Genotoxic Stress Induces Senescence-Associated ADAM10-Dependent Release of NKG2D MIC Ligands in Multiple Myeloma Cells

Alessandra Zingoni, Francesca Cecere, Elisabetta Vulpis, Cinzia Fionda, Rosa Molfetta, Alessandra Soriani, Maria Teresa Petrucci, Maria Rosaria Ricciardi, Daniel Fuerst, Maria Giulia Amendola, Joannis Mytilineos, Cristina Cerboni, Rossella Paolini, Marco Cippitelli and Angela Santoni
J Immunol July 15, 2015, 195 (2) 736-748; DOI: https://doi.org/10.4049/jimmunol.1402643

Abstract

Genotoxic stress can promote antitumor NK cell responses by upregulating the surface expression of activating ligands on cancer cells. Moreover, a number of studies suggested a role for soluble NK group 2D ligands in the impairment of NK cell tumor recognition and killing. We investigated whether genotoxic stress could promote the release of NK group 2D ligands (MHC class I–related chain [MIC]A and MICB), as well as the molecular mechanisms underlying this event in human multiple myeloma (MM) cells. Our results show that genotoxic agents used in the therapy of MM (i.e., doxorubicin and melphalan) selectively affect the shedding of MIC molecules that are sensitive to proteolytic cleavage, whereas the release of the short MICA*008 allele, which is frequent in the white population, is not perturbed. In addition, we found that a disintegrin and metalloproteinase 10 expression is upregulated upon chemotherapeutic treatment both in patient-derived CD138+/CD38+ plasma cells and in several MM cell lines, and we demonstrate a crucial role for this sheddase in the proteolytic cleavage of MIC by means of silencing and pharmacological inhibition. Interestingly, the drug-induced upregulation of a disintegrin and metalloproteinase 10 on MM cells is associated with a senescent phenotype and requires generation of reactive oxygen species. Moreover, the combined use of chemotherapeutic drugs and metalloproteinase inhibitors enhances NK cell–mediated recognition of MM cells, preserving MIC molecules on the cell surface and suggesting that targeting of metalloproteinases in conjunction with chemotherapy could be exploited for NK cell–based immunotherapeutic approaches, thus contributing to avoid the escape of malignant cells from stress-elicited immune responses.

Introduction

Increasing evidence supports a pivotal role for innate immune effector cells, such as NK cells, in tumor surveillance. NK cells’ rapid response is related primarily to their ability to release cytolytic mediators, such as perforin and granzymes, or to express ligands able to trigger death receptors on target cells. NK cell activation is tightly regulated by a delicate balance between activating and inhibitory signals, with the latter being primarily transduced by receptors for MHC class I molecules (KIRs, CD94/NKG2A). Recognition of abnormal self on tumor cells triggers a number of non-MHC class I–restricted activating receptors, such as NK group 2D (NKG2D), DNAX accessory molecule-1 (CD226), and the natural cytotoxicity receptors (1).

NKG2D, a C-type lectin-like receptor expressed on NK cells, γδ T cells, CD8+ T cells, and some autoreactive or immunosuppressive CD4+ T cells (2), represents a major recognition receptor for the detection and elimination of transformed cells (3, 4). In humans, NKG2D binds to MHC class I–related chain (MIC)A, MICB, and UL16-binding proteins, whose expression is either null or low on normal tissues; however, it is induced by different types of stress, including tumoral transformation, viral and bacterial infections, and autoimmune diseases (5, 6).

NKG2D ligand (NKG2DL) expression is finely regulated by transcriptional, posttranscriptional, and posttranslational mechanisms, with the latter comprising shedding through the action of metalloproteinases or exosome release. Thus, in addition to being displayed on the cell surface, NKG2DLs can be shed or secreted from different cell types (7–9), and they appear as soluble forms in the sera of most cancer patients (10, 11). The release of soluble (s)MIC was suggested to be a major mechanism of tumor cell evasion from NKG2D-mediated immune surveillance. Indeed, sMICA/B can cause systemic downregulation of NKG2D surface expression, thereby impairing NKG2D-dependent lysis of tumor cells (11, 12). Concomitantly, the reduction in NKG2DL density on the tumor cell surface likely triggers less efficient killing.

Shedding of MIC molecules primarily depends on a proteolytic process involving multiple enzymes, including a disintegrin and metalloproteinase (ADAM)9, ADAM10, ADAM17, some members of the matrix metalloproteinases (MMP) family (MMP9 and MMP14), and the disulfide isomerase Erp5 (13). Another level of complexity is reflected by the fact that MICA displays a high degree of polymorphism leading to an allele-dependent shedding, as demonstrated for the allelic variant MICA*008 released in association with exosomes (14).

Multiple myeloma (MM) is a clonal B cell malignancy characterized by expansion of plasma cells (PCs) in the bone marrow. It is an incurable disease with a median survival of a few years, and its prognosis has been improved by the use of autologous hematopoietic stem cell transplantation and new immunochemotherapeutic approaches (15).

Among immune cells controlling MM progression, NK cells have long been considered key players, because they are expanded and activated during the early stages of disease and can recognize and kill the tumor cells (16). NKG2D was reported to play an important role in the NK cell–mediated killing of MM cells (17–19), and high levels of sMICA in sera of MM patients correlate with the disease stage, strongly suggesting sMICA as an independent prognostic factor in newly diagnosed MM patients (20, 21).

