Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jun 29.
Published in final edited form as: Br J Haematol. 2015 May 14;170(5):704–718. doi: 10.1111/bjh.13492

Stromal CYR61 Confers Resistance to Mitoxantrone via Spleen Tyrosine Kinase Activation in Human Acute Myeloid Leukaemia

Xin Long 1, Yang Yu 1, Laszlo Perlaky 1,2, Tsz-Kwong Man 1, Michele S Redell 1,2
PMCID: PMC5490799  NIHMSID: NIHMS869996  PMID: 25974135

Summary

Approximately 50% of children with acute myeloid leukaemia (AML) relapse, despite aggressive chemotherapy. The bone marrow stromal environment protects leukaemia cells from chemotherapy (i.e., stroma-induced chemoresistance), eventually leading to recurrence. Our goal is to delineate the mechanisms underlying stroma-mediated chemoresistance in AML. We used two human bone marrow stromal cell lines, HS-5 and HS-27A, which are equally effective in protecting AML cells from chemotherapy-induced apoptosis in AML-stromal co-cultures. We found that CYR61 was highly expressed by stromal cells, and was upregulated in AML cells by both stromal cell lines. CYR61 is a secreted matricellular protein and is associated with cell-intrinsic chemoresistance in other malignancies. Here, we show that blocking stromal CYR61 activity, by neutralization or RNAi, increased mitoxantrone-induced apoptosis in AML cells in AML-stromal co-cultures, providing functional evidence for its role in stroma-mediated chemoresistance. Further, we found that spleen tyrosine kinase (SYK) mediates CYR61 signalling. Exposure to stroma increased SYK expression and activation in AML cells, and this increase required CYR61. SYK inhibition reduced stroma-dependent mitoxantrone resistance in the presence of CYR61, but not in its absence. Therefore, SYK is downstream of CYR61 and contributes to CYR61-mediated mitoxantrone resistance. The CYR61-SYK pathway represents a potential target for reducing stroma-induced chemoresistance.

Keywords: marrow stroma, adhesion, drug resistance, signalling, myeloid leukaemia

Acute myeloid leukaemia (AML) is a devastating malignancy of the bone marrow in which haematopoietic precursors undergo uncontrolled proliferation and arrest in an early stage of development as myeloid blasts. Childhood AML has a high relapse rate and subsequent low survival rate, despite very toxic chemotherapy (Rubnitz et al, 2010; Gamis et al, 2014). Therefore new, rational therapies to circumvent chemotherapy resistance are desperately needed.

De novo chemotherapy resistance is present at the outset, in contrast to acquired chemotherapy resistance, which develops as a result of selective pressure during treatment (Meads et al, 2009). De novo resistance includes environment-mediated drug resistance (EMDR) (Meads et al, 2008, 2009). The tumour microenvironment is composed of stromal cells, extracellular matrix (ECM) and soluble factors, all of which can contribute to drug resistance (Matsunaga et al, 2003; Li & Dalton, 2006; Meads et al, 2008; Shain et al, 2009). Human and mouse bone marrow stromal cells have been shown to protect cells of various haematological malignancies against chemotherapy-induced apoptosis, thereby allowing disease recurrence (Mudry et al, 2000; Konopleva et al, 2002; Nefedova et al, 2003; Sison & Brown, 2011). In one salient report, over 700 anti-cancer drugs were less active against cancer cells when co-cultured with stromal cells (McMillin et al, 2010).

Stroma-mediated drug resistance includes soluble factor-dependent drug resistance and contact-dependent drug resistance (Meads et al, 2008, 2009). Soluble factors refer to cytokines, chemokines and growth factors secreted by stromal cells, which bind to their receptors and activate signalling pathways in malignant cells (Li & Dalton, 2006). Contact-dependent interactions exist between tumour and stromal cells, and between tumour cells and ECM. The ECM is composed of large molecular weight proteins like collagens and fibronectin, as well as smaller proteins, including the CCN family of matricellular ligands. The CCN family was named after cysteine-rich 61 (CYR61/CCN1), connective tissue growth factor (CTGF/CCN2) and nephroblastoma overexpressed (NOV/CCN3) (Perbal, 2004). As a secreted protein, CYR61 binds to integrins (Perbal, 2004), and has been linked to chemotherapy resistance in solid tumours (Lin et al, 2004; Menendez et al, 2005; Lai et al, 2011). However, the contribution of CYR61 to stroma-induced chemoresistance in leukaemia has not yet been described.

In the present study, we sought to determine the underlying mechanisms by which the bone marrow stroma protects AML cells from chemotherapy. Two human bone marrow stromal cell lines, HS-5 and HS-27A, were used. Both provide physical contact with malignant cells, but HS-5 cells secrete many more soluble factors than HS-27A cells (Roecklein & Torok-Storb, 1995). Using an in vitro AML-stroma co-culture model, we identified CYR61 as a mediator of resistance to mitoxantrone. Further, we demonstrated that CYR61-dependent resistance occurs via activation of SYK. SYK is a non-receptor tyrosine kinase widely expressed in haematopoietic cells. SYK is activated downstream of integrin ligation, and inhibition of SYK has been shown to induce apoptosis in AML models (Hahn et al, 2009; Miller et al, 2013). This is the first report of a CYR61-SYK pathway mediating stroma-induced chemotherapy resistance in AML.

Materials and methods

Reagents

Etoposide, mitoxantrone and cytarabine were obtained from the Texas Children’s Hospital pharmacy. Selective SYK inhibitor R406 was purchased from Selleck (Houston, TX, USA). CYR61 neutralizing antibody was purchased from Novus Biologicals (Littleton, CO, USA). Normal rabbit IgG was from R&D Systems (Minneapolis, MN, USA). Hoechst 33342 was from Sigma-Aldrich (St. Louis, MO, USA).

Cell lines

NB-4 and THP-1 human AML cells were gifts of Drs. Shuo Dong and Terzah Horton (Baylor College of Medicine, Houston, TX, USA), respectively. Kasumi-1 human AML cells, HS-5, HS-27A human stromal cells and HEK293T cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). HS-5 and HS-27A cells were stably transduced with pCDH.CMV.mOrange (gift of Dr. Stephen Gottschalk, Baylor College of Medicine), to allow distinction from AML cells by flow cytometry. For all experiments except where indicated, HS-5-mOrange and HS-27A-mOrange cells are denoted as HS-5 and HS-27A cells. HS-5 and HEK293T cells were maintained in Dulbecco’s modified Eagle’s medium (Gibco/Life Technologies, Grand Island, NY, USA) with 10% foetal bovine serum (FBS, Invitrogen/Life Technologies Carlsbad, CA, USA), 2 mM L-Glutamine (Invitrogen), 100 units/ml penicillin and 100 μg/ml streptomycin (Pen/Strep; Invitrogen). All other cell lines were maintained in RPMI-1640 medium (ATCC) with 10% FBS and other additives as above. All cells were grown in a 90% humidified, 37°C incubator with 5% CO2.

Primary AML samples

De-identified primary paediatric AML samples were obtained from the Research and Tissue Support Services (RTSS) Core of Texas Children’s Cancer and Hematology Centers at Texas Children’s Hospital and from the Children’s Oncology Group (COG) AML Reference Laboratory (Fred Hutchinson Cancer Research Center, Seattle, WA, USA). All samples were enriched for mononuclear cells by density centrifugation, and cryopreserved. Samples were carefully thawed into warm Iscove’s minimal essential medium with 20% FBS, and rested for 2 h before counting viable cells. Only samples with <40% spontaneous apoptosis after overnight culture were analysed. Written informed consent was obtained in accordance with the Declaration of Helsinki, for bone marrow and peripheral blood cells to be stored for research. These studies were approved by the Institutional Review Board of Baylor College of Medicine. Basic clinical data associated with the samples used in this study are provided in Table SI.

Co-culture and Transwell co-culture experiments

For co-culture experiments, human AML cell lines (NB-4, THP-1) or primary AML patient samples were seeded at 1 × 105 cells/ml and co-cultured with HS-5 or HS-27A cells pre-seeded the day before (ratio of AML:stromal cells 1:1), for 1 h (primary patient samples) or 24 h (cell lines). Cells were then treated with etoposide, mitoxantrone or cytarabine at increasing doses for 24 h before performing Annexin V apoptosis assays. As a control, AML cells were cultured alone and treated with chemotherapy drugs in parallel. For the Transwell® co-culture experiments, AML cell lines were plated inside the Transwell® microporous inserts (0·4 μm pore size, polyester or polycarbonate) while stromal cells were seeded on the Transwell® plate surface (Corning Incorporated, Corning, NY, USA), underneath the inserts (ratio of AML:stromal cells 1:1), thus eliminating the physical contact between AML and stromal cells but allowing soluble factors to cross. The regimen of co-culture and chemotherapy treatment was the same as standard co-culture. Polycarbonate membranes were used for mitoxantrone treatment after preliminary experiments indicated that mitoxantrone adsorbs to the polyester membrane.

