Background
As a malignant hematological condition, adult B-cell acute lymphoblastic leukemia (B-ALL) is caused by the abnormal proliferation of B-lymphocyte lineage precursor cells in the bone marrow. Studies have shown that complete remission is achievable in most patients with B-ALL after receiving induction chemotherapy [
1,
2]. However, the recurrence, drug resistance and extramedullary infiltration during treatment remain the chief reasons for the poor efficacy in B-ALL patients, and the long-term efficacy is not optimistic [
3,
4].
The tumor microenvironment (TME) is a dynamic network, which can offer a supportive environment for the emergence and development of tumor cells [
5]. Lately, the indispensable and decisive role of TME in tumor progression has attracted a growing scholarly attention [
6,
7]. During the tumorigenesis and evolution, tumor cells exist in close proximity to their surrounding stromal microenvironment, and due to the extensive “cross-talks” between them, TME can promote their survival and metastasis [
8,
9].
As a kind of crucial stromal cells in the TME, bone marrow mesenchymal stem cells (BM-MSCs) are often important factors in protecting leukemia cells from chemotherapeutics and promoting their migration and invasion [
10‐
12]. However, in vitro experiments, previous reports have shown the presence of certain differences between MSCs cultured alone and MSCs in a leukemia niche [
13,
14], and the later could be simulated by constructing a co-culture system of leukemia cells and BM-MSCs in vitro [
13]. To our knowledge, the differences in biological functions and gene expression profiles between mono-cultured MSCs and MSCs in B-ALL-derived leukemia niches remain unclear in some respects.
In the present work, it was revealed that MSCs in the leukemia niche have varying degrees of changes in biological characteristics and gene expression profiles. Through differentially expressed genes (DEGs) analysis, we focused on the abnormal expression of interferon alpha-inducible protein 6 (IFI6), which is a gene activated by interferon capable of facilitating the tumorigenesis and development of several solid tumors [
15‐
17]. We elucidated IFI6's positive role in the leukemia cell multiplication, and explored the underlying molecular mechanisms. These findings perhaps offer a novel theoretical and experimental foundation for unraveling the leukemia evolution-facilitating mechanism of the microenvironment, and for finding potential therapeutic targets.
Methods
Cell cultures
Human ALL (Nalm-6 and RS4;11) cells, which were acquired from the Laboratory of Hematopoietic Stem Cell Transplantation Center of Guizhou Province (Guiyang, China), were cultured in a RPMI-1640 medium involving FBS (10%), streptomycin (100 mg/mL) and penicillin (100 units/mL) at 37 °C with 5% CO2.
For the acquisition of BM-MSCs, we separated BM-MSCs from the bone marrow aspirates of B-ALL patients (n = 37) with their consent. The aspirates were subjected to ficoll gradient centrifugation, and subsequently cultured in the complete medium of adult BM-MSCs (Saiye, Shanghai, China). The phenotypic traits and differentiation capacities of BM-MSCs have been confirmed by us priorly [
18].
Co-culture system and establishment of leukemia niche in vitro
For the construction of co-culture system, the culture plate was initially seeded with BM-MSCs at 1 × 10
5/ml for 24 h, and then added with leukemia cells at a 4:1 ratio. For the establishment of the leukemia niche, MSCs were subjected to a 72-h co-cultivation with leukemia cells with reference to previously published literature [
13]. In each independent repeated experiment, separately cultured MSCs from the same patient as control. For the collection of suspended leukemia cells in the co-culture system, Nalm-6/RS4;11 cells were pipetted cautiously from the monolayer MSCs. For some leukemia cells adhering to MSCs surface, PBS-EDTA 1 was used to wash the co-culture extensively to remove all the leukemia cells.