A novel strategy to improve NK cell–mediated antitumor activity may involve the upregulation of activating ligands on tumor cells. Enhancement of NK cell–mediated recognition of MM cells was reported by us and other investigators, showing increased surface expression of NKG2DLs on tumor cells following treatment with genotoxic agents (22) or with GSK3, HSP-90, and histone deacetylase inhibitors (23–25). Importantly, drug-induced expression of these ligands on MM cells was always associated with their ability to trigger increased NK cell degranulation in an NKG2D-dependent manner. Our previous findings indicate that upregulation of NKG2DL expression on MM cells in response to genotoxic drugs depends on DNA damage response activation and reactive oxygen species (ROS) signaling and is associated with the induction of a drug-induced senescent phenotype (22, 26).

In the current study, we investigated the mechanisms regulating the shedding of different allelic variants of MICA and MICB following treatment of MM cells with sublethal doses of genotoxic drugs: doxorubicin (DOX) and melphalan (MEL). To our knowledge, we first demonstrate that drug treatment can stimulate the shedding of MICA long alleles and MICB by a mechanism that is primarily dependent on ADAM10, whereas the release of the MICA*008 short allele is not perturbed. We also provide novel evidence that genotoxic agents result in upregulation of ADAM10 expression that is associated with a senescent phenotype and requires ROS generation. Finally, our data show that the combined use of chemotherapeutic drugs and metalloproteinase inhibitors strongly increases NK cell–mediated recognition of MM cells, preserving MIC molecules on the cell surface.

Materials and Methods

Abs and reagents

Anti-MICA (clone 159227, IgG2b), anti-MICB (clone 236511; IgG2b), anti-ADAM10 (clone 163003; IgG2b), and anti-ADAM17 (clone 111633; IgG1) were from R&D Systems (Minneapolis, MN). Anti-PDIA6 (Erp5) Ab (ab11432) was purchased from Abcam (Cambridge, U.K.); anti-Erp5 (sc-365260; IgG2b) was from Santa Cruz Biotechnology (Santa Cruz, CA). Control mouse IgG1 (clone MOPC-21) was from BioLegend (San Diego, CA); F(ab)2 fragments of allophycocyanin-conjugated goat anti-mouse ([GAM]-allophycocyanin or GAM-PE) IgG were from Jackson ImmunoResearch Laboratories (West Grove, PA). Other Abs used were anti-MICA (clone AMO1; IgG1; BamOmaB, Tübingen, Germany); anti–β-actin (clone AC-74; IgG2a; Sigma-Aldrich, St Louis, MO); anti–CD3/allophycocyanin-H7 (clone SK7), anti-CD56/PE (clone NCAM16.2), anti–CD16/PerCP-CY5.5 (clone 3G8), anti-CD107a/allophycocyanin (clone H4A3), anti-CD38/allophycocyanin (clone HIT2), anti-CD138/FITC, CD138/PerCP (clone MI15) (all from BD Biosciences); and Alexa Fluor 488 GAM IgG Ab and Alexa Fluor 594 cholera toxin subunit B conjugate (C22842) (Life Technologies, Gaithersburg, MD). ELISA kits for sMICA, sMICB, and syndecan (CD138) were from R&D Systems. The ADAM10 inhibitor GI254023X was kindly provided by Dr. A. Ludwig (Institute for Biochemistry, Christian-Albrechts-University Kiel, Kiel, Germany). Other reagents used were MMP inhibitor marimastat, paraformaldehyde, propidium iodide (PI), trypan blue, Polybrene (8 mg/ml) (hexadimethrine bromide), N-acetyl cysteine (NAC), and PI-PLC (all from Sigma-Aldrich).

Cell lines and clinical samples

The human MM cell lines SKO-007(J3), U266, RPMI 8226, ARP, ARK, OPM-2, and LP1 were provided by P. Trivedi (“Sapienza” University of Rome). The cell lines were maintained at 37°C and 5% CO2 in RPMI 1640 (Life Technologies, Gaithersburg, MD) supplemented with 10% FCS. All cell lines were mycoplasma free (EZ-PCR Mycoplasma test kit; Biological Industries). Bone marrow samples from patients with MM were managed at the Division of Hematology (“Sapienza” University of Rome). Informed consent, in accordance with the Declaration of Helsinki, was obtained from all patients, and approval was obtained from the Ethics Committee of “Sapienza” University of Rome. The bone marrow aspirates were processed, as previously described (22). In some experiments, myeloma cells were purified using anti-CD138 magnetic beads (Miltenyi Biotec, Auburn, CA). More than 95% of the purified cells expressed CD138 and CD38.

RNA isolation, RT-PCR, and real-time PCR

One microgram of total RNA, isolated by TRIzol reagent (Life Technologies, Grand Island, NY), was used for cDNA first-strand synthesis in a 25-μl reaction volume, according to the manufacturer’s protocol for Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI). Real-time PCR was performed using the ABI Prism 7900 Sequence Detection system (Applied Biosystems, Foster City, CA). cDNAs were amplified in triplicate with primers for ADAM10 (Hs00153853_m1), ADAM17 (Hs01041915_m1), and β-actin (Hs99999903_m1) conjugated with fluorochrome FAM (Applied Biosystems). The level of ADAM expression was measured using threshold cycle (Ct), which was obtained by subtracting the Ct value of the gene of interest from that of the housekeeping gene (β-actin). In the current study, we used the Ct of the untreated sample as the calibrator. The fold change was calculated as 2−ΔΔCt, where ΔΔCt is the difference between the Ct of the sample and the Ct of the calibrator (according to the formula, the value of the calibrator in each run is 1).