Annexin V apoptosis assay

For the apoptosis assay, cells were labelled using fluorescein isothiocyanate (FITC) Annexin V (BD Biosciences, San Jose, CA, USA). The mOrange-positive stromal cells were excluded, and the percentage of apoptotic AML cells (FITC-positive and mOrange-negative) was determined from 10 000 mOrange-negative events acquired for each sample by LSRII (BD Biosciences). Data were acquired and analysed with Diva (BD Biosciences). The percentage of spontaneous apoptosis was subtracted from the drug-treated samples to yield the percentage of apoptosis attributed to drug treatment, as described (Redell et al, 2011).

Multiplex cytokine assay

Medium was conditioned by HS-5 and HS-27A parental and mOrange stromal cells for 48 h at 1 × 105/ml, and fresh RPMI medium with 10% FBS was used to define the background. Forty-two soluble factors in freshly harvested stroma-conditioned medium were quantified with the MIL-LIPLEX® MAP Human Cytokine/Chemokine Magnetic Bead Panel (Millipore, Billerica, MA, USA), using the Luminex 200™ (Luminex Corporation, Austin, TX, USA).

Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR)

AML cells were cultured alone, or co-cultured with HS-5 or HS-27A cells for 24 h (primary paediatric samples) or 48 h (cell lines) before mOrange-negative AML cells were sorted by fluorescence-activated cell sorting (FACS) (MoFlo, Beckman Coulter, Indianapolis, IN, USA) at Flow Cytometry Core at Texas Children’s Cancer and Hematology Centers. Total RNA from AML cell lines was extracted, and reverse transcribed into cDNA as described (Redell et al, 2011). Messenger RNA from primary samples was isolated and cDNA synthesized using μMACS mRNA isolation kit and cDNA synthesis module (Miltenyi Biotec, San Diego, CA, USA). The cDNA was used as the template in qRT-PCR (ABI Prism 7900HT; Applied Biosystems/Life Technologies). Parallel reactions were performed for GAPDH, which served as endogenous control for normalization. The fold change in mRNA levels for the co-culture conditions, relative to the culture alone conditions, was calculated by the ΔΔCt method (Ono et al, 2004).

Oligonucleotide microarray assay and data analysis

Human AML cells were cultured for 48 h under 3 conditions: alone, co-culture with HS-5 cells or co-culture with HS-27A cells. Cells were then sorted by FACS (MoFlo, Flow Cytometry Core at Texas Children’s Cancer and Hematology Centers) to isolate the mOrange-negative population. Three biological replicates were prepared for each culture condition, for two AML cell lines (NB-4 and THP-1), yielding a total of 18 samples. Total RNA was isolated and analysed for acceptable quantity and purity. Oligonucleotide microarray gene expression profiling was performed with the Affymetrix U133Plus 2 platform (Affymetrix, Santa Clara, CA, USA) by the Genomic and RNA Profiling Core at Baylor College of Medicine. The raw intensity data were log transformed, normalized and summarized by the Robust Multichip Average (RMA) algorithm (Irizarry et al, 2003). The invariant genes were filtered if <20% of expression value of a gene has at least a 1·5-fold change from the gene’s median value or >50% missing intensity values. Then, the normalized intensities of the resulting genes were analysed by Significance Analysis of Microarray (SAM) to perform class comparisons of alone v. co-culture with HS-5 or HS-27A cells (Tusher et al, 2001). Genes, denoted by specific probesets, with a proportion of false discovery (PFD) ≤0·1 at 90th percentile after 1000 permutations were considered to be significant. The data were analysed by BRB ArrayTools, developed by Dr. Richard Simon and the BRB-Array Tools Development Team (Simon et al, 2007). The functional analysis of the significantly altered genes was performed using Ingenuity Pathway Analysis (Ingenuity Systems, Redwood City, CA, USA).

Immunocytochemistry

As described above, AML cells were co-cultured with stromal cells or cultured alone, and sorted for the mOrange-negative population. AML cells were then cytospun onto glass slides, fixed, and permeabilized as described (Perlaky et al, 1992). Slides were stained with rabbit-anti-human CYR61 primary antibody (Abcam, Cambridge, MA, USA) and donkey-anti-rabbit Cy3-conjugated secondary antibody (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA). The cells were co-stained for myeloperoxidase (MPO) using mouse-anti-human MPO primary antibody (Abcam) and chicken-anti-mouse Alexa Fluor® 488 secondary antibody (Invitrogen/Life Technologies). In parallel, the primary antibodies were omitted for some slides, which showed no non-specific binding by secondary antibodies themselves. MPO is a diagnostic marker for AML, and was used to confirm that CYR61-expressing cells were AML cells. Cell nuclei were counterstained with Hoechst 33342. The images were captured by Nikon Eclipse TE300 Fluorescent Microscope (Nikon, Tokyo, Japan) and analysed with Roper Scientific Image software (Media Cybernetics, Inc., Rockville, USA) at the RTSS Core, or with Nikon A1R Confocal Microscope (Nikon) and analysed with NIS Elements AR software (Nikon) at the Integrated Microscopy Core at Baylor College of Medicine.

CYR61 neutralization assays

THP-1 cells, as well as stromal cells, were pre-incubated with 20 μg/ml rabbit-anti-human CYR61 neutralizing antibody or normal rabbit IgG for 1 h at 37°C before either co-culture with stromal cells or continuing culture alone for 24 h. Then cells were treated with etoposide (1·5 μM) or mitoxantrone (50 nM) for 24 h, with addition of CYR61 neutralizing antibody or normal rabbit IgG again, followed by Annexin V apoptosis assays.

CYR61 knockdown

Packaging plasmids (VSV-G, Tat1b, Hgpm2, Rev1b) and lentivirus expression pGIPZ plasmids with short hairpin RNA (shRNA) constructs targeting human CYR61, or non-silencing control (purchased from the Cell-Based Assay Screening Service, C-BASS, core facility at Baylor College of Medicine), were added to HEK293T cells along with Lipofectamine 2000 (Invitrogen/Life Technologies). Stromal cells were transduced with lentiviral supernatant from HEK293T producer cells 48 h and 72 h after plasmid packaging with addition of 4 μg/ml polybrene (Sigma-Aldrich). Transduced stromal cells were confirmed with >97% transduction efficiency with turboGFP and then selected with 0·5 μg/ml puromycin (Sigma-Aldrich) for 5 days.

Western blotting

AML and stromal cells were treated with Brefeldin A (BD Biosciences, 5 μl/ml culture medium) and monensin (BD Biosciences, 2·5 μl/ml culture medium) for 5 h in order to block secretion of CYR61. Preparation of whole cell protein lysate and Western blotting were performed as described (Redell et al, 2011). CYR61 was detected with rabbit-anti-human CYR61 neutralizing antibody (1:200 dilution) and Alexa Fluor® 680 goat-anti-rabbit antibody (1:5000 dilution, Invitrogen/Life Technologies) on an Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE, USA). β-actin (ACTB, Sigma-Aldrich) served as loading control.

Intracellular flow cytometry

AML cells were cultured alone, or co-cultured with control/CYR61-KD HS-27A stromal cells for 24 h (primary patient samples) or 48 h (cell lines), before they were harvested, fixed, permeabilized and intracellularly stained as described (Redell et al, 2011). Cells were stained with Alexa Fluor® 488 anti-pY348-SYK (BD Biosciences), Alexa Fluor® 647 anti-pY352-SYK (BD Biosciences), phycoerythrin (PE) anti-pY525/526-SYK (Cell Signaling Technology, Danvers, MA, USA), FITC anti-SYK (BD Biosciences), or isotype control and analysed by LSRII. The mOrange-positive stromal cells were excluded, and 10 000 mOrange-negative events were acquired for AML cells. Similarly, Kasumi-1 AML cells were serum starved for 1 h, then stimulated with 100 ng/ml granulocyte colony-stimulating factor (G-CSF), HS-5- or HS-27A- conditioned medium for 15 min, before they were harvested, fixed, permeabilized, and intracellularly stained with PE anti-pY705-STAT3 (BD Biosciences) or isotype control as described above. 10 000 events were acquired and analysed for Kasumi-1 cells by LSRII. Data were further analysed with FCS Express 4 (De Novo Software, Glendale, CA, USA).

Inhibition of p-SYK by R406

NB-4, THP-1 cells or primary AML patient samples were pre-incubated with 3 μM R406 or dimethyl sulfoxide (DMSO) for 30 min at 37°C before either co-culture with control/CYR61-KD HS-27A cells, or continuing culture alone, for 1 h (primary patient samples) or 24 h (cell lines). Then all cells were treated with 300 nM mitoxantrone for 24 h before Annexin V apoptosis assays.

