677-5p, CXCL12
, and CXCR4 and EMI in M5 Patients. Te results of GEPIA website showed that the expression levels of CXCL12 and CXCR4 in AML patients were signifcantly higher than those in healthy people (P < 0.05), as shown in Figure 3(a). Te relative expression level of miR-3677-5p in AML-M5 patients was signifcantly lower than that in normal control group (P < 0.01). Meanwhile, the relative expression level of miR-3677-5p in EMI group was 0.36 ± 0.10, which was lower than the 0.73 ± 0.11 in N-EMI group (P < 0.01). Te mRNA relative expression levels of CXCL12 and CXCR4 in AML-M5 patients was signifcantly higher than those in the normal control group (P < 0.01). Meanwhile, the mRNA relative expression levels of CXCL12 and CXCR4 in the EMI group were 2.92 ± 0.37 and 2.63 ± 0.66, respectively, which were signifcantly higher than 1.79 ± 0.24 and 1.60 ± 0.22 in N-EMI group (P < 0.01), as shown in Figure 3(b). Te leukemia cells in the EMI group showed abnormally low expression of miR-3677-5p but high expression of CXCL12 and CXCR4 mRNA, indicating that the miR-3677-5p/ CXCL12 axis may be closely related to the occurrence of EMI in monocytic leukemia. Meanwhile, clinical data showed that the proportion of bone marrow blasts in patients of the EMI group was 68.4 ± 12.5%, signifcantly higher than the 46.5 ± 16.7% in the N-EMI group. Furthermore, the patients of the EMI group lacked sensitivity to chemotherapy, and the remission rate after chemotherapy was low, as shown in Table 1.

CXCL12 Is the Downstream Target Gene of miR-3677-5p.
Te specifc binding sites between miR-3677-5p and CXCL12 were predicted by using biological information database, as shown in Figure 6(a). Furthermore, luciferase assay showed that CXCL12 3′ UTR-Wt and miR-3677-5p mimic were co-transfected into 293T cells, and the luciferase activity was signifcantly decreased in compared with NC mimic group (P < 0.01). CXCL12 3′ UTR-Mut and miR-3677-5p mimic were co-transfected into 293T cells, but there was no signifcant diference between the two groups, as shown in Figure 6(b). Te Western blot results showed that overexpression of miR-3677-5p signifcantly inhibited the expression level of CXCL12 protein in SHI-1 cells (P < 0.01), as shown in Figure 6(c).

Te Expression of miR-3677-5p, CXCL12, and CXCR4 in SHI-1 Cells after Treatment of Rk3.
RT-qPCR showed that the expression level of miR-3677-5p in SHI-1 cells treated by Rk3 at 60, 120, and 180 mol/L was signifcantly higher than that in the control group, but the mRNA expression level of CXCL12 was signifcantly lower than that in the control group (P < 0.05 or 0.01), as shown in Figure 8 Evidence-Based Complementary and Alternative Medicine 7 expression level of CXCR4 in the Rk3 120 and 180 mol/L groups was signifcantly lower than that in the control group (P < 0.01). Te result of the Western blot was consistent with that of the RT-qPCR above. Te protein levels of both CXCL12 and CXCR4 in Rk3 treated cells were signifcantly lower than those in the control group (P < 0.05), as shown in Figure 8(b).

Rk3 Regulated the Expression of EMI-Related Functional
Proteins. Te results of immunofuorescence showed the fuorescence intensity of MMP2 protein related to promoting migration and invasion, which was signifcantly weakened in Rk3 treated SHI-1 cells, while the fuorescence intensity of TIMP2 related to inhibiting migration and invasion was signifcantly enhanced. Te results indicated that Rk3 could not only directly inhibit the expression of MMP2 protein but also upregulate the expression of TIMP2 protein, so as to enhance the inhibitory efect of Rk3 on MMP2 protein, as shown in Figure 9.