Plasmids

The pMSCV retroviral vector (Clontech Laboratories, Mountain View, CA) was used to clone the cDNA encoding the MICA*008, MICB, and MICA*019 sequences. MICA*008 and MICB cDNA were amplified from the SKO-007(J3) cell line, whereas MICA*019 cDNA was provided by Lewis L. Lanier (University of California, San Francisco). For knocking down ADAM10 and ADAM17, we used pLKO.1-sh-ADAM10 (TRCN0000006672) and pLKO.1-sh-ADAM17 (TRCN0000052168) lentivirus vectors and the control vector pLKO nontarget short hairpin RNA (shRNA) (all from Sigma-Aldrich).

Virus production and in vitro transduction

For retrovirus production, the Phoenix retrovirus packaging cell line HEK293 was transfected with viral DNA (pMSCV, pMSCV/MICA*008, pMSCV/MICB, and pMSCV/MICA*019) and the packaging vectors (pCMVgag-pol and pMD2.G) using Lipofectamine Plus (Invitrogen, San Diego, CA). With regard to lentivirus production, Phoenix cells were transfected with pLKO-sh-ADAM10, pLKO-sh-ADAM17, or the control vector pLKO plus psPAX2 and pMD2.G packaging plasmids. After 48 h, virus-containing supernatants were harvested, filtered, and used immediately for infection as follows: 2 ml viral supernatant containing Polybrene (8 mg/ml) was used to infect 1 × 106 SKO-007(J3) or LP1 cells for 2 h. Two infection cycles were performed.

Drug treatment

Cells were cultured in six-well tissue culture plates for different times (24–72 h) at a concentration of 3 × 105 cells/ml with different drugs. In some experiments, marimastat (10 μM) or GI254023X (10 μM) was added to the cell culture. A dose-response curve was constructed for both metalloproteinase inhibitors to identify the doses that did not affect cell viability. DOX and MEL doses used to treat the different cell lines were determined as previously described (22) and corresponded to sublethal doses 10 times lower than the IC50 values: DOX: ARK (0.06 μM), LP-1 (0.06 μM), OPM-2 (0.08 μM), RPMI-8226 (0.05 μM), SKO-007(J3) (0.05 μM), and U266 (0.1 μM); MEL: ARK (7 μM), LP-1 (21.5 μM), OPM-2 (1.6 μM), RPMI-8226 (1.5 μM), SKO-007(J3) (22 μM), and U266 (15 μM). In some experiments, cells were incubated with 10 mM NAC for 1 h before the drug treatment, as previously described (26). Patient-derived PCs were incubated with MEL (10 μM) or DOX (0.05 μM) for 48–72 h at 37°C and 5% CO2.

Immunofluorescence, FACS, and microscopic analysis

The expression of the NKG2DLs and ADAM10 and ADAM17 on MM cells was analyzed by immunofluorescence staining using unconjugated mAbs, followed by secondary GAM-allophycocyanin or GAM-PE Abs. In some experiments, cells were stained with PI (1 μg/μl) to assess cell viability. The analysis of ligand and ADAM expression on patient-derived PCs was performed by gating on the CD38+CD138+ PC population. Samples were analyzed using a FACSCanto (BD Biosciences, San Jose, CA). For microscopy experiments, samples were prepared and visualized using a fluorescence microscope equipped with the ApoTome system, as previously described (27).

Degranulation assay

As the source of effector cells, we used PBMCs isolated from healthy donors by Lymphoprep (Nycomed, Oslo, Norway) gradient centrifugation and then cocultured for 10 d with the irradiated (30 Gy) EBV-transformed B cell line RPMI 8866 at 37°C in a humidified 5% CO2 atmosphere, as previously described (28). On day 10, the cell population was routinely >95% CD56+CD16+CD3+, as assessed by immunofluorescence and flow cytometry analysis. NK cells were activated overnight with 200 U/ml human rIL-2 (R&D Systems). Drug-treated SKO-007(J3) cells were incubated with activated NK cells in a U-bottom 96-well tissue culture plate in complete medium at 37°C and 5% CO2 for 2 h. Thereafter, cells were washed with PBS/2% FCS and incubated with the lysosomal marker CD107a/allophycocyanin for 45 min at 4°C. In some experiments, before the assay, target cells were incubated with anti-MICA (clone 159227; IgG2b; R&D Systems) and anti-MICB (BamOmaB clone AUMO1; mouse IgG2a) at 2 μg/106 cells for 20 min at room temperature. Cells were washed with complete medium and used as target in the degranulation assay. Blocking of the NKG2D receptor was performed as previously described (29).

When patient-derived PCs were used as targets, as the source of effector cells, we used autologous bone marrow purified NK cells or autologous bone marrow CD138− cells cultured for 2 d in complete medium supplemented with 200 U/ml IL-2. Drug-treated patient-derived PCs were incubated with effector cells in complete medium at 37°C and 5% CO2 for 2 h. Thereafter, cells were washed with PBS/2% FCS and stained with CD107a/allophycocyanin and anti–CD3/allophycocyanin-H7, anti-CD56/PE, and anti-CD16/PerCP to gate the CD3−CD56+CD16+ NK population. Analyses were performed using a FACSCanto (BD Biosciences).