Statistical analysis

Values are means ± standard error of the mean (SEM). Dose-response apoptosis curves were generated by 4-parameter logistic equation using GRAPHPAD PRISM 5·02 software (GraphPad Software Inc., La Jolla, CA, USA). Two-way analysis of variance was used to compare co-culture v. alone or CYR61 neutralization v. rabbit IgG under different conditions, followed by Bonferroni’s least significant difference post hoc test using GraphPad Prism. Unpaired, two-tailed, t-test was used to compare apoptosis rates for AML cells in co-culture with non-silencing control (NS Ctrl) stroma v. CYR61-KD stromal cells, treated with different chemotherapies. P < 0·05 was considered significant.

Results

HS-5 and HS-27A stromal cells have distinct profiles of secreted soluble factors

In order to be able to distinguish stromal cells from leukaemia cells after co-culture, we stably transduced HS-5 and HS-27A cells with mOrange. To compare HS-5 to HS-27A cells, and to confirm the similarity of mOrange-transduced to parental stromal cells, we first examined the profiles of soluble factors in stroma-conditioned medium by performing a multiplex cytokine assay for 42 secreted factors. HS-5 cells secreted high levels of over 20 factors, including G-CSF and interleukin-6 (IL6), while HS-27A cells secreted only a few, e.g., vascular endothelial growth factor (VEGFA 165), monocyte chemoattractant protein-1 (MCP1, or chemokine C-C motif ligand 2, CCL2) and IL8 (Fig 1A–D). This assay also confirmed that the soluble factor secretion profiles were similar between parental and mOrange stromal cells.

Fig 1.

Fig 1

Open in a new tab

HS-5 cells secreted many more factors than HS-27A cells. Both parental (pa) and mOrange (mO) stromal cells were cultured for 48 h, and 42 soluble factors were examined in the stroma-conditioned medium by multiplex cytokine assay. Detectable soluble factors are grouped by concentrations (A–D). N = 4–5.

Given that G-CSF and IL6 are known activators of STAT3 (Redell et al, 2011, 2013), and that both are highly abundant in HS-5-conditioned medium but not in HS-27A-conditioned medium (Fig 1C), we next compared the stroma-conditioned media in the ability to activate STAT3 in Kasumi-1 AML cells. We found that HS-5-conditioned medium, but not HS-27A-conditioned medium, induced STAT3 phosphorylation (Y705) in Kasumi-1 cells, similar to treatment with G-CSF alone (Fig S1A). Again, conditioned media from the mOrange-transduced and parental stromal cells were equally effective in activating STAT3 in Kasumi-1 cells (Fig S1B).

Both stromal cell lines protected AML cell lines and primary AML patient samples from chemotherapy-induced apoptosis

After validation of the mOrange-transduced stromal cells, we established an in vitro model of the bone marrow environment by co-culturing AML cells with mOrange-transduced stromal cells. We tested the ability of HS-5 and HS-27A cells to protect AML cells from apoptosis induced by exposure to etoposide, mitoxantrone or cytarabine, three chemotherapy drugs used to treat AML patients. Both stromal cell lines significantly protected NB-4 cells (Fig 2A1–3) and THP-1 cells (Fig 2B1–3) from apoptosis induced by all three chemotherapy drugs in a dose-dependent manner. Interestingly, the two stromal cell lines conferred equivalent levels of resistance to etoposide and mitoxantrone. For cytarabine, HS-5 co-culture was more protective than HS-27A co-culture at intermediate drug doses (Fig 2A2, B2; Annexin V positive rate of NB-4 treated with 3 μM cytarabine was 5·1 ± 0·6% for HS-5 co-culture vs 13·2 ± 2·0% for HS-27A co-culture, P < 0·05).

Fig 2.

Fig 2

Open in a new tab

Bone marrow stromal cells conferred resistance to chemotherapy-induced apoptosis in acute myeloid leukaemia (AML) cell lines and primary paediatric patient samples. NB-4 and THP-1 AML cell lines and primary patient samples were kept in suspension, or co-cultured with HS-5 or HS-27A stromal cells, for 24 h (cell lines) or 1 h (patient samples), followed by exposure to chemotherapy drugs for 24 h before Annexin V apoptosis assay. Dose-response apoptosis curves of NB-4 (A1-3), THP-1 (B1-3), a representative patient sample (C1-3) and the averaged results from five patient samples (D1-3) are depicted. N = 4–5. BM, bone marrow mononuclear cells.

*, P < 0·05, alone v. HS-5 or HS-27A co-culture; #, P < 0·05, HS-5 v. HS-27A co-culture.

In addition to AML cell lines, we tested the ability of both stromal cell lines to protect primary paediatric AML patient samples against these chemotherapy agents. We found that primary AML cells were also significantly and similarly protected from chemotherapy-induced apoptosis by co-culture with both stromal cell lines (Fig 2C, D). Both stromal cells also protected primary patient samples against spontaneous apoptosis at baseline (data not shown). These results support the assertion that the stroma-mediated chemotherapy resistance demonstrated in AML cell lines is similar to the resistance that occurs in patients.

Different stroma-dependent mechanisms contribute to resistance for different chemotherapies

To investigate the underlying mechanisms of stroma-induced chemotherapy resistance for AML cells, we performed Transwell co-culture experiments. The Transwell inserts provide a microporous membrane which allows soluble factors, but not cells, to cross, thus preventing physical contact between the AML cells on top of the insert and the stromal cells on the well surface, below the membrane. In contrast to co-culture with direct contact (Fig 2A1, B1), Transwell co-culture significantly abrogated the protection conferred by both stromal cell lines against etoposide, for both NB-4 cells and THP-1 cells (Fig 3A1, B1; Annexin V positive rate of NB-4 treated with 3 μM etoposide was 24·1 ± 3·7% for HS-5 Transwell co-culture vs 29·8 ± 5·3% for NB-4 cells alone, not significant). Further, incubation of NB-4 and THP-1 cells with stroma-conditioned medium before and during etoposide treatment had very little protective effect on apoptosis (Fig S2A, B). These findings indicated that the mechanism of stroma-induced protection from etoposide in AML cells mainly requires direct cell-cell contact.

Fig 3.

Fig 3

Open in a new tab

Cell-cell contact and soluble factors differentially contributed to resistance to different chemotherapies in acute myeloid leukaemia (AML) cells. NB-4 and THP-1 AML cells were cultured alone, or co-cultured in Transwell inserts physically separated from stromal cells, and treated with chemotherapy before Annexin V apoptosis assay. Stroma-induced protection of AML cells from etoposide (A1, B1), cytarabine (A2, B2) and mitoxantrone (A3, B3) was shown. N = 3–4. *, P < 0·05, alone v. HS-5, or HS-27A co-culture; +, P < 0 05, alone v. HS-5 co-culture; #, P < 0·05, HS-5 v. HS-27A co-culture.

For cytarabine treatment, the Transwell membrane reduced the protection from HS-5 cells, suggesting some contribution from contact-dependent pathways to HS-5-induced cytarabine resistance. However, both cell lines still provided significant protection, suggesting both stromal cell lines also secrete soluble factors that promote resistance to cytarabine (Fig 3A2, B2). For example, the Annexin V positive rate of NB-4 cells treated with 3 μM cytarabine was 9·0 ± 1·1% for HS-5 Transwell co-culture and 11·6 ± 2·0% for HS-27A Transwell co-culture, vs. 20·0 ± 4·0% for NB-4 cells alone (P < 0·05). A protective effect of soluble factors was further supported by the finding that both stroma-conditioned media significantly protected NB-4 and THP-1 cells from cytarabine (Fig S2C, D).

Stroma-induced resistance to mitoxantrone also was partially reduced by the Transwell inserts, with more reduction of HS-27A- than HS-5-induced resistance (Fig 3A3, B3). The Annexin V positive rate of NB-4 cells treated with 100 nM mitoxantrone was 30·7 ± 3·4% for HS-27A Transwell co-culture vs 22·2 ± 2·3% for HS-5 Transwell co-culture (P < 0·05). A substantial role for HS-5-derived soluble factors was confirmed by our finding that HS-5-conditioned medium significantly reduced mitoxantrone-induced apoptosis (Fig S2E, F; Annexin V positive rate of NB-4 cells treated with 300 nM mitoxantrone was 35·0 ± 1·1% for HS-5-conditioned medium vs 50·9 ± 1·0% for control medium, P < 0·001). In contrast, HS-27A-conditioned medium provided minimal protection against mitoxantrone. Taken together, our results indicated that resistance to mitoxantrone, and to a lesser extent cytarabine, can be conferred by soluble factors secreted by stromal cells, while direct contact with stromal cells confers resistance to all three chemotherapy agents.