Reagents and antibodies
Vincristine sulfate, which was product of Taoshu Biotechnology (Shanghai, China), was prepared in phosphate buffer solution (PBS). AMD3100 and PD98059 from MedChemExpress (Shanghai, China) were formulated separately in dimethyl sulfoxide (DMSO) and anhydrous ethanol. The AKT, ERK, phospho-AKT and phospho-ERK antibodies were procured from Cell Signaling Technology (Danvers, MA, USA)., while the CXCR4 and IFI6 antibodies were procured separately from Solibao Biotechnology (Beijing, China) and ImmunoWay Biotechnology (Plano, TX, USA). Proteintech Group (Wuhan, China) was the provider of the secondary antibody used herein for Western blot.
Cell proliferation
MSCs (500 ul) were seeded in a 24-well culture plate at 5 × 104/ml. After the MSCs adhered overnight, the MSCs culture medium was aspirated, and 500 ul of Nalm-6/RS4;11 cells (1-3 × 105/ml) were inoculated on MSCs. The mono-cultured Nalm-6/RS4;11 cells were set as a control, where RPMI 1640 medium was used as a culture medium. The leukemia cells in each well were collected after incubation for 24, 48, 72 and 96 h, respectively. Then, the leukemia cells were quantified by direct counting using a cell counter, and the cell growth curves were drawn according to the results.
Apoptosis and cell cycles
After harvesting and PBS-washing, the cells were subjected to Annexin-V/propidium iodide (PI) staining to assay the apoptotic ratio as per the advised protocol (7Sea Pharmatech, Shanghai, China). To assess the cell cycle, RNase A and PI (7Sea Pharmatech, Shanghai, China) were utilized to treat the gathered MSCs, followed by flow cytometry (BD Biosciences, San Jose, USA).
Cell migration and invasion
Matrigel-coated and uncoated Transwell chambers were used for invasion and migration experiments, respectively. The lower chamber was added with 650 ul of MSCs (1 × 105/mL) and incubated overnight. Next, the upper chamber (pore size: 8.0 um, Corning Incorporated, Costar) was added with 100 ul of Nalm-6/RS4; 11 cells (4 × 105/mL) and subjected to a 24-h incubation. An inverted microscope was utilized to surveil the leukemia cell migration and invasion in the lower chamber, followed by photographing under 40 × magnification. Quantification of migrated and invaded cells was accomplished by direct counting using a cell counter.
β-galactosidase staining
For adherent BM-MSCs, after washing with PBS in a 6-well plate, the procedure for β-galactosidase staining was implemented as per the specific guidelines. then followed by observation of the proportion of cells stained blue (senescent cells) under a microscope and photographing (200 ×). The proportion of senescent cells in each group was evaluated with 5 random fields of view, and then the senescence of cells in each group was assessed by value averaging.
Lentiviral transduction
Human IFI6-silencing RNA (si-IFI6) and IFI6-overexpressing clone lentiviral particles (LV-IFI6) were the Genechem (Shanghai, China) products. Transfection of si-IFI6/LV-IFI6 was accomplished as per the manufacturer protocol. Controls used were the empty vector (EV)-transfected BM-MSCs.
Quantitative real-time PCR
Trizol reagent (Qiagen, Hilden, Germany) was utilized to extract the total RNA of cells, the FastKinggDNA Dispelling RT SuperMix (Qiagen, Hilden, Germany) was utilized to reversely transcribe the RNA extract to cDNA. Then, cDNA was analyzed by quantitative real-time PCR (qRT-PCR) in accordance with the protocols of primers and Talent qPCR PreMix (SYBR Green) (Qiagen, Hilden, Germany). For the target gene, their relative expression levels were estimated by the comparative CT (2
−△CT) approach following normalization to β-actin. Table
1 details the human primers (Generay Bioteach, Shanghai, China) used.