Western blot analysis

For Western blot analyses, SKO-007(J3) or LP1 cells were pelleted, washed once with cold PBS, resuspended in lysis buffer (1% Triton X-100 [v/v], 50 mM Tris HCl [pH 7.6], 1 mM EGTA, 150 mM NaCl, 10 mM NaF, complete protease inhibitor mixture [Sigma-Aldrich]), and incubated for 20 min on ice. The lysate was centrifuged at 14,000 × g for 15 min at 4°C, and the supernatant was collected as whole-cell extract. Protein concentration was determined with the Bio-Rad Protein Assay. A total of 30–50 μg cell extract was run on 8% denaturing SDS-polyacrylamide gels. Proteins were electroblotted onto nitrocellulose membranes (Schleicher & Schuell, Keene, NJ) and blocked in 5% milk in TBST buffer. Immunoreactive bands were visualized on nitrocellulose membranes, using HRP-coupled goat anti-rabbit or GAM Igs and an ECL detection system (GE Healthcare Amersham), following the manufacturer’s instructions.

ADAM10 catalytic activity

ADAM10 catalytic activity was evaluated using a SensoLyte 520 ADAM10 Activity Assay Kit *Fluorimetric* (cat. #72226; AnaSpec). SKO-007(J3) cells were treated with DOX for 72 h. Cells were harvested, and 100 μg total cell lysate, prepared according to the manufacturer’s instructions, was incubated with 5-FAM/QXL 520 ADAM10 substrate for 30 or 60 min. Fluorescence signal was measured at 490/520 nm. Fluorescence readings are expressed in relative fluorescence units.

MICA gene typing

For the genotyping of MICA, genomic DNA, derived from patients’ PBMCs or from different MM cell lines, was isolated from 1 × 106 cells using the Genomic DNA purification kit, according to the manufacturer’s instructions (Bioline, London, U.K.). Sequence-based typing of MICA was performed as described (30).

Analysis and isolation of senescent cells

We performed a senescence-associated β-galactosidase (βGal) assay, using the fluorogenic substrate C12FDG (Invitrogen), to measure βGal activity by flow cytometry, as previously described (26). Surface staining of ADAM10 was performed on SKO-007(J3) cells or on ex vivo primary PCs, as previously described, and the C12-fluorescein signal was measured on the FL-1 detector. In some experiments, 72-h DOX-treated SKO-007(J3) cells were incubated with C12FDG, and βGalhigh and βGallow MM cells were sorted using a FACSAria (BD Biosciences).

Statistics

Statistical analysis was performed with the Student paired test. Error bars represent SD or, where indicated, SEM.

Results

Drug treatment of MM cells selectively affects the release of MIC molecules sensitive to proteolytic cleavage

To investigate whether genotoxic stress results in MICA and MICB release from MM cells, we initially determined the expression levels and the release of sMICB and sMICA together with MICA genotype in different MM cell lines (Table I). Notably, all MM cell lines expressed the MICA*008 short allele that is present more frequently in nearly all populations. Despite the fact that MICA was detected on the cell surface (data not shown) and in cell lysates (as evaluated by FACS and ELISA, respectively), it was not detectable in the supernatants of SKO-007(J3), U266, ARP-1, and RPMI cells. Instead, MICB was released from SKO-007(J3) and U266 cell lines (Table I).

View this table:
Table I. Expression levels and shedding profiles of endogenous MICA and MICB determined in different MM cell lines

We then examined whether the release of MICB in the supernatants of SKO-007(J3) and U266 cell lines could be modulated by treatment with DOX and MEL at doses not affecting cell viability (22). Interestingly, drug treatment strongly promoted the release of sMICB at 48 h, which increased further at 72 h (Fig. 1A, 1B, left panels). Because these treatments block cell proliferation, data also were normalized taking into account cell number; we noted an even greater difference between untreated and drug-treated cells (Fig. 1A, 1B, right panels). When the relative shedding of MICB was analyzed, a significant increase was observed in drug-treated cells compared with untreated cells, indicating that increased MICB levels in the supernatant were attributable to drug-induced MICB expression (data not shown) (22) and were dependent on stimulation of the shedding process (Fig. 1C, 1D). In contrast, we did not detect sMICA after drug treatment (data not shown), despite the enhancement of its cell surface expression (22) (data not shown).

FIGURE 1.
FIGURE 1.

Drug treatment of MM cells selectively affects the shedding process of MIC molecules sensitive to proteolytic cleavage. SKO-007(J3) (A) and U266 (B) cells were plated, at 3 × 105 cells/ml, in the presence of suboptimal doses of DOX and MEL. After 24, 48, and 72 h of incubation, cells were harvested, and supernatants were collected and analyzed for the presence of sMICB using specific ELISA. The mean of three independent experiments is shown. Relative shedding of MICB in SKO-007(J3) (C) and U266 (D) cells. Relative shedding, evaluated after 72 h, represents the ratio of the amount of MICB (ng/ml) detected in conditioned medium/cellular lysates (ng/ml). The mean of four independent experiments is shown. (E) LP1 MM cells transduced with MICA*008, MICA*019, or MICB were treated as described in (A). After 48 h, cells were washed and plated at 0.5 × 106 cells/ml in fresh medium and incubated for an additional 18 h. Supernatants were collected and analyzed for the presence of sMICA or sMICB by specific ELISA. Mean of four independent experiments ±SEM is shown. *p < 0.05, **p < 0.01. ns, not significant.