Stroma-induced genes and proteins in AML cells

To search for candidate genes responsible for stroma-dependent chemoresistance in AML cells, we compared the gene expression profiles of NB-4 and THP-1 AML cell lines under 3 conditions: cultured alone, co-cultured with HS-5 cells, or co-cultured with HS-27A cells. After 48 h in culture, cells were flow-sorted to isolate the mOrange-negative population. RNA was extracted for gene expression profiling using the Affymetrix U133Plus2 platform. Forty genes were found to be significantly up-regulated in both NB-4 and THP-1 AML cells after co-culture with HS-5 cells (Table SII), while over 1000 genes were altered, either up- or down-regulated, in both AML cell lines after co-culture with HS-27A cells (Tables SIII, SIV). Additionally, using SAM analysis, we identified genes that were more robustly induced in the AML cells by one stromal cell line compared to the other, and therefore distinguish between the two stromal cell lines (PFD <0·1). The 29 differentially expressed probesets, representing 24 genes, are listed in Table SV. For example, CD52, RPS6 (ribosomal protein S6), and FGR (FGR proto-oncogene, Src family tyrosine kinase) were all more robustly upregulated in AML cells by HS-27A cells than by HS-5 cells. GREM1 [gremlin 1, cysteine knot superfamily, homolog (Xenopus laevis)], ABI3BP [ABI family, member 3 (NESH) binding protein], and SERPINB2 [serpin peptidase inhibitor, clade B (ovalbumin), member 2] were more upregulated by HS-5 than by HS-27A cells.

To further evaluate the functions of those differentially expressed stroma-induced genes with regard to drug resistance, functional ontology analysis of the significantly differentially expressed genes was performed using Ingenuity Pathway Analysis. In the Diseases and Disorders category, Cancer was the most significantly enriched. Over 30 genes induced by HS-5 cells and over 100 genes induced by HS-27A cells were related to cancer. In the Molecular and Cellular Function category, the stroma-induced gene lists were significantly enriched in functions including cellular growth, cell-cell signalling and cell death (Table SVI).

Of particular interest were 19 transcripts that were commonly increased in both AML cell lines after co-culture with both stromal cell lines (Fig 4A, Table SVII). Given that both stromal cells provided similar degrees of protection for AML cells against chemotherapy, those 19 genes represent likely candidates contributing to chemoresistance. In addition, several genes, including CYR61, CAV1 (Caveolin 1, caveolae protein, 22 kDa), CCL2, CTGF and AXL (AXL receptor tyrosine kinase) are associated with chemotherapy resistance in other malignancies (Ho et al, 2008; Hong et al, 2008; Qian et al, 2010; Yin et al, 2010; Lai et al, 2011). To verify stroma-upregulated transcripts, we cultured AML cells alone or co-cultured with either stromal cell line, flow sorted to isolate the mOrange-negative population, and extracted RNA for qRT-PCR. We confirmed the stroma-induced upregulation of CYR61 (Fig 4B), as well as CAV1, TM4SF1 (Transmembrane-4-L-six-family-1), CTGF, CCL2 and AXL (data not shown), in both AML cell lines and in primary samples.

Fig 4.

Fig 4

Open in a new tab

A subset of stroma-induced genes was commonly upregulated by both stromal cell lines in acute myeloid leukaemia (AML) cells. NB-4 (N) and THP-1 (T) cells were cultured alone, or co-cultured with stromal cells for 48 h, then flow sorted to isolate the mOrange-negative population, from which total RNA was extracted for gene expression profiling. (A) Nineteen genes, including CYR61, were found upregulated in both AML cell lines by both stromal cell lines. The levels of those genes (mean-centred, denoted by gene-specific probesets) are shown. Red/green denotes relatively high/low gene expression, respectively. (B) CYR61 mRNA level was determined by qRT-PCR in NB-4 (N = 4), THP-1 (N = 4), and two AML patient samples cultured alone, or co-cultured with stromal cells. *, P < 0·05, alone v. HS-5 or HS-27A co-culture. (C) CYR61 protein expression was determined by immunocytochemistry in THP-1 cells and primary patient samples cultured alone or co-cultured with HS-5 cells. Representative images of CYR61 (column 1), myeloperoxidase (MPO; AML marker, column 2) staining and merged images of CYR61, MPO, and Hoechst DNA staining for cell nuclei (column 3) are shown. The scale bars shown are 25 μm.

Next, we examined the protein expression for selected genes by immunocytochemistry. AML cells were co-cultured with stromal cells or cultured alone, sorted for the mOrange-negative population, cytospun onto glass slides, fixed, permeabilized and stained for candidate stroma-upregulated proteins. The cells were co-stained for MPO, a clinically used marker for AML, in order to confirm that the imaged cells were AML cells and not contaminating stromal cells. CYR61, TM4SF1 and CCL2 proteins were detected in AML cells cultured alone (data not shown). Of these, only CYR61 protein was clearly up-regulated in THP-1 cells and primary paediatric AML samples (Fig 4C) after co-culture with HS-5 stromal cells. Similar CYR61 upregulation was detected in THP-1 cells and primary AML cells co-cultured with HS-27A cells and NB-4 cells co-cultured with either stromal cells (data not shown). These results suggested that stromal cells may confer resistance to chemotherapy partly through the upregulation of CYR61.

Reducing CYR61 activity partially reversed the stroma-induced resistance to mitoxantrone in AML cells

As an ECM-associated signalling molecule, CYR61 promotes cell adhesion, the mitogenic effects of growth factors (Kireeva et al, 1996; Babic et al, 1998), and taxol resistance in breast cancer (Menendez et al, 2005), so we hypothesized that CYR61 contributed to stroma-induced chemotherapy resistance in AML. We used anti-CYR61 neutralizing antibody to block CYR61 activity in THP-1 cells cultured alone or co-cultured with stromal cells, and treated with chemotherapy. Neutralization of CYR61 did not alleviate stroma-induced resistance to etoposide in THP-1 cells (Fig 5A). However, blocking CYR61 activity significantly reduced the protection provided by HS-27A cells to mitoxantrone, with a similar trend toward increased apoptosis in THP-1 cells co-cultured with HS-5 cells or cultured alone (Fig 5B; Annexin V positive rate of THP-1 cells co-cultured with HS-27A cells and treated with 50 nM mitoxantrone was 30·9 ± 4·0% for CYR61 neutralizing antibody vs 12·6 ± 3·0% for rabbit IgG control, P < 0·05).

Fig 5.

Fig 5

Open in a new tab

Neutralizing CYR61 reduced stroma-induced mitoxantrone resistance in THP-1 cells. THP-1 and stromal cells were pre-incubated with anti-CYR61 neutralizing antibody or normal rabbit IgG (both at 20 μg/ml) prior to placing in co-culture. After co-culture for 24 h, chemotherapy agent was added, along with anti-CYR61 or control IgG, and apoptosis was measured after another 24 h. The contributions of CYR61 activity to the stroma-induced protection from (A) 1·5 μM etoposide and (B) 50 nM mitoxantrone are shown. N = 3. *, P < 0·05, anti-CYR61 antibody v. normal rabbit IgG.

Next, we wanted to confirm the pharmacological neutralization results by RNAi knockdown. Although CYR61 expression was increased in co-cultured AML cells, we found that stromal cells express much higher levels of CYR61 at baseline, and therefore stromal cells are likely to be the major source of CYR61 in the bone marrow (Fig S3). Therefore, we transduced both HS-5 and HS-27A stromal cells with lentiviral supernatant containing shRNA constructs targeting CYR61, then confirmed the reduction in expression by qRT-PCR (Fig 6A) and Western blotting (Fig 6B). Among the 4 different shRNA constructs targeting CYR61, stromal cells transduced with constructs 2 (C2) and 4 (C4) achieved the highest knockdown efficiency, compared to stroma transduced with non-silencing control (NS Ctrl). In regards to their function, the mitoxantrone resistance conferred by C4-transduced stromal cells was significantly reduced, compared to NS Ctrl-transduced stromal cells (Fig 6C, D). The Annexin V positive rate of NB-4 cells co-cultured with HS-27A cells and treated with 100 nM mitoxantrone was 12·2 ± 0·7% for co-culture with C4-transduced HS-27A vs. 7·2 ± 0·7% for co-culture with NS Ctrl-transduced HS-27A (P < 0·05). Similarly, the Annexin V positive rate of THP-1 cells was 10·8 ± 0·8% for co-culture with C4-transduced HS-27A v. 6·8 ± 1·1% for co-culture with NS Ctrl-transduced HS-27A (P < 0·05). However, the resistance to etoposide and cytarabine was not changed by CYR61 knockdown in HS-27A stromal cells (Fig 6E–H). The results of these CYR61 knockdown experiments are consistent with the CYR61 neutralization results, and support a role for CYR61 in mediating stroma-induced resistance to mitoxantrone in AML cells.

Fig 6.

