Background
Medical advances have improved the survival of adult patients with acute lymphoblastic leukemia (ALL) over the past few decades. T cell acute lymphoblastic leukemia (T-ALL) is one of the most aggressive hematologic malignancies, accounting for up to 10–15% of pediatric ALL and 25% of adult ALL cases [
1], and arises from the malignant transformation of T cell progenitors [
2,
3]. Although high-dose multi-agent chemotherapy is clinically effective in most cases, primary drug resistance and relapse are frequently observed [
4], preventing T-ALL from being cured [
5,
6].
Recent studies have shown that the bone marrow milieu, especially mesenchymal stem cells (MSCs), has pro-survival effects on leukemia cells and protects leukemic cells from chemotherapy [
7‐
11]. On one hand, soluble factor-mediated drug resistance has been proposed to contribute to MSC-induced chemoresistance. Iwamoto and colleagues found that MSCs secreted asparagine that was taken up by ALL cells and thus protected ALL cells from asparaginase treatment [
12]. On the other hand, cell adhesion-mediated drug resistance is also an important mechanism of MSC-induced chemoresistance. For example, Mudry et al. reported that MSCs interacted with leukemia cells by increasing the expression of vascular cell adhesion molecule-1 (VCAM-1), protecting leukemia cells from cytarabine and etoposide cytotoxicity [
13]. However, the role of MSCs in T-ALL cell drug resistance remains unclear, and thus, intensive studies on the mechanisms by which MSCs protect T-ALL cells are needed to develop T-ALL treatments.
Excessive intracellular reactive oxygen species (ROS) can induce the apoptosis of cells [
14‐
16]. An important mechanism for chemotherapeutic agents is to induce cancer cell apoptosis by enhancing intracellular ROS levels. Such agents include paclitaxel, anthracyclines, ara-C, and methotrexate (MTX), among others [
17]. Mitochondria are the most important source of cellular ROS [
18‐
22]. Therefore, upregulating mitochondrial ROS levels is a potential strategy for killing cancer cells [
23,
24], including T-ALL cells [
25]. Jitschin et al. reported that induction of mitochondrial ROS in chronic lymphocytic leukemia (CLL) cells with PK11195, a drug that can generate mitochondrial superoxide, resulted in cell apoptosis [
26]. Our previous research showed that MSCs reduced mitochondrial ROS levels in T-ALL cells through the ERK pathway and thus protected T-ALL cells from chemotherapeutics ara-C or MTX. Accordingly, inhibition of the ERK activation with the ERK inhibitor PD325901 increased mitochondrial ROS levels and the cell death rate of T-ALL cells [
27]. These results indicated that MSCs protect T-ALL cells by decreasing mitochondrial ROS levels in T-ALL cells. Interestingly, in the past few years, several studies have reported that mitochondria can move between cells, through tunneling nanotubes (TNTs), microvesicles, or gap junctions, leading to protection against tissue injury or resistance to therapeutic agents [
21,
28‐
37]. However, there are few studies on the mechanisms of mitochondria transfer between MSCs and T-ALL cells. As mitochondrial ROS plays a major role in the intracellular redox balance [
38], it is important to determine whether mitochondrial transfer can be used to modify ROS levels.
In this study, we examined bidirectional mitochondria transfer between MSCs and T-ALL cells and found that T-ALL cells exposed to chemotherapeutic drugs transferred many mitochondria to MSCs but received few from MSCs. This process facilitates the proliferation and survival of leukemia cells by reducing ROS levels. Furthermore, by impeding cell adhesion, the mitochondria transfer was disturbed, therefore decreasing the survival rate under chemotherapy. These results suggest that mitochondria transfer may be a candidate target for T-ALL treatment.
Methods
Cell culture
Human T-ALL cell line Jurkat was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The culture medium was RPMI 1640 (Hyclone, Logan, UT, USA) supplemented with fetal bovine serum (FBS; Gibco, Grand Island, NY, USA), penicillin, and streptomycin (Sigma, St. Louis, MO, USA).