The notion that different MIC alleles exhibit distinct molecular structures and different shedding modalities (14, 31) prompted us to examine whether the chemotherapeutic agents could differentially affect the release of distinct MICA alleles. To this end, the MM cell line LP1, which does not express endogenous MICA and MICB, was stably transduced with cDNA encoding MICA*008 (carrying a truncated cytoplasmic tail and released by exosomes or through exocytosis), MICA*019 (which represents the prototype of the long form of MICA alleles), or MICB and exposed to DOX treatment. Transfection resulted in ligand expression, as evaluated by immunofluorescence, FACS analysis, fluorescence microscopy, and Western blot, and, importantly, transduced MIC molecules were released in the supernatants (Supplemental Fig. 1). Interestingly, DOX treatment upregulated only the shedding of the long MICA*019 allele, whereas it did not perturb the release of MICA*008 (Fig. 1E). As expected, MICB shedding also was strongly stimulated, confirming our observations for SKO-007(J3) and U266 cells.

Altogether, these data demonstrate that drug treatment of MM cells stimulates the shedding of MICA and MICB and selectively affects the release of MIC molecules that are sensitive to metalloproteinase cleavage (i.e., MICA*019 and MICB).

Inhibition of MICB shedding on drug-treated cells by treatment with metalloproteinase inhibitors increases their susceptibility to NK cell–mediated killing

To investigate whether drug-induced MICB release involves metalloproteinase activity (32, 33), we evaluated MICB levels in the supernatants of DOX- and MEL-treated SKO-007(J3) cells in the presence of the metalloproteinase inhibitor marimastat. As shown in Fig. 2A, marimastat inhibited the amount of sMICB, in a dose-dependent manner, on both untreated and drug-treated cells, and this inhibition was accompanied by a concomitant increase in cell surface MICB (Fig. 2B, 2C). In accordance with the evidence that the MICA*008 allele is resistant to proteolytic cleavage (14), marimastat treatment did not result in a significant upregulation of cell surface MICA expression (Fig. 2B, 2C). We explored whether the differential behavior of MICA*008 and MICB could be affected by the association mode to the plasma membrane. Because one of the modifications acquired by many NKG2DLs is the replacement of a short transmembrane region with a glycosylphosphatidylinositol (GPI) moiety (13), we investigated whether MIC molecules could be sensitive to PI-PLC, an enzyme that specifically cleaves the GPI anchor. We found that PI-PLC treatment of SKO-007(J3) cells affected MICA*008, but not MICB, cell surface expression, thus indicating that MICA*008 molecules can be anchored to the cell membrane through GPI (Supplemental Fig. 2). To investigate the functional consequences of marimastat-induced expression of MICB on drug-treated SKO-007(J3) MM cells, we evaluated their ability to trigger NK cell degranulation. Upregulation of NKG2DLs on MM cells by the combination of the metalloproteinase inhibitor and genotoxic agents was always verified before degranulation assays (data not shown). As shown in Fig. 2D and 2E, in accordance with the marimastat-induced enhancement of cell surface levels of MICB, but not MICA (Fig. 2B), NK cells contacting drug-treated MM cells exhibited higher expression of CD107a, a lysosomal marker that correlates with NK cell cytotoxicity; this expression was further enhanced upon marimastat target treatment. To directly prove the contribution of MICB, we performed degranulation experiments using a neutralizing Ab against MICB, alone or in combination with anti-MICA. Our results show that anti-MICB alone moderately, but significantly, reduced NK cell degranulation in untreated and drug-treated target cells, indicating that stabilization of this ligand on the cell surface renders tumor cells more susceptible to NK cell lysis. In addition, the combined usage of anti-MICA and anti-MICB Abs further reduced NK cell degranulation, indicating that the NKG2D-dependent lysis of the SKO-007(J3) cell line is mediated primarily by these ligands (Fig. 2F). In line with these observations, UL16-binding protein molecules, expressed at low levels on SKO-007(J3) cells, are not upregulated by drug treatment, thus excluding their contribution in the NKG2D-dependent killing of drug-treated MM cells (22) (data not shown). Collectively, these data show that chemotherapeutic treatment of SKO-007(J3) cells, while enhancing the surface expression of both NKG2DLs, promotes MICB, but not MICA*008, shedding by enhancing metalloproteinase activity. Inhibition of MICB shedding by marimastat in drug-treated cells results in increased susceptibility of MM cells to NK cell lysis by preserving cell surface MICB expression.

FIGURE 2.
FIGURE 2.

Inhibition of MICB shedding on drug-treated cells increases their susceptibility to NK cell–mediated killing. (A) SKO-007(J3) cells were plated at 3 × 105 cells/ml and treated with DOX and MEL, as described in Fig. 1. After 48 h, cells were washed and plated at 0.5 × 106 cells/ml in fresh medium. Different doses of the broad metalloproteinases inhibitor marimastat were added and left for an additional 18 h. Supernatants were collected and analyzed for the presence of sMICB by specific ELISA. A representative experiment of three is shown. (B) SKO-007(J3) cells were treated as described above in the presence of 10 μM of marimastat. Cell surface expression of MICA and MICB was evaluated by immunofluorescence and cytofluorimetric analysis by gating on PI− cells. Shaded line graphs represent the isotype-control Ig. A representative experiment of four is shown. (C) Cells were treated as described in (B). The mean of four independent experiments is shown. Values represent the mean fluorescence intensity (MFI) of MICA or MICB subtracted from the MFI value of the isotype-control Ig. (D) SKO-007(J3) cells, treated as described in (B), were used as targets, and long-term cultured NK cells were used as effectors in a standard CD107a degranulation assay using an E:T ratio of 2.5:1. NK cells were distinguished from target cells by gating on forward scatter and side scatter. Numbers indicate the percentage of CD107a+ cells. A representative experiment of four is shown. (E) Data are expressed as fold increase in CD107a values (%) obtained on NK cells cocultured with drugs and inhibitor-treated MM cells divided by CD107a values (%) of NK cells cocultured with untreated MM cells. The mean of four independent experiments is shown. (F) Degranulation assay was performed as described in (D). Before the assay, target cells were incubated with cIg, anti-MICB, or anti-MICA plus anti-MICB Abs, as described in Materials and Methods. Anti-NKG2D mAb was used instead to block NKG2D receptor on effector cells. The mean of four independent experiments ±SEM is shown. *p < 0.05, **p < 0.01, ***p < 0.001. ns, not significant.