Te Efect of miR-3677-5p-Regulating CXCL12 on the Expression of EMI-Related Functional Proteins.
Te immunofuorescence results showed that compared with the miR-NC group and control group, the fuorescence intensity of MMP2 in the miR-3677-5p mimic group and AMD3100 group was signifcantly weakened, while the fuorescence intensity of TIMP2 was signifcantly enhanced, as shown in Figure 10.

Discussion
Acute monocytic leukemia belongs to type M5 of AML, classifed by FAB, with two diferent morphological subtypes (M5a and M5b), characterized by abnormal proliferation of immature monocytes in peripheral blood and bone marrow, eventually resulting in inefective hematopoiesis and bone marrow failure. Clinical complications are anemia, infection, bleeding, and so on. M5 patients often have EMI in the liver, spleen, lymph nodes, skin, and gums. Infltration of the central nervous system occurs mostly in the spinal cord RT-qPCR showed that the expression of miR-3677-5p in the EMI group was signifcantly lower than that in both the N-EMI group and the normal control group. However, the mRNA expression of CXCL12 and CXCR4 was higher than those of the N-EMI group and the normal control group. * * P < 0.01]s. the control group. 8 Evidence-Based Complementary and Alternative Medicine membrane, brain parenchyma, and spinal cord. Tat is, leukemia cells directly spread or hematogenous transfer into the central nervous system, which is one of the important factors leading to recurrence and death [24]. Te mechanism of EMI is a complex process of multistep and multifactor regulation, involving multiple factors of activation or inhibition. First, leukemia cells escape from bone marrow to peripheral blood, undergo chemotaxis, and adhere to the vascular endothelium through adhesion molecules, migrate and degrade the extracellular matrix, then survive and proliferate in extramedullary tissues, and fnally, form infltrating lesions [25]. In terms of cytogenetic factors, the occurrence of EMI is usually associated with the alterations in chromosome 8 alteration, FLT3-ITD and NPM1 mutations [26]. Most patients with EMI had a low remission rate and a short overall survival time after chemotherapy [27]. Cytotoxic chemotherapy drugs, targeted therapies, and immunotherapies cannot prevent and cure EMI, so it is necessary to research and develop a safe and efective new Chinese medicine as an adjuvant therapy for anti-EMI of leukemia. Ginsenoside Rk3 is isolated from total saponins of panax ginseng, it has been reported that Rk3 is efective to inhibit the proliferation and induce apoptosis of tumor cells in esophageal cancer, liver cancer, nonsmall cell lung cancer, and other tumors. For instance, Rk3 could decrease the expression of infammatory factors, arrest cell cycle, promote apoptosis of hepatocellular carcinoma cells, and therefore, inhibit the occurrence and development of hepatocellular carcinoma by targeting the LPS-TLR4 signaling pathway [28]. Rk3 could downregulate the expression of cyclin D1 and CDK4, upregulate the expression of P21 protein, and inhibit the proliferation and colony formation of lung cancer cells [29]. Consistently, in this study, we observed Rk3 was efective to inhibit the proliferation, migration, and invasion of SHI-1 cells, and regulate miR-3677-5p/CXCL12 axis and EMI-related functional proteins.
Some studies indicated that targeted upregulation of miR-3677 expression level would inhibit the proliferation and migration of human osteosarcoma cells [18]. Our study showed that the primary leukemia cells of patients in the EMI group showed abnormally low expression of miR-3677-5p. Meanwhile, these patients with EMI were not sensitive to chemotherapy with a low remission rate. After miR-3677-5p mimic was transfected, the expression of miR-3677-5p in SHI-1 cells increased, and the proliferation ability decreased signifcantly. Te Transwell experiment found that overexpression of miR-3677-5p could inhibit the migration and invasion in SHI-1 cells and regulate EMI-related functional proteins. Our results showed that Rk3 upregulated the  Figure 7: Intervening CXCL12 could inhibit proliferation, migration, and invasion of SHI-1 cells. (x ± s, n � 5) (a) Colony formation assay showed that both the AMD3100 and Rk3 could inhibit the proliferation of SHI-1 cells either alone or combination. (b) Transwell assay showed that the number of cell migrations and invasions of AMD3100 and Rk3 either alone or combination were signifcantly lower than those of the control group. * * P < 0.01 vs. the control group.