Fig 6

Open in a new tab

CYR61 knockdown in stromal cells alleviated stroma-induced mitoxantrone resistance in NB-4 and THP-1 cells. HS-5 and HS-27A cells were transduced with shRNA constructs targeting CYR61. Knockdown efficiency was evaluated by (A) quantitative reverse transcription polymerase chain reaction and (B) Western blotting, respectively. Compared to non-silencing control (NS Ctrl), constructs 2 (C2) and C4 had the highest knockdown efficiency, and these two were used for further functional analyses. The contribution of stromal CYR61 to resistance to 100 nM mitoxantrone (C, D), 1·5 μM etoposide (E, F) or 3 μM cytarabine (G, H) was determined in AML cells co-cultured with C2- or C4-transduced stromal cells in comparison to those co-cultured with NS Ctrl stromal cells. N = 4. *, P < 0·05 as indicated.

CYR61 increases SYK expression and SYK activation

According to the microarray mRNA gene profiling, SYK was among the genes in AML cells that were upregulated by co-culture with HS-27A cells (Table SIII). Given that CYR61 is reported to bind to integrin β3 (Menendez et al, 2005; Su et al, 2010; Long et al, 2013), and integrin β3 signals through SYK in AML cells (Miller et al, 2013), we hypothesized that SYK may be downstream of CYR61 in mediating HS-27A-induced resistance to mitoxantrone. We first determined total SYK (t-SYK) expression in AML cells cultured alone or co-cultured with NS Ctrl or CYR61-KD HS-27A cells by flow cytometry. T-SYK was upregulated in NB-4 and THP-1 cells co-cultured with NS Ctrl HS-27A, but not in AML cells co-cultured with CYR61-KD HS-27A (Fig 7A, B). Next, we measured SYK activation by phospho-SYK (p-SYK) in AML cells, in the same culture conditions, by flow cytometry. As the phosphorylation of tyrosine 348 (Y348) and Y352 of SYK is important for its conformational activation (Gradler et al, 2013) and pY525/526 also contributes to its tyrosine kinase activity (Zhang et al, 2000), we investigated pY348, pY352 and pY525/526 SYK. Levels of pY348 were significantly increased in NB-4 and THP-1 cells co-cultured with NS Ctrl HS-27A, but not in cells co-cultured with CYR61-KD HS-27A (Fig 7C, D). The mean fluorescence intensity (MFI) of pY348 was 19·0 ± 4·3 for NB-4 co-cultured with NS Ctrl HS-27A vs. 10·1 ± 1·6 for NB-4 co-cultured with CYR61-KD HS-27A (P < 0·05), and 7·5 ± 1·0 for NB-4 alone, P < 0·05). Similarly, pY348 was significantly increased in primary AML samples after co-culture with NS Ctrl HS-27A cells, but not after co-culture with CYR61-KD cells (Fig 7E). In contrast, pY352 and pY525/526 did not change in NB-4 and THP-1 cells co-cultured with NS Ctrl HS-27A (Fig S4A–D). These results suggested that the increase in stroma-induced SYK activation was specific for pY348 and was mediated by the ECM ligand, CYR61.

Fig 7.

Fig 7

Open in a new tab

Stromal CYR61 upregulated total-SYK (t-SYK) and phospho-SYK (Y348) in acute myeloid leukaemia (AML) cells. T-SYK was determined by flow cytometry in NB-4 (A, N = 3) and THP-1 cells (B, N = 3), cultured alone or co-cultured with non-silencing control (NS Ctrl) HS-27A or CYR61-KD HS-27A cells for 48 h. P-SYK (Y348) was also determined in NB-4 (C, N = 6), THP-1 (D, N = 4) and paediatric AML patient samples (E, N = 9) in the similar culture conditions. The expression levels were quantified by mean fluorescence intensity (MFI). *, P < 0·05, **, P < 0·01, + NS Ctrl v. alone.

CYR61-induced SYK activation confers resistance to mitoxantrone

To determine the role of SYK in CYR61-mediated mitoxantrone resistance, we used the small molecule SYK inhibitor, R406 (Braselmann et al, 2006). To determine the optimal dose of R406 for inhibition of SYK activation (pY348), we pre-treated AML cells with different doses of R406, prior to co-culture with NS Ctrl HS-27A and measured p-SYK (Y348) in AML cells by flow cytometry. At 3 μM, R406 achieved significant inhibition of SYK activation (pY348) in AML cells co-cultured with NS Ctrl HS-27A (Fig 8A, B), but not in AML cells cultured alone or co-cultured with CYR61-KD HS-27A (data not shown), further supporting the notion that SYK activation via pY348 is mediated by stromal CYR61.

Fig 8.

Fig 8

Open in a new tab

Inhibition of p-SYK partially reversed HS-27A-induced mitoxantrone resistance. NB-4 (A, N = 3) and THP-1 cells (B, N = 4) were pre-incubated with SYK inhibitor R406 (30 nM, 300 nM, or 3 μM) or dimethyl sulfoxide (DMSO) vehicle control for 30 min, before co-culture with NS Ctrl HS-27A cells for 48 h. Then p-SYK (Y348) was determined in acute myeloid leukaemia (AML) cells by flow cytometry and quantified by MFI. NB-4 (C, N = 4), THP-1 cells (D, N = 3) and primary AML patient samples (E, N = 5) were pre-incubated with 3 μM R406 or DMSO for 30 min, then were either cultured alone, or co-cultured with NS Ctrl HS-27A or CYR61-KD HS-27A cells. All cells were treated with 300 nM mitoxantrone for 24 h before the Annexin V apoptosis assay. The contribution of SYK activation (Y348) to stromal CYR61-induced resistance to mitoxantrone was depicted. *, P < 0·05 as indicated.

Next, we pre-treated NB-4, THP-1 cells and primary AML patient samples with 3 μM R406, and then either cultured alone or co-cultured with NS Ctrl HS-27A or CYR61-KD HS-27A and treated with mitoxantrone for 24 h. R406 treatment of AML cells cultured alone did not increase apoptosis, indicating that this drug is not directly toxic to AML cells (Fig 8C, D). In contrast, R406 treatment of AML cells co-cultured with NS Ctrl HS-27A cells did result in increased mitoxantrone-induced apoptosis, suggesting that p-SYK (Y348) contributes to this stroma-mediated resistance pathway (Fig 8C, D; Annexin V positive rate of NB-4 cells co-cultured with NS Ctrl HS-27A cells and treated with 300 nM mitoxantrone was 29·6 ± 3·1% for R406 vs. 20·4 ± 1·8% for DMSO, P < 0·05). Importantly, R406 did not further reduce mitoxantrone resistance beyond that already achieved by CYR61 knockdown (Fig 8C, D). Likewise, in primary paediatric AML samples, inhibition of SYK activity dramatically reversed mitoxantrone resistance conferred by CYR61-expressing stromal cells (Fig 8E). The Annexin V positive rate for primary AML cells co-cultured with NS Ctrl HS-27A cells and treated with 300 nM mitoxantrone was 71·4 ± 9·3% for R406 vs. 14·8 ± 18·1% for DMSO (P < 0·05). The difference in apoptosis achieved by SYK inhibition in primary AML cells co-cultured with CYR61-KD stromal cells was not significant (Fig 8E). These results place SYK (Y348) downstream of CYR61 in mediating HS-27A-induced mitoxantrone resistance in AML cell lines as well as primary patient samples.

Discussion

In the current study, we used an in vitro model of the bone marrow microenvironment to investigate the mechanisms of stroma-mediated chemotherapy resistance in AML, and to identify targets for overcoming this resistance. We compared the effects of co-culture with two distinct stromal cell lines, HS-5 and HS-27A. Both provided similar degrees of protection for AML cells against three chemotherapy drugs used clinically to treat patients with AML. However, chemoresistance is a diverse and complex process, as evidenced by our findings that the two cell lines confer resistance by different mechanisms, and that resistance to different chemotherapy agents can be achieved via different routes.

HS-5 and HS-27A stromal cell lines are both derived from the fibroblast lineage, but they differ in morphology, gene expression patterns (Graf et al, 2002), and secreted factors (Roecklein & Torok-Storb, 1995). Using a comprehensive multiplex bead assay, we further defined the differences in soluble factor levels between HS-5 and HS-27A cells. Although HS-5 cells secrete many more soluble factors at higher levels than HS-27A cells, both stromal cells are able to maintain colony-forming cells at similar levels (Roecklein & Torok-Storb, 1995), and our results indicate that both are equally effective at promoting chemoresistance. It is somewhat surprising that a small number of genes were found to be significantly differentially expressed in the HS-5 co-culture relative to the HS-27A co-culture. The two AML cell lines may respond differently to soluble factors, due to different levels of receptors or intracellular mediators, for example, resulting in fewer genes meeting our selection criteria. This is supported by the fact that NB-4 and THP-1 cells showed different magnitudes of upregulation of SOCS3 induced by G-CSF treatment (Redell et al, 2011). Also, we observed a higher variability between the replicates of AML cells co-cultured with HS-5 cells relative to those co-cultured with HS-27A cells. This would also decrease the power of detecting significant genes in the HS-5 co-culture condition.