The combined use of genotoxic drugs and metalloproteinase inhibitors on patient-derived malignant PCs enhances autologous NK cell degranulation

We next extended the analysis of the effects on NK cell degranulation by the combined use of metalloproteinase inhibitors and MEL in patient-derived PCs that express both MICA long alleles and MICB (Fig. 3A, 3B). To this end, highly purified patient-derived PCs were treated with MEL and marimastat for 48 h and then allowed to interact with IL-2–treated autologous bone marrow–derived NK cells to assess their ability to trigger NK cell degranulation. The combined treatment with MEL and marimastat enhanced the expression of MICA long alleles and MICB on patient PCs (Fig. 3A). We found that IL-2–treated bone marrow–derived autologous NK cells contacting CD138+ drug-treated cells exhibited higher expression of CD107a, and this expression was further enhanced upon marimastat target treatment (Fig. 3B, 3C). Similar results were obtained using CD138− bone marrow cells (Fig. 3B) or highly purified CD3−CD16+CD56+ NK cells (Fig. 3C).

FIGURE 3.
FIGURE 3.

The combined use of chemotherapeutic drugs and metalloproteinase inhibitors on patient-derived malignant PCs enhances autologous NK cell degranulation. (A) Total cells isolated from the bone marrow of MM patients were purified with CD138+ immunomagnetic beads and cultured in the presence of IL-3 (20 ng/ml), IL-6 (2 ng/ml), MEL (10 μM), marimastat (10 μM), or MEL+marimastat for 48 h, and expression of MICA and MICB on PCs was evaluated by immunofluorescence and FACS analysis. Results relative to two patients (P7 and P13) are shown. Values represent the mean fluorescence intensity (MFI) of MICA and MICB subtracted from the MFI value of the isotype-control Ig. (B) Highly purified PCs were treated as described in (A) and used as targets in a standard degranulation assay. IL-2–treated bone marrow CD138− cells were used as effectors. The assay was performed at an E:T ratio of 2.5:1. After 2 h at 37°C, cells were stained with anti-CD56, anti-CD3, anti-CD16, and anti-CD107a mAbs. Cell surface expression of CD107a was analyzed on CD56+CD16+CD3− NK cells. (C) Highly purified PCs were treated as described in (A) and used as targets in a standard degranulation assay. As effectors, autologous IL-2–treated highly purified bone marrow–derived NK cells, as evaluated by CD45+CD56+CD16+CD3− expression, were used. The assay was performed at an E:T ratio of 2.5:1. After 2 h at 37°C, cells were stained with anti-CD56, anti-CD16, and anti-CD107a mAbs. Cell surface expression of CD107a was analyzed on CD56+CD16+ NK cells. MICA gene typing of patients P39, P38, and P42 was *006/009:01, *008:1/018:01, and *007:01/008:01, respectively. Patient characteristics are shown in Supplemental Fig. 4C.

Overall, our findings indicate that the increased expression of MIC molecular species sensitive to proteolytic cleavage on MM cells results in enhanced NK cell recognition and killing and further potentiates the immunostimulatory effects exerted by genotoxic agents.

ADAM10 is involved in drug-induced MICB release from MM cells

The relative roles for ADAM10 and ADAM17 metalloproteinases on MIC shedding were primarily investigated in steady-state conditions in a variety of tumor cell lines or transfected cells (9, 32–34). To identify the ADAM enzymes responsible for the drug-induced MICB release, we examined whether the inhibition of ADAM10 and ADAM17 expression on drug-treated MM cells interfered with MICB shedding. To this end, ADAM10 and ADAM17 were knocked down in SKO-007(J3) cells using lentiviral vectors containing ADAM10 or ADAM17 shRNA. As shown in Fig. 4A–C, ADAM10 and ADAM17 expression were clearly suppressed and ∼70% reduction was observed at the mRNA and protein levels using real-time PCR, Western blot, immunofluorescence, and FACS analysis. Interestingly, ADAM10, but not ADAM17, knockdown led to a dramatic inhibition of both constitutive and drug-induced MICB shedding, as observed in untreated and DOX-treated MM cells (Fig. 4D). Of note, CD138, which is highly expressed and released as a soluble molecule from MM cells, including SKO-007(J3) cells (35), was reduced in ADAM17-silenced cells, making evident the specific effect of ADAM10-silencing on MICB shedding (Fig. 4E). The involvement of ADAM10 was further confirmed using the specific ADAM10 inhibitor GI254023X (36), which affected MICB release in a dose-dependent manner in both untreated and drug-treated MM cells (Fig. 4F). In addition, GI254023X inhibited MICA*019 release from LP1-transduced cells in response to drug treatment, revealing a role for this sheddase in drug-induced MICA shedding as well (data not shown).

FIGURE 4.
FIGURE 4.