For collection of human primary T-ALL cells, 10 enrolled T-ALL patients were previously untreated and newly diagnosed at the Department of Haematology, Nanfang Hospital, Southern Medical University (Guangzhou, China), and Department of Pediatrics, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University (Guangzhou, China). Consent was provided according to the Declaration of Helsinki. Informed consent was obtained following institutional guidelines, and approval was obtained from the institutional review board of Sun Yat-Sen University. All human bone marrow or peripheral blood samples were obtained with written informed consent. Primary CD3+ T-ALL cells were isolated through density gradient centrifugation on standard Ficoll-HyPaque and subjected to fluorescence-activated cell sorting (FACS; BD Bioscience Influx, Franklin Lakes, NJ, USA).
MSCs were collected from bone marrow aspirates of healthy volunteers with informed consent. Isolation and characterization of MSCs were performed as we previously described [
39,
40]. Briefly, the bone marrow aspirates were diluted, have undergone the density gradient centrifugation, and were counted and planted before purification [
41,
42]. The culture medium was low-glucose DMEM (Hyclone, Logan, UT, USA) supplemented with 10% FBS and 100 IU/ml penicillin and streptomycin. To generate GFP-labeled MSCs, MSCs were transfected with lentivirus containing lentiviral expression vector pLV/puro-EF1a-GFP [
43] using the X-treme GENE HP reagent (Roche) according to the manufacturer’s instructions. Three days after transfection, GFP-labeled MSCs were purified by FACS (Influx, Becton Dickinson).
Several culture models were used in this article. (1) Monoculture: Jurkat cells/human primary T-ALL cells (5 × 105 /ml) or MSCs (5 × 104 /ml) were respectively seeded in 24-well plates. (2) Coculture: MSCs (5 × 104 /ml) and Jurkat cells/human primary T-ALL cells (5 × 105 /ml) were suspended in RPMI 1640 and seeded in 24-well plates. (3) Transwell: Jurkat cells/human primary T-ALL cells (5 × 105 /ml) were seeded in the Transwell inserts (Millipore), which were inserted into the 24-well plates with preseeded MSCs (5 × 104 /ml). The relative measurements were performed after coculture for 1 to 3 days.
Reagents and antibodies
Ara-C and MTX were purchased from Pharmacia Pty Ltd. (NSW, Australia) and Calbiochem (San Diego, CA, USA), respectively. 300nM ara-C or 100 nM MTX were used to cause cytotoxicity in Jurkat cells. 18-α-GA, dynasore, and cytochalasin D were purchased from Sigma-Aldrich and were used in a concentration of 50, 50, and 1 μM, respectively. Neutralizing anti-ICAM-1 antibody (MS305PABX) was purchased from Invitrogen (Carlsbad, CA, USA). Alexa Fluor™ 647 phalloidin was purchased from Thermo Fisher Scientific.
Cell viability assay
Cell viability was determined using a CCK-8 assay kit (Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer’s instructions. The principle of this assay is that some components of the CCK-8 assay kit will be reduced by mitochondria to produce formazan so as to be detectable. Briefly, 5 × 104 Jurkat cells in 100 μl culture media were plated to a 96-well plate in suspension. Then, the samples were incubated with CCK-8 solution (10 μl) for 4 h at 37 °C; the absorbance in each well was quantified at 450 nm using an automated enzyme-linked immunosorbent assay reader (Tecan, Salzburg, Austria). Cell viability was calculated according to the manufacturer’s instructions.
Fluorescence staining of mitochondria
MitoTracker Red (Molecular Probes) was used to label mitochondria. Jurkat cells or MSCs were incubated with 200 nM MitoTracker Red in culture media for 10 min at 37 °C. Excess of the dye was washed out with PBS. Then, 4 days later, stained cells were then seeded for monoculture and coculture. We verified the feasibility of mitochondria dye method in Additional file
1: Figure S1.
Staining of F-actin in MSCs
To visualize TNTs which consist of F-actin, MSCs were immersion-fixed in 4% paraformaldehyde. After a brief permeabilization with 0.1% Triton X-100, cell coverslips were incubated in AlexaFluor 647-conjugated phalloidin for 20 min at room temperature.