Table 1
Primer sequences used for real-time PCR
hMX-1 (F) | GTTTCCGAAGTGGACATCGCA |
hMX-1 (R) | CTGCACAGGTTGTTCTCAGC |
hIFITM1 (F) | CCAAGGTCCACCGTGATTAAC |
hIFITM1 (R) | ACCAGTTCAAGAAGAGGGTGTT |
hIFIT3 (F) | AAAAGCCCAACAACCCAGAAT |
hIFIT3 (R) | CGTATTGGTTATCAGGACTCAGC |
hISG15 (F) | CGCAGATCACCCAGAAGATCG |
hISG15 (R) | TTCGTCGCATTTGTCCACCA |
hIFI6 (F) | GGTCTGCGATCCTGAATGGG |
hIFI6 (R) | TCACTATCGAGATACTTGTGGGT |
hIFI44L (F) | ACAGAGCCAAATGATTCCCTATG |
hIFI44L (R) | TCGATAAACGACACACCAGTTG |
hIFIT1 (F) | AGAAGCAGGCAATCACAGAAAA |
hIFIT1 (R) | CTGAAACCGACCATAGTGGAAAT |
hP-53 (F) | CTGCCCTCAACAAGATGTTTTG |
hP-53 (R) | CTATCTGAGCAGCGCTCATGG |
hP-21 (F) | GCCTGGACTGTTTTCTCTCG |
hP-21 (R) | ATTCAGCATTGTGGGAGGAG |
hP-16(F) | GAAGGTCCCTCAGACATCCCC |
hP-16 (R) | CCCTGTAGGACCTTCGGTGAC |
hSOX2 (F) | GCCGAGTGGAAACTTTTGTCG |
hSOX2 (R) | GGCAGCGTGTACTTATCCTTCT |
hNANOG (F) | TTTGTGGGCCTGAAGAAAACT |
hNANOG (R) | AGGGCTGTCCTGAATAAGCAG |
hADIPOQ (F) | AACATGCCCATTCGCTTTACC |
hADIPOQ (R) | TAGGCAAAGTAGTACAGCCCA |
hPPAR-γ (F) | GGGATCAGCTCCGTGGATCT |
hPPAR-γ (R) | TGCACTTTGGTACTCTTGAAGTT |
hRUNX2 (F) | CCGCCTCAGTGATTTAGGGC |
hRUNX2 (R) | GGGTCTGTAATCTGACTCTGTCC |
hBGLAP (F) | CACTCCTCGCCCTATTGGC |
hBGLAP (R) | CCCTCCTGCTTGGACACAAAG |
hRUNX1 (F) | CTGCCCATCGCTTTCAAGGT |
hRUNX1 (R) | GCCGAGTAGTTTTCATCATTGCC |
hHOXB4 (F) | CGTGAGCACGGTAAACCCC |
hHOXB4 (R) | CGAGCGGATCTTGGTGTTG |
hPOU5F1 (F) | CTTGAATCCCGAATGGAAAGGG |
hPOU5F1 (R) | GTGTATATCCCAGGGTGATCCTC |
hSDF-1 (F) | CACTTTAGCTTCGGGTCAATG |
hSDF-1 (R) | ACACTCCAAACTGTGCCCTTCA |
hβ-actin (F) | CTACCTCATGAAGATCCTCACCGA |
hβ-actin (R) | TTCTCCTTAATGTCACGCACGATT |
Western blotting
Cells were collected and lysed with Radio ImmunoprecipitationAssay (RIPA) lysis buffer (Beyotime, Shanghai, China) involving phenylmethylsulfonyl fluoride (PMSF; 1%). Protein (10–30 μg) was isolated on SDS-PAGE and then shifted onto the PVDF membrane. A 1–2 h blockade of the membrane proceeded using skimmed milk (5%) at room temperature, followed by an overnight incubation using primary antibodies at 4 ℃. Thereafter, an extra 1-h incubation of the membranes was accomplished using secondary antibody at ambient temperature, and then the protein expression was assayed with electrochemiluminescence reagent. Gray value analysis was performed via the Image J software against the internal β-actin reference.