Our work showed that adhesion and soluble factors differentially contributed to stroma-induced resistance to different chemotherapies. For example, we demonstrated that resistance to etoposide requires direct contact with stromal cells, which is supported by previous work in pre-B acute lymphoblastic leukaemia (Mudry et al, 2000). In contrast, resistance to cytarabine and mitoxantrone has been reported to involve contact-dependent interactions and soluble factor-induced pathways (Mudry et al, 2000; Konopleva et al, 2002; Nefedova et al, 2003; Zeng et al, 2006), and our results implicate both processes for both drugs, as well. Etoposide induces DNA strand breaks by inhibition of topoisomerase II. It is most effective against cells in the G2 and S phases of the cell cycle. Mitoxantrone is also a potent inhibitor of topoisomerase II, but it can also intercalate into DNA and RNA. Mitoxantrone affects dividing and nondividing cells, so its actions are not cell cycle phase-specific. Cytarabine inhibits DNA polymerase and kills cells in the S phase and blocks the progression of cells from the G1 phase to the S phase. As etoposide and mitoxantrone both are topoisomerase II inhibitors, we believe that the mechanisms of action of the drugs do not fully explain the differences in stroma-dependent protective mechanisms. As for mitoxantrone and cytarabine, although they share similar protective effects induced by stroma (direct contact or soluble factor-induced), their mechanisms of action are very distinct. Therefore, the mechanisms of stroma-induced resistance described here are likely to be independent of the functions of the drugs.

While the contributions of adhesion and of soluble factors are well-studied, the downstream signalling pathways involved have not been completely described. A very recent report from Jacamo et al (2014), identified increased nuclear factor (NF)-κB signalling, induced in both stromal cells and leukaemia cells by VCAM1/VLA4 interaction, as a mediator of resistance to vincristine and to doxorubicin. Our study points to another pathway, in which the ECM ligand CYR61 increases SYK expression and activity in AML cells.

Our experiments led to the discovery of CYR61 as a potential mediator of stroma-induced resistance to mitoxantrone. CYR61 promotes breast cancer and prostate carcinoma growth (Sampath et al, 2002; Franzen et al, 2009), enhances pancreatic and gastric carcinogenesis (Babic et al, 1998; Haque et al, 2011), and is linked to intrinsic chemotherapy resistance in breast cancer and cervical cancer (Rho et al, 2009; Lai et al, 2011). This is the first time CYR61 has been shown to mediate stroma-induced chemotherapy resistance and the first time that CYR61 has been studied in any haematological malignancy. As an ECM-associated signalling molecule, CYR61 is thought to promote cell attachment and potentiate the effects of growth factors on DNA synthesis (Kireeva et al, 1996; Babic et al, 1998). Therefore, in the context of AML, CYR61 may enhance the proliferation and survival of AML cells by augmenting their adhesion to stromal cells, or by increasing the secretion or the activity of pro-survival growth factors; both mechanisms lead to activation of downstream survival signalling pathways. We found that CYR61 was more abundant in HS-27A cells than in HS-5 cells, and that CYR61 inhibition, by neutralizing antibody and by knockdown, was more effective for reversing HS-27A-induced mitoxantrone resistance, compared to HS-5. Therefore, the CYR61 pathway may be more important for resistance conferred by HS-27A cells than HS-5 cells. It is interesting to note that CYR61 neutralization did not alter etoposide resistance or cytarabine resistance. As etoposide and mitoxantrone are both topoisomerase II inhibitors, this difference in susceptibility to CYR61 signalling implies that this pathway does not induce resistance via altered topoisomerase activity.

Previous studies have shown that NF- κB (Lin et al, 2004), ERK1/2 (Menendez et al, 2005) and PI3K-AKT (Lai et al, 2011) are activated downstream of CYR61 and are associated with innate chemoresistance in breast cancer cells. In AML cells our gene expression profiling results led us to investigate SYK as a potential downstream mediator of CYR61-induced mitoxantrone resistance. SYK inhibition has been reported to inhibit cell proliferation, reduce cell viability, and increase cell apoptosis in AML (Hahn et al, 2009), and activated SYK promotes resistance to FLT3 inhibitors (Puissant et al, 2014). Thus, SYK may contribute to chemoresistance in AML. SYK includes two N-terminal Src Homology 2 (SH2) domains, one C-terminal kinase domain and two interdomain linkers. SYK is maintained in an autoinhibitory conformation in the basal condition. Upon appropriate stimulation, SYK undergoes autophosphorylation, or phosphorylation by Src family kinases, at Y348 and/or Y352 (Gradler et al, 2013), which disrupts the autoinhibitory conformation and causes its activation. Y348 and Y352 lie within the interdomain linker associated with the autoinhibitory conformation, and pY348 and pY352 are reported to be most important for SYK activation (Gradler et al, 2013). Y525/526 are located in the activation loop of the kinase domain, and pY525/526 may be less important for SYK activation than pY348 and pY352 (Gradler et al, 2013). Increased phosphorylation on Y348 therefore is expected to result in increased kinase activity. Our results confirmed a role for pY348 in SYK activation, and demonstrated for the first time a CYR61-SYK (pY348) pathway mediating stroma-dependent chemotherapy resistance. The confirmation of our results in patient-derived AML cells further validates the clinical relevance of this resistance mechanism. Understanding how the stromal microenvironment protects AML cells against chemotherapy, and how these processes may be overcome by inhibition of adhesion pathways, soluble factors, or their downstream targets, is likely to result in promising mechanism-oriented new therapies to reduce chemotherapy resistance and relapse, and thereby improve survival for patients with AML.

Supplementary Material

Suppl Fig 1

Fig S1. HS-5-, not HS-27A-, conditioned medium induced STAT3 phosphorylation in Kasumi-1 AML cells.

Suppl Fig 2

Fig S2. HS-5-, HS-27A- conditioned medium differentially protected AML cells against chemotherapy.

Suppl Fig 3

Fig S3. Stromal cells are the main source of CYR61 in bone marrow.

Suppl Fig 4

Fig S4. Phospho-SYK (Y352 or Y525/526) levels were not changed in AML cells by exposure to stromal CYR61.

Suppl Tables

Data S1. Supporting Methods.

Table SI. Clinical characteristics associated with primary paediatric AML patient samples used for this study

Table SII. HS-5-upregulated gene probesets in AML cells

Table SIII. HS-27A-upregulated gene probesets in AML cells

Table SIV. HS-27A-downregulated gene probesets in AML cells

Table SV. Gene probesets in both NB-4 and THP-1 AML cells that distinguish between HS-5 and HS-27A cells

Table SVI. Gene ontology and functional category analysis of stroma-induced genes

Table SVII. Genes in both NB-4 and THP-1 AML cells commonly upregulated by both HS-5 and HS-27A cells

Acknowledgments

This work was supported by the National Institutes of Health (NIH), USA (K08 HL085018; R01 CA 175026, M.S.R), Hyundai Hope on Wheels, USA (M.S.R.) and Gillson Longenbaugh Foundation, USA (M.S.R.). The authors thank Claudia Gerken, Tatiana Goltsova, Colby Navarro, Marcos Ruiz, Radhika Dandekar, Zaowen Chen and Dr. Donald Shaffer for advice and technical assistance, and Drs. David Tweardy and Edward Allan R. Sison for helpful discussions. Resources and expert technical assistance were provided by the Molecular Biology Core at Baylor College of Medicine (BCM), and the Flow Cytometry Core and the Research and Tissue Support Services (RTSS) Core at Texas Children’s Cancer and Hematology Centers. This project was supported in part by the Genomic and RNA Profiling Core at BCM with funding from the NIH NCI grant (P30CA125123) and the expert assistance of Dr. Lisa D. White, Ph.D. This project was supported by the Integrated Microscopy Core at BCM with funding from the NIH (HD007495, DK56338, and CA125123), the Dan L. Duncan Cancer Center, and the John S. Dunn Gulf Coast Consortium for Chemical Genomics. We would like to acknowledge the support of the C-BASS Shared Resource of the Dan L. Duncan Cancer Center at BCM (P30CA125123). The authors thank Dr. Stephen Gottschalk for kindly providing the HS-5 mOrange and HS-27A mOrange stromal cells, and for access to the Luminex 200. The authors thank Dr. Jianhua Yang for generously supplying the lentivirus packaging plasmids. Lastly, the authors are grateful to the RTSS facility and the COG AML Reference Lab for primary samples, and to the families who consented to the use of their cells for research.

Footnotes

Supporting Information

Additional Supporting Information may be found in the online version of this article.

Author contributions

Drs. XL and LP performed the experiments and acquired the data; Drs. MSR, XL, YY, LP and T-K Man contributed to the research design, data analysis and the manuscript writing.

Conflict of interest

The authors have no competing interests.