Mitochondrial ROS assessment
The levels of mitochondrial ROS were detected using the fluorescent probes MitoSOX™ Red (Molecular Probes, Life Technologies, Carlsbad, CA, USA), and fluorescent intensity was measured by flow cytometry (FACScan; Becton Dickinson, San Diego, CA, USA).
Annexin V/PI flow cytometry analysis
Jurkat cells from monoculture or coculture system were treated with ara-C or MTX for 2 days and harvested by centrifugation. Jurkat cells were then stained with annexin V/propidium iodide (PI) assay kit (BIOSCI BIOTECH, Shanghai, China) according to the manufacturer’s instruction. The apoptotic population was immediately evaluated by flow cytometry. The percentages of early apoptotic cells (annexin V+/PI−) and late apoptotic cells (annexin V+/PI+) were analyzed and graphed.
RNA isolation and qRT-PCR analysis
Total mRNA from MSCs was extracted using an RNeasy Mini Kit (Qiagen), and complementary DNA (cDNA) was synthesized using a QuantiTect Reverse Transcription Kit (Qiagen) according to the manufacturers’ protocols. qRT-PCRs were carried out using SYBR Green qPCR SuperMix (Roche, Indianapolis, IN, USA) and a LightCycler 480 Detection System (Roche) as described by the manufacturer. Target mRNA levels were normalized with respect to those of β-actin. The primer sequences used for qRT-PCR are listed in Additional file
1: Table S1.
Statistical analyses
All experiments were performed at least three separate times. All data are expressed as the mean ± S.E.M. Comparisons among groups were performed using one-way analysis of variance (ANOVA) or Student’s t test. Statistical differences were determined by GraphPad Prism 5.0 software (GraphPad Software Inc., CA, USA). A two-sided P value < 0.05 was considered to be statistically significant.
For the other experimental procedures, please see Additional file
1.
Discussion
As one of the most aggressive hematologic malignancies, T-ALL is usually treated with multiple chemotherapeutic drugs clinically, but observations of primary drug resistance during treatment are quite frequent. Although it is widely accepted that MSCs are involved in the pro-survival effects, the exact role of MSCs under the chemotherapy remains unclear. Here, we demonstrated that upon the induction of oxidative stress by chemotherapeutic drugs, T-ALL cells were able to transfer mitochondria to MSCs. This process was mediated by TNTs and ICAM-1 contributing to the cell adhesion-mediated drug resistance. A graphical abstract is shown to describe this mechanism briefly (Fig.
6f).
MSCs can trigger the drug resistance of tumor cells via two main strategies, soluble factor-mediated drug resistance and cell adhesion-mediated drug resistance [
47]. For the former, drug resistance can be triggered by MSCs secreting cytokines, chemokines, growth factors [
27], and exosomes [
48], and for the latter, MSCs can induce drug resistance by adhering to cancer cells, including melanoma cells [
40] and leukemia cells [
49]. In this study, we found that mitochondria transfer between MSCs and T-ALL cells is also a mechanism that induces chemoresistance in tumor cells. According to the literature, intercellular mitochondria transfer can be mediated by TNTs, microvesicles, or gap junctions. We confirmed that TNTs played a major role in the mitochondria transfer between MSCs and T-ALL cells, and inhibition of TNTs led to decreased MSC-induced chemoresistance. This finding may thus provide a novel strategy for T-ALL treatment.
Many studies have demonstrated that the intercellular transferred mitochondria were still functional and can affect cellular fate. For example, mitochondria transferred from MSCs to tumor cells could increase the oxidative phosphorylation and ATP production [
47]. Some cancer cells could also induce stromal cells to produce oncometabolites to fuel their metabolism through mitochondria transfer [
50]. Meanwhile, it is reported that mitochondrial loss in MSCs could decrease ATP concentrations in these cells, thereby decreasing their secretory capacity and interfering the cytokine secretion which played an important role in maintaining the microenvironment [
28]. On the other hand, MSCs might eliminate the transferred mitochondria to stabilize the intracellular homeostasis. Phinney et al. firstly figured out that MSCs eliminated their damaged mitochondria by exporting them to neighboring macrophages for recycling [
51], so as to decrease the oxidative pressure in the microenvironment, associated with better survival and increased regrowth potential. In our study, mitochondria transfer helps reduce the ROS level in Jurkat cells, so as to induce chemoresistance. Taken together, investigating the fate of transferred mitochondria helps to understand the crosstalk in leukemic microenvironment and offers probable therapy strategies.