Transcriptome sequencing analysis
For transcriptome sequencing analysis, the processed MSCs and leukemia cells were collected respectively. After cells were lysed by trizol reagent, the samples were sent to Shanghai Liebing Information Technology and Hangzhou Lianchuan Biotechnology for transcriptome sequencing analyses. Meanwhile, the gene expression profile dataset GSE101454 related to MSCs derived from patients with B-Cell Precursor ALL (BCP-ALL) in GEO database was selected. This dataset provided microarray analysis data of mono-cultured BM-MSCs and BCP-ALL cells- co-cultured BM-MSCs (for 40 h).
Analysis of gene datasets and screening of DEGs
The DEGs in the datasets were analyzed and screened, where the screening conditions were: |log2FC|> 1, Flase Discovery Rate (FDR) < 0.05, P < 0.05. During the experimentation, Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment were performed on the DEGs via the Database For Annotation Visualization and Integrated Discovery (DAVID) software, and the biological functions and signaling pathway changes enriched by the differential genes were screened out. Gene set enrichment analysis (GSEA) 4.2.3 was utilized to assess the enrichment of gene sets.
Xenografted tumor model
Xenograft assays in mice were approved by the Guizhou Medical University's Animal Care Welfare Committee. 4–6 weeks-old female nonobese diabetes/severe combined immunodeficiency (NOD/SCID) mice were chosen, each of which was given subcutaneous injection of RS4;11/Nalm-6 cells, a RS4;11/Nalm-6-MSC mixture, a RS4;11/Nalm-6-MSCs-EV mixture, and a RS4;11/Nalm-6-MSCs-LV-IFI6 mixture (1 × 106 MSCs mixed with 4 × 106 RS4;11/Nalm-6 cells) into the left chest. Growth of tumors was surveilled every 3 d through length (L) and width (W) determination, and the computational formula for tumor volume was: 0.5 LW2. After injection for 25d/34d, the tumor tissues were extracted from mice and embedded in paraffin for further study. The procedure for immunohistochemistry (IHC) staining was implemented as per the specific guidelines. The primary and secondary antibodies used were 1:100 dilutions.
Statistical analysis
SPSS 20.0 was utilized to assess the data. Independent student's t-test was employed for making two groups comparison, while one-way ANOVA was adopted for homogeneity of variance assessment among multiple groups. Non-normal data were subjected to the Kruskal–Wallis non-parametric test. P-values were indicated as follows: *P < 0.05;**P < 0.01; ***P < 0.001.
Discussion
For leukemia cells, a vital site for their survival and multiplication is the bone marrow microenvironment. TME has been reported to function crucially in the occurrence and development of leukemia [
19‐
21]. BM-MSCs, as a pivotal microenvironmental constituent of bone marrow, are the main factor promoting leukemia progression in the TME [
22,
23]. However, it remains unclear what changes have taken place in MSCs in the B-ALL niches. Through establishment of the leukemia microenvironment in vitro, this study found that MSCs in leukemia niche had undergone changes in several aspects.
Regarding the stemness markers of MSCs, Zhang et al. [
24] revealed that the stemness of leukemia cell-derived MSCs did not change significantly compared with donor-derived MSCs. Nevertheless, whether MSCs are altered in leukemic niches is still unknown. In the present work, the difference of the stemness between MSCs in the leukemic niche and MSCs in mono-culture was not significant. Regarding the self-renewal traits of MSCs, Vanegas et al. [
14] reported that MSCs in the co-culture system of MSCs and REH cells showed increased expression. But in our study, these indicators just presented upward trends, without revealing statistical differences. In terms of the multi-directional differentiation potential of MSCs, Zhao et al. [
25] showed that MSCs derived from ALL were similar to normal MSCs, while Vicente et al. [
26] suggested that ALL-MSCs have increased adipogenic capacity compared to normal MSCs. In our current work, MSCs in the leukemia niche had attenuated osteogenic and adipogenic differentiation abilities, and displayed varying degrees of senescence changes, showing agreement with prior studies by Bonilla et al. [
13], Yang et al. [
27] and Vanegas et al. [
14]. Yuan et al. [
28] also showed that after co-culture of MSCs with T-ALL cell-derived extracellular vesicles, the MSCs exhibited suppressed differentiation towards osteogenesis. In terms of apoptosis and cell cycle, we just observed changes in the MSC–Nalm-6 co-culture system. Bonilla et al. [
13] found that the cell cycle of MSCs in the MSCs-REH cells co-culture system showed stagnation in G2/M phase. These results suggested that the leukemia niches constructed by different cell lines might have inconsistent experimental results, but the MSCs in the leukemia niches indeed have some changes in several aspects.