References

  1. Babic AM, Kireeva ML, Kolesnikova TV, Lau LF. CYR61, a product of a growth factor-inducible immediate early gene, promotes angiogenesis and tumor growth. Proceedings of the National Academy of Sciences of the United States of America. 1998;95:6355–6360. doi: 10.1073/pnas.95.11.6355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Braselmann S, Taylor V, Zhao H, Wang S, Sylvain C, Baluom M, Qu K, Herlaar E, Lau A, Young C, Wong BR, Lovell S, Sun T, Park G, Argade A, Jurcevic S, Pine P, Singh R, Grossbard EB, Payan DG, Masuda ES. R406, an orally available spleen tyrosine kinase inhibitor blocks fc receptor signaling and reduces immune complex-mediated inflammation. Journal of Pharmacology and Experimental Therapeutics. 2006;319:998–1008. doi: 10.1124/jpet.106.109058. [DOI] [PubMed] [Google Scholar]
  3. Franzen CA, Chen CC, Todorovic V, Juric V, Monzon RI, Lau LF. Matrix protein CCN1 is critical for prostate carcinoma cell proliferation and TRAIL-induced apoptosis. Molecular Cancer Research. 2009;7:1045–1055. doi: 10.1158/1541-7786.MCR-09-0017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Gamis AS, Alonzo TA, Meshinchi S, Sung L, Gerbing RB, Raimondi SC, Hirsch BA, Kahwash SB, Heerema-McKenney A, Winter L, Glick K, Davies SM, Byron P, Smith FO, Aplenc R. Gemtuzumab ozogamicin in children and adolescents with de novo acute myeloid leukemia improves event-free survival by reducing relapse risk: results from the randomized phase III Children’s Oncology Group Trial AAML0531. Journal of Clinical Oncology. 2014;32:3021–3032. doi: 10.1200/JCO.2014.55.3628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Gradler U, Schwarz D, Dresing V, Musil D, Bomke J, Frech M, Greiner H, Jakel S, Rysiok T, Muller-Pompalla D, Wegener A. Structural and biophysical characterization of the Syk activation switch. Journal of Molecular Biology. 2013;425:309–333. doi: 10.1016/j.jmb.2012.11.007. [DOI] [PubMed] [Google Scholar]
  6. Graf L, Iwata M, Torok-Storb B. Gene expression profiling of the functionally distinct human bone marrow stromal cell lines HS-5 and HS-27a. Blood. 2002;100:1509–1511. doi: 10.1182/blood-2002-03-0844. [DOI] [PubMed] [Google Scholar]
  7. Hahn CK, Berchuck JE, Ross KN, Kakoza RM, Clauser K, Schinzel AC, Ross L, Galinsky I, Davis TN, Silver SJ, Root DE, Stone RM, DeAngelo DJ, Carroll M, Hahn WC, Carr SA, Golub TR, Kung AL, Stegmaier K. Proteomic and genetic approaches identify Syk as an AML target. Cancer Cell. 2009;16:281–294. doi: 10.1016/j.ccr.2009.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Haque I, Mehta S, Majumder M, Dhar K, De A, McGregor D, Van Veldhuizen PJ, Banerjee SK, Banerjee S. Cyr61/CCN1 signaling is critical for epithelial-mesenchymal transition and stemness and promotes pancreatic carcinogenesis. Molecular Cancer. 2011;10:1–17. doi: 10.1186/1476-4598-10-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ho CC, Kuo SH, Huang PH, Huang HY, Yang CH, Yang PC. Caveolin-1 expression is significantly associated with drug resistance and poor prognosis in advanced non-small cell lung cancer patients treated with gemcitabine-based chemotherapy. Lung Cancer. 2008;59:105–110. doi: 10.1016/j.lungcan.2007.07.024. [DOI] [PubMed] [Google Scholar]
  10. Hong CC, Lay JD, Huang JS, Cheng AL, Tang JL, Lin MT, Lai GM, Chuang SE. Receptor tyrosine kinase AXL is induced by chemotherapy drugs and overexpression of AXL confers drug resistance in acute myeloid leukemia. Cancer Letters. 2008;268:314–324. doi: 10.1016/j.canlet.2008.04.017. [DOI] [PubMed] [Google Scholar]
  11. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Research. 2003;31:e15. doi: 10.1093/nar/gng015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Jacamo R, Chen Y, Wang Z, Ma W, Zhang M, Spaeth EL, Wang Y, Battula VL, Mak PY, Schallmoser K, Ruvolo P, Schober WD, Shpall EJ, Nguyen MH, Strunk D, Bueso-Ramos CE, Konoplev S, Davis RE, Konopleva M, Andreeff M. Reciprocal leukemia-stroma VCAM-1/VLA-4-dependent activation of NF-kappaB mediates chemoresistance. Blood. 2014;123:2691–2702. doi: 10.1182/blood-2013-06-511527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kireeva ML, Mo FE, Yang GP, Lau LF. Cyr61, a product of a growth factor-inducible immediate-early gene, promotes cell proliferation, migration, and adhesion. Molecular and Cellular Biology. 1996;16:1326–1334. doi: 10.1128/mcb.16.4.1326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Konopleva M, Konoplev S, Hu W, Zaritskey AY, Afanasiev BV, Andreeff M. Stromal cells prevent apoptosis of AML cells by up-regulation of anti-apoptotic proteins. Leukemia. 2002;16:1713–1724. doi: 10.1038/sj.leu.2402608. [DOI] [PubMed] [Google Scholar]
  15. Lai D, Ho KC, Hao Y, Yang X. Taxol resistance in breast cancer cells is mediated by the hippo pathway component TAZ and its downstream transcriptional targets Cyr61 and CTGF. Cancer Research. 2011;71:2728–2738. doi: 10.1158/0008-5472.CAN-10-2711. [DOI] [PubMed] [Google Scholar]
  16. Li ZW, Dalton WS. Tumor microenvironment and drug resistance in hematologic malignancies. Blood Reviews. 2006;20:333–342. doi: 10.1016/j.blre.2005.08.003. [DOI] [PubMed] [Google Scholar]
  17. Lin MT, Chang CC, Chen ST, Chang HL, Su JL, Chau YP, Kuo ML. Cyr61 expression confers resistance to apoptosis in breast cancer MCF-7 cells by a mechanism of NF-kappaB-dependent XIAP up-regulation. The Journal of Biological Chemistry. 2004;279:24015–24023. doi: 10.1074/jbc.M402305200. [DOI] [PubMed] [Google Scholar]
  18. Long QZ, Zhou M, Liu XG, Du YF, Fan JH, Li X, He DL. Interaction of CCN1 with alphavbeta3 integrin induces P-glycoprotein and confers vinblastine resistance in renal cell carcinoma cells. Anti-Cancer Drugs. 2013;24:810–817. doi: 10.1097/CAD.0b013e328363046d. [DOI] [PubMed] [Google Scholar]
  19. Matsunaga T, Takemoto N, Sato T, Takimoto R, Tanaka I, Fujimi A, Akiyama T, Kuroda H, Kawano Y, Kobune M, Kato J, Hirayama Y, Sakamaki S, Kohda K, Miyake K, Niitsu Y. Interaction between leukemic-cell VLA-4 and stromal fibronectin is a decisive factor for minimal residual disease of acute myelogenous leukemia. Nature Medicine. 2003;9:1158–1165. doi: 10.1038/nm909. [DOI] [PubMed] [Google Scholar]
  20. McMillin DW, Delmore J, Weisberg E, Negri JM, Geer DC, Klippel S, Mitsiades N, Schlossman RL, Munshi NC, Kung AL, Griffin JD, Richardson PG, Anderson KC, Mitsiades CS. Tumor cell-specific bioluminescence platform to identify stroma-induced changes to anticancer drug activity. Nature Medicine. 2010;16:483–489. doi: 10.1038/nm.2112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Meads MB, Hazlehurst LA, Dalton WS. The bone marrow microenvironment as a tumor sanctuary and contributor to drug resistance. Clinical Cancer Research. 2008;14:2519–2526. doi: 10.1158/1078-0432.CCR-07-2223. [DOI] [PubMed] [Google Scholar]
  22. Meads MB, Gatenby RA, Dalton WS. Environment-mediated drug resistance: a major contributor to minimal residual disease. Nature Reviews Cancer. 2009;9:665–674. doi: 10.1038/nrc2714. [DOI] [PubMed] [Google Scholar]
  23. Menendez JA, Vellon L, Mehmi I, Teng PK, Griggs DW, Lupu R. A novel CYR61-triggered ‘CYR61-alphavbeta3 integrin loop’ regulates breast cancer cell survival and chemosensitivity through activation of ERK1/ERK2 MAPK signaling pathway. Oncogene. 2005;24:761–779. doi: 10.1038/sj.onc.1208238. [DOI] [PubMed] [Google Scholar]
  24. Miller PG, Al-Shahrour F, Hartwell KA, Chu LP, Jaras M, Puram RV, Puissant A, Callahan KP, Ashton J, McConkey ME, Poveromo LP, Cowley GS, Kharas MG, Labelle M, Shterental S, Fujisaki J, Silberstein L, Alexe G, Al-Hajj MA, Shelton CA, Armstrong SA, Root DE, Scadden DT, Hynes RO, Mukherjee S, Stegmaier K, Jordan CT, Ebert BL. In Vivo RNAi screening identifies a leukemia-specific dependence on integrin beta 3 signaling. Cancer Cell. 2013;24:45–58. doi: 10.1016/j.ccr.2013.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mudry RE, Fortney JE, York T, Hall BM, Gibson LF. Stromal cells regulate survival of B-lineage leukemic cells during chemotherapy. Blood. 2000;96:1926–1932. [PubMed] [Google Scholar]
  26. Nefedova Y, Landowski TH, Dalton WS. Bone marrow stromal-derived soluble factors and direct cell contact contribute to de novo drug resistance of myeloma cells by distinct mechanisms. Leukemia. 2003;17:1175–1182. doi: 10.1038/sj.leu.2402924. [DOI] [PubMed] [Google Scholar]
  27. Ono M, Yu B, Hardison EG, Mastrangelo MA, Tweardy DJ. Increased susceptibility to liver injury after hemorrhagic shock in rats chronically fed ethanol: role of nuclear factor-kappa B, interleukin-6, and granulocyte colony-stimulating factor. Shock. 2004;21:519–525. doi: 10.1097/01.shk.0000126905.75237.07. [DOI] [PubMed] [Google Scholar]
  28. Perbal B. CCN proteins: multifunctional signalling regulators. Lancet. 2004;363:62–64. doi: 10.1016/S0140-6736(03)15172-0. [DOI] [PubMed] [Google Scholar]
  29. Perlaky L, Valdez BC, Busch RK, Larson RG, Jhiang SM, Zhang WW, Brattain M, Busch H. Increased growth of NIH/3T3 cells by transfection with human p120 complementary DNA and inhibition by a p120 anti-sense construct. Cancer Research. 1992;52:428–436. [PubMed] [Google Scholar]
  30. Puissant A, Fenouille N, Alexe G, Pikman Y, Bassil CF, Mehta S, Du J, Kazi JU, Luciano F, Ronnstrand L, Kung AL, Aster JC, Galinsky I, Stone RM, DeAngelo DJ, Hemann MT, Stegmaier K. SYK is a critical regulator of FLT3 in acute myeloid leukemia. Cancer Cell. 2014;25:226–242. doi: 10.1016/j.ccr.2014.01.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Qian DZ, Rademacher BL, Pittsenbarger J, Huang CY, Myrthue A, Higano CS, Garzotto M, Nelson PS, Beer TM. CCL2 is induced by chemotherapy and protects prostate cancer cells from docetaxel-induced cytotoxicity. Prostate. 2010;70:433–442. doi: 10.1002/pros.21077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Redell MS, Ruiz MJ, Alonzo TA, Gerbing RB, Tweardy DJ. Stat3 signaling in acute myeloid leukemia: ligand-dependent and -independent activation and induction of apoptosis by a novel small-molecule Stat3 inhibitor. Blood. 2011;117:5701–5709. doi: 10.1182/blood-2010-04-280123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Redell MS, Ruiz MJ, Gerbing RB, Alonzo TA, Lange BJ, Tweardy DJ, Meshinchi S. FACS analysis of Stat3/5 signaling reveals sensitivity to G-CSF and IL-6 as a significant prognostic factor in pediatric AML: a Children’s Oncology Group report. Blood. 2013;121:1083–1093. doi: 10.1182/blood-2012-04-421925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rho SB, Woo JS, Chun T, Park SY. Cysteine-rich 61 (CYR61) inhibits cisplatin-induced apoptosis in ovarian carcinoma cells. Biotechnology Letters. 2009;31:23–28. doi: 10.1007/s10529-008-9845-8. [DOI] [PubMed] [Google Scholar]
  35. Roecklein BA, Torok-Storb B. Functionally distinct human marrow stromal cell lines immortalized by transduction with the human papilloma virus E6/E7 genes. Blood. 1995;85:997–1005. [PubMed] [Google Scholar]
  36. Rubnitz JE, Inaba H, Dahl G, Ribeiro RC, Bowman WP, Taub J, Pounds S, Razzouk BI, Lacayo NJ, Cao X, Meshinchi S, Degar B, Airewele G, Raimondi SC, Onciu M, Coustan-Smith E, Downing JR, Leung W, Pui CH, Campana D. Minimal residual disease-directed therapy for childhood acute myeloid leukaemia: results of the AML02 multicentre trial. The Lancet Oncology. 2010;11:543–552. doi: 10.1016/S1470-2045(10)70090-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sampath D, Winneker RC, Zhang Z. The angiogenic factor Cyr61 is induced by the progestin R5020 and is necessary for mammary adenocarcinoma cell growth. Endocrine. 2002;18:147–159. doi: 10.1385/ENDO:18:2:147. [DOI] [PubMed] [Google Scholar]
  38. Shain KH, Yarde DN, Meads MB, Huang M, Jove R, Hazlehurst LA, Dalton WS. Beta1 integrin adhesion enhances IL-6-mediated STAT3 signaling in myeloma cells: implications for microenvironment influence on tumor survival and proliferation. Cancer Research. 2009;69:1009–1015. doi: 10.1158/0008-5472.CAN-08-2419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Simon R, Lam A, Li MC, Ngan M, Menenzes S, Zhao Y. Analysis of gene expression data using BRB-ArrayTools. Cancer Informatics. 2007;3:11–17. [PMC free article] [PubMed] [Google Scholar]
  40. Sison EAR, Brown P. The bone marrow microenvironment and leukemia: biology and therapeutic targeting. Expert Review of Hematology. 2011;4:271–283. doi: 10.1586/ehm.11.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Su JL, Chiou J, Tang CH, Zhao M, Tsai CH, Chen PS, Chang YW, Chien MH, Peng CY, Hsiao M, Kuo ML, Yen ML. CYR61 regulates BMP-2-dependent osteoblast differentiation through the {alpha}v {beta}3 integrin/integrin-linked kinase/ERK pathway. The Journal of Biological Chemistry. 2010;285:31325–31336. doi: 10.1074/jbc.M109.087122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proceedings of the National Academy of Sciences of the United States of America. 2001;98:5116–5121. doi: 10.1073/pnas.091062498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Yin D, Chen W, O’Kelly J, Lu D, Ham M, Doan NB, Xie D, Wang C, Vadgama J, Said JW, Black KL, Koeffler HP. Connective tissue growth factor associated with oncogenic activities and drug resistance in glioblastoma multiforme. International Journal of Cancer. 2010;127:2257–2267. doi: 10.1002/ijc.25257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zeng Z, Samudio IJ, Munsell M, An J, Huang Z, Estey E, Andreeff M, Konopleva M. Inhibition of CXCR4 with the novel RCP168 peptide overcomes stroma-mediated chemoresistance in chronic and acute leukemias. Molecular Cancer Therapeutics. 2006;5:3113–3121. doi: 10.1158/1535-7163.MCT-06-0228. [DOI] [PubMed] [Google Scholar]
  45. Zhang J, Billingsley ML, Kincaid RL, Siraganian RP. Phosphorylation of Syk activation loop tyrosines is essential for Syk function. An in vivo study using a specific anti-Syk activation loop phosphotyrosine antibody. The Journal of Biological Chemistry. 2000;275:35442–35447. doi: 10.1074/jbc.M004549200. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Suppl Fig 1