expression of miR-3677-5p. Te abovementioned results all indicated that miR-3677-5p could inhibit the malignant behavioral activities of monocytic leukemia cells and can be used as a specifc target for targeted therapy of monocytic leukemia. Te literature reported that CXCL12/CXCR4 axis was extensively involved in the migration and invasion of colorectal cancer and other cancers [30,31]. Adhesion molecules were identifed as key proteins in relation to tumor cell invasion, and CXCL12 could also upregulate the expression of adhesion molecules such as VLA-4. Te results of the GEPIA website showed that the expression levels of CXCL12 and CXCR4 in AML patients were signifcantly higher than those in healthy people. Meanwhile, our results showed that patients' leukemia cells in the EMI group showed abnormally high expression of CXCL12 and CXCR4 mRNA. In combination with the clinical situation, these patients with EMI were not sensitive to chemotherapy with a low remission rate. After interfering with the expression of CXCL12, the proliferation, migration, and invasion of SHI-1 cells were inhibited. In addition, TargetScan database found that miR-3677-5p and CXCL12 have specifc binding sites, it have been confrmed that miR-3677-5p could directly bind to the 3′ UTR of CXCL12 and inhibit its activity, by means of dual luciferase reporter gene assay. Te experimental results above indicate that the miR-3677-5p/CXCL12 axis may be closely related to the occurrence of EMI in monocytic leukemia. Terefore, miR-3677-5p and CXCL12 may be used as indicators of EMI and poor prognosis in monocytic leukemia. In view of the small number of cases in this study, the cases need to be expanded for further verifcation.
EMI and N-EMI have the same pathogenesis, but they are two manifestations of AML. Te reason for the diference between EMI and N-EMI was the heterogeneity and complexity of leukemia cells. Tat is, leukemia cells in the EMI group had a higher degree of malignancy and possessed extremely strong proliferation potential, so the proportion of bone marrow blasts in patients with the EMI group was signifcantly higher than that in the N-EMI group. Te potency of both extramedullary migration and invasion was very strong for leukemia cells in the EMI group, which is one of the important factors for poor prognosis.
Te MMPs are zinc-dependent proteolytic enzymes of the extracellular matrix (ECM) [32]. As a member of MMPs, MMP2 is involved in the invasion of a variety of tumor cells by digesting the extracellular matrix barrier [33][34][35][36][37]. Excessive secretion of MMP2 by leukemia cells would increase the permeability of the blood-brain barrier via disrupting the tight junction proteins, leading to enhance the invasion of leukemia cells into the central nervous system [38]. In addition, the TIMPs family is a tissue inhibitor of matrix metalloproteinases, among them, TIMP2 is capable of specifcally suppressing the expression and activity of MMP2. After overexpression of miR-3677-5p or interference with CXCL12 expression, the fuorescence intensity of MMP2 weakened, and the fuorescence intensity of TIMP2 enhanced. Our results showed that Rk3 downregulated the expression of MMP2, but upregulated the expression of TIMP2. It shows that the high expression of miR-3677-5p and the low expression of CXCL12 can signifcantly inhibit the cell EMI process, reducing the migration and invasion. In summary, our results indicated that the ginsenoside Rk3 is efective to inhibit proliferation, migration, and invasion by targeting miR-3677-5p/CXCL12 axis in SHI-1 cells, it provided experimental evidence for the role of Rk3 on antimonocytic leukemia, but to elucidate the exact mechanism remains to be further explored.

Data Availability
All data generated or analyzed during this study are included in this article.

Ethical Approval
Te patient's samples and clinical data of this study were approved by the Ethics Committee of the First Afliated Hospital of Zhejiang Chinese Medical University.

Conflicts of Interest
All authors declare that they have no conficts of interest.