In contrast to the myriad reports demonstrating that MSCs could transfer mitochondria to various kinds of cells, including cortical neurons [
52], cardiomyocytes [
53], renal tubular cells [
30], lung epithelium cells [
28], lung adenocarcinoma cells [
32], osteosarcoma cells [
54], macrophage [
51], and acute myeloid leukemia cells [
55], mitochondria transfer from other cells to MSCs has rarely been demonstrated. In our study, we found that T-ALL cells could transfer mitochondria to adhering MSCs when treated with ara-C or MTX. This newly identified process complements the existing knowledge of mitochondria transfer and provides a novel perspective regarding mitochondria transfer in intercellular relationships. On the other hand, in cancers such as acute myeloid leukemia (AML), MSCs are impaired in their growth properties and osteogenic differentiation potential [
56]. Interestingly, here, we demonstrated that MSCs could receive mitochondria from T-ALL cells, and it is likely that this transfer would lead to MSC damage in T-ALL patients. Since MSCs play an important role in tissue repair [
57,
58], they are worthy of further investigation. Moreover, it is reported that AML cells can import mitochondria from MSCs in order to better withstand chemotherapy [
55,
59]. Thus, we verified our conclusion by comparing ALL cells with AML cells and found they have different adhesive capacity and mitochondria transfer direction (Additional file
1: Figure S5). The difference in transfer direction may also be due to their different metabolic state. T-ALL cells prefer glycolysis after coculture, while AML cells have more oxidative phosphorylation [
60]. ALL cells export mitochondria to reduce intracellular ROS, while AML cells import mitochondria for the demand of oxidative phosphorylation.
Disrupted oxidative stress metabolism is a common feature of cancer cells [
61,
62], and this phenomenon has also been observed in T-ALL cells [
63]. As a result, ROS levels are higher in T-ALL cells than in non-leukemic cells. Since excess ROS can lead to leukemia cell deaths, the induction of intracellular oxidative stress has been shown to be an important anti-cancer mechanism of leukemia chemotherapy [
64]. Thus, the promotion of mitochondrial ROS production can be observed in T-ALL cells treated with paclitaxel, anthracyclines, ara-C, and MTX, among others
. In our previous study, we found that MSC-mediated chemoresistance of T-ALL cells was dependent on decreased mitochondrial ROS in T-ALL cells, and the ERK/Drp1 signaling pathway was involved in the downregulation of ROS levels. However, mitochondrial ROS in T-ALL cells decreased to a larger extent in a coculture system than in a Transwell system, indicating that an unknown mechanism mediated the MSC-induced chemoresistance. Here, we found that cell adhesion-mediated mitochondria transfer from T-ALL cells to MSCs can reduce oxidative stress by decreasing mitochondrial ROS. This finding solved the question raised by our previous study. Additionally, Ishikawa et al. observed the direct transfer of intercellular ROS mediated by connexin-43 from hematopoietic stem cells to bone marrow stromal cells [
65]. This finding suggested that the direct transfer of mitochondrial ROS via TNTs might also be another mechanism for decreasing ROS in T-ALL cells. Unfortunately, due to the limitations of the existing experimental techniques, the two mechanisms have yet to be discriminated.
Additionally, we also found that treatment with anti-ICAM-1 significantly blocked mitochondria transfer, indicating that mitochondria transfer was mediated by T-ALL cell/MSC adhesion. Combined with our finding that blocking mitochondria transfer with cytochalasin D abolished the capacity of MSCs to protect T-ALL cells, we concluded that T-ALL cell/MSC adhesion-mediated mitochondria transfer contributed to MSC-induced chemoresistance. Thus, inhibition of T-ALL cell/MSC adhesion-mediated mitochondria transfer may be a novel strategy for T-ALL treatment.
Acknowledgements
We thank the Department of Haematology in Nanfang Hospital and Department of Pediatrics in Sun Yat-Sen Memorial Hospital for sample collection.