ADAM10 strongly contributes to drug-induced MICB release. (A–C) SKO-007(J3) cells were infected with the pLKO lentiviral vector containing shRNA for silencing ADAM10 (shADAM10) or ADAM17 (shADAM17) or with a scrambled sequence. After 72 h, real-time PCR, immunofluorescence, FACS analysis, and Western blot were performed to evaluate the efficiency of the ADAM10 and ADAM17 silencing. One representative experiment of three is shown. (D and E) pLKO/shADAM10, pLKO/shADAM17, and pLKO/scramble transduced cells were treated as described in Fig. 1. After 48 h, supernatants were collected and analyzed for the presence of sMICB or CD138. The mean of three independent experiments is shown. (F) SKO-007(J3) cells were treated as described in Fig. 1. After 48 h, cells were washed and plated at 0.5 × 106 cells/ml in fresh medium. Different doses of the specific ADAM10 inhibitor GI254023X were added and left for an additional 18 h. Supernatants were collected and analyzed for the presence of sMICB. A representative experiment of three is shown. *p < 0.05.

Because efficient MICB proteolysis depends on ligand recruitment to detergent-resistant microdomains (32), we also investigated the localization of MICB, ADAM10, and Erp5 within the membrane rafts in response to chemotherapeutic drugs. Our findings show that a high proportion of MICB and ADAM10 colocalize in the membrane rafts, and their distribution is not altered by drug treatment. In contrast, Erp5 does not localize within raft microdomains, suggesting that Erp5 and MIC form transitory mixed disulfide complexes on the cell membrane outside of the raft microdomains (Supplemental Fig. 3).

Altogether, these data demonstrate that ADAM10 plays a predominant role on drug-induced MICB release from human MM cells, with the proteolytic process primarily occurring in the membrane rafts.

ADAM10 is upregulated in MM cell lines and in patient-derived malignant PCs in response to genotoxic stress

Because we defined an important role for ADAM10 in the regulation of MIC shedding by genotoxic agents, we asked whether this could be attributable to the drug-induced modulation of ADAM10 expression and activity. Fig. 5A and 5B show that ADAM10 increases sharply at 48 h after DOX or MEL treatment, as evaluated by immunofluorescence and FACS analysis, and this increase persists at 72 h. Similar results were obtained by Western blot analysis, which did not reveal any change in the expression of other enzymes involved in MIC proteolysis: ADAM17 and Erp5 (Fig. 5C). In addition, increased levels of ADAM10 mRNA were observed in drug-treated cells (Fig. 5D). Accordingly, increased ADAM10 expression resulted in enhanced enzymatic activity in DOX-treated cells (Fig. 5E).

FIGURE 5.
FIGURE 5.

ADAM10 is upregulated in response to genotoxic stress in several MM cell lines and in ex vivo malignant CD138+/CD38+ PCs. SKO-007(J3) cells were plated at 3 × 105 cells/ml and treated with DOX and MEL, as described in Fig. 1. (A and B) After 24, 48, and 72 h, cell surface expression of ADAM10 was evaluated by performing immunofluorescence and cytofluorimetric analysis by gating on PI− cells. Values reported represent the mean fluorescent intensity (MFI) of ADAM10 subtracted from the MFI value of the isotype-control Ig. One representative experiment (A) and the mean of three independent experiments (B) are shown. (C) After 48 h of drug treatment, Western blot analysis was performed on total cell lysates using anti-ADAM10–, anti-ADAM17–, or anti-ERp5–specific Abs. Protein loading was normalized using β-actin. One representative experiment of three is shown. (D) Real-time PCR analysis of total mRNA obtained from MM cells, unstimulated or treated with drugs, as described above. Data, expressed as fold change units, were normalized with β-actin and compared with the untreated cells, which were considered calibrators. Data represent the mean of at least four independent experiments. (E) Cells treated with drugs for 72 h were harvested, and ADAM10 catalytic activity was evaluated on total cellular lysates in the presence of 5-FAM/QXL 520 ADAM10 substrate after 30 and 60 min, as described in Materials and Methods. The mean of three independent experiments is shown. (F) Different MM cell lines were plated at 3 × 105 cells/ml and treated with low doses of DOX and MEL, as indicated in Materials and Methods, for 48 h. Cell surface expression of ADAM10 was evaluated as described in (A). Values reported represent the MFI of ADAM10 subtracted from the MFI value of the isotype-control Ig. Mean of at least three independent experiments is shown. (G) Total cells isolated from the bone marrow of MM patients were cultured in the presence of IL-3 (20 ng/ml), IL-6 (2 ng/ml), and DOX (0.05 μM) or MEL (10 μM) for 48 h. Cells were stained with anti-ADAM10 mAb plus GAM-PE and with anti-CD38/FITC and CD138/allophycocyanin. ADAM10 expression was analyzed on CD38+/CD138+ cells. Three representative patients (P1, P5, and P9) are shown. (H) MFI values of ADAM10 subtracted from the MFI value of the isotype-control Ig on CD138+/CD38+ cells from nine patients (P1–P9). Each patient is represented by an individual symbol. (I) CD138+ cells were purified using immunomagnetic beads and treated as described in (A). After 48 h, cells were harvested, RNA was extracted, and real-time PCR was performed. Data, expressed as fold change units, were normalized with β-actin and compared with the untreated cells, which were considered calibrators. Values reported represent the mean of four independent experiments relative to patients P5, P9, P11, and P12. *p < 0.05, **p < 0.01. ns, not significant.