For the gene expression profiles, the current study found that MSCs in leukemia niches had significant expression changes. In a previous study, the data also showed different gene expression profiling in MSCs co-cultured with primary BCP-ALL cells; besides, survival benefit was observed in leukemia cells after co-culture with MSCs [
29]. This finding is consistent with ours. As for the possible biological functions of MSCs, studies have shown that MSCs could promote leukemia progression [
10,
18]. In the present work, we found that the differential genes of MSCs in leukemia niche were enriched to include several biological functions that promote tumor progression, which suggested that MSCs might be critical to the persistence and deterioration of leukemia cells.
To further describe how DEGs in MSCs affect the leukemia cells, in this study, we screened IFI6, an interferon-stimulated gene, which though has not been explored in B-ALL. During the occurrence and development of viral infectious diseases [
30,
31], autoimmune diseases [
32,
33] and some tumors [
16,
34‐
37], IFI6 is often highly expressed, which exerts the functions of resisting apoptosis and viruses, as well as promoting tumor progression. Liu et al. [
17] found that IFI6 was increased in patients with esophageal squamous cell carcinoma, the overexpression of IFI6 was closely related to the invasive phenotype and poor outcome. In an ovarian cancer research, the overexpression of IFI6 could facilitate the multiplication of tumor cells and mediate their chemoresistance [
15]. In addition, IFI6 is regarded as a crucial predictor of poor outcome in breast cancer [
34,
38]. In the study of hematological tumors, aberrantly expressed IFI6 in multiple myeloma is an important factor leading to the chemoresistance of myeloma cells [
39]. In the present work, we found that IFI6 might be an important component in the B-ALL microenvironment that promotes the proliferation of leukemia cells.
Regarding the tumor progression-promoting mechanism of IFI6 at the molecular level, Cheriyath et al. [
39] found that IFI6 regulated the balance between Bcl-2 and Bim expression to resist apoptosis. And Liu et al. [
17] revealed that the mitochondrial Ca
2+ overload could be induced by down-regulation of IFI6 to induce tumor cells apoptosis. In this study, we explored and found that highly expressed IFI6 in MSCs promoted the activation of the SDF-1/CXCR4 axis initiation in the TME, which served as a mediator in the stromal component–tumor cell interaction [
40,
41]. Multiple studies have attempted to make leukemia cells more sensitive to chemotherapeutics by disrupting their interaction using the CXCR4 inhibitor AMD3100 [
42‐
44]. In this study, AMD3100 also effectively attenuate the pro-proliferative effect of IFI6 on leukemia cells. AKT and ERK signaling pathways are key pathways that promote tumor progression in leukemia [
45,
46]. In the present work, IFI6 was found capable of initiating the ERK signaling pathway via the SDF-1/CXCR4 axis, thereby facilitating the leukemia cell multiplication. Suggesting that targeting ERK pathway in leukemia niches is probably a valid strategic option for reducing the leukemia cell multiplication.
Finally, we also found in this work that the increased expression of IFI6 in MSCs had some effects on the gene expression profile and biological functions of leukemia cells through RNA sequencing. Although this has not been reported in other studies, due to the small number of differential genes enriched in the GO/KEGG entries of our interest in this dataset and the small differences between the two groups, further study is needed. In addition, although this study interestingly found that increased expression of IFI6 in MSCs might be a key factor leading to the proliferation of B-ALL cells through in vitro and in vivo experiments, the current exploration is preliminary and limited to the B-ALL cell lines, more in-depth studies are needed to demonstrate the role of IFI6 in ALL.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.