Fig S1. HS-5-, not HS-27A-, conditioned medium induced STAT3 phosphorylation in Kasumi-1 AML cells.

NIHMS869996-supplement-Suppl_Fig_1.tif (12.7MB, tif)
Suppl Fig 2

Fig S2. HS-5-, HS-27A- conditioned medium differentially protected AML cells against chemotherapy.

NIHMS869996-supplement-Suppl_Fig_2.tif (17.1MB, tif)
Suppl Fig 3

Fig S3. Stromal cells are the main source of CYR61 in bone marrow.

NIHMS869996-supplement-Suppl_Fig_3.tif (7.1MB, tif)
Suppl Fig 4

Fig S4. Phospho-SYK (Y352 or Y525/526) levels were not changed in AML cells by exposure to stromal CYR61.

NIHMS869996-supplement-Suppl_Fig_4.tif (21.6MB, tif)
Suppl Tables

Data S1. Supporting Methods.

Table SI. Clinical characteristics associated with primary paediatric AML patient samples used for this study

Table SII. HS-5-upregulated gene probesets in AML cells

Table SIII. HS-27A-upregulated gene probesets in AML cells

Table SIV. HS-27A-downregulated gene probesets in AML cells

Table SV. Gene probesets in both NB-4 and THP-1 AML cells that distinguish between HS-5 and HS-27A cells

Table SVI. Gene ontology and functional category analysis of stroma-induced genes

Table SVII. Genes in both NB-4 and THP-1 AML cells commonly upregulated by both HS-5 and HS-27A cells

NIHMS869996-supplement-Suppl_Tables.docx (203.7KB, docx)

RESOURCES