To exclude that the drug-induced enhancement of ADAM10 expression was unique to the SKO-007(J3) MM cell line, we extended our analysis to a large panel of MM cell lines and malignant PCs from MM patients by gating on CD138+/CD38+ PCs (Supplemental Fig. 4A). Drug treatment induced a significant upregulation of ADAM10 cell surface expression, from its basal level of expression, in all MM cell lines tested independently (Fig. 5F). With regard to ADAM10 cell surface expression on malignant PCs, we found that patient-derived PCs displayed different levels of ADAM10 independent of the clinical stage (Supplemental Fig. 4B, 4C). In addition, consistent with the data obtained in MM cell lines, drug treatment of ex vivo PCs resulted in higher levels of surface ADAM10 (Fig. 5G, 5H, Supplemental Fig. 4C), and this increase was accompanied by a concomitant increase in ADAM10 mRNA (Fig. 5I). We did not observe any change in the expression of ADAM10 on CD138−CD38− cells upon drug treatment (data not shown).

These results show that ADAM10 expression is upregulated in MM cell lines and in primary malignant PCs in response to genotoxic agents.

Drug-induced ADAM10 upregulation is dependent on ROS generation and is associated with a senescent phenotype

A large body of evidence describes an increased expression of several metalloproteinases in cells undergoing senescence and developing a senescence-associated secretory phenotype (SASP) (37). In addition, our recent data show that drug-induced NKG2DLs are preferentially expressed on senescent MM cells (22), and their expression requires DNA damage response activation and ROS signaling (26).

Thus, we asked whether MICB secretion and ADAM10 expression were upregulated in MM drug-induced senescent cells. To this end, SKO-007(J3) MM cells were treated with DOX for 72 h and then incubated with the fluorogenic substrate C12FDG to measure βGal activity, which is a marker of senescence; subsequently, βGalhigh and βGallow cells were sorted, as previously described (26). Generally, the percentage of drug-induced senescent SKO-007(J3) cells was ∼60% (22, 26) (data not shown). We found that ADAM10 expression was more prominent on βGalhigh cells at the mRNA and protein levels (Fig. 6A, 6B). Accordingly, MICB release was dramatically increased on βGalhigh senescent cells (Fig. 6C). To further investigate whether drug-induced ADAM10 upregulation was associated with a senescent phenotype in primary PCs, we assayed its expression on βGalhigh and βGallow drug-treated PCs. By simultaneously evaluating cell cycle and βGal activity, we found that the ability of primary PCs to undergo senescence was related to the percentage that was able to divide and to consequently arrest following chemotherapeutic treatment (data not shown). Interestingly, and similar to the data obtained in the SKO(J3)-007 cell line, our results indicate that ADAM10 expression was significantly augmented on βGalhigh cells (Fig. 6D, 6E).

FIGURE 6.
FIGURE 6.

Drug-induced ADAM10 upregulation is associated with a senescent phenotype. (A) A senescence-associated βGal assay was performed using the fluorogenic substrate C12FDG to measure βGal activity by flow cytometry (26). Cells were treated with DOX for 72 h, incubated with C12FDG, and analyzed by immunofluorescence. βGalhigh and βGallow cells were selected by a FACSAria cell sorter through FL-1 green fluorescence emission. mRNA levels of ADAM10 were tested by RT-PCR. β-Actin was used to normalize. (B) ADAM10 expression was evaluated on βGalhigh- and βGallow-sorted cells by immunofluorescence and FACS analysis. (C) βGalhigh- and βGallow-sorted cells were incubated for an additional 24 h in fresh medium, and the presence of sMICB in the supernatants was evaluated by ELISA. Results are representative of one of three independent experiments. (D and E) Total cells isolated from the bone marrow of MM patients were cultured in the presence of IL-3 (20 ng/ml), IL-6 (2 ng/ml), and DOX (10 μM) for 72 h. A senescence-associated βGal assay was performed as described in (A) and then cells were stained with anti-ADAM10/PE, anti-CD38/PerCP, and CD138/allophycocyanin. In (D), SA-βGal activity evaluated by gating on CD38+/CD138+ cells is shown. In (E), ADAM10 expression was analyzed on CD38+/CD138+ βGalhigh and βGallow cells. Values reported represent the mean fluorescence intensity (MFI) of ADAM10 subtracted from the MFI value of the isotype-control Ig. Four representative patients (P42, P39, P43, and P44) are shown.

It is noteworthy that the inhibition of oxidative stress by the antioxidant agent NAC, a glutathione precursor acting as a ROS scavenger, resulted in a marked reduction in ADAM10 on drug-treated SKO-007(J3) MM cells but did not substantially affect its basal expression (Fig. 7A, 7B), suggesting that ROS generation is involved in the upregulation of ADAM10 on senescent cells. In addition, we validated these data using ex vivo primary malignant PCs derived from three patients. As shown in Fig. 7C, MEL treatment of malignant PCs in the presence of NAC reduced ADAM10 expression.

FIGURE 7.
FIGURE 7.

Drug-induced ADAM10 upregulation is dependent on oxidative stress. (A) SKO-007(J3) were pretreated with 10 mM of NAC for 1 h at 37°C and then treated with DOX or MEL for 72 h. ADAM10 expression was evaluated by immunofluorescence and FACS analysis. (B) Values reported represent the mean fluorescence intensity (MFI) of ADAM10 subtracted from the MFI value of the isotype-control Ig. Mean of three independent experiments is shown. (C) Total cells isolated from the bone marrow of MM patients were cultured in the presence of IL-3 (20 ng/ml), IL-6 (2 ng/ml), and MEL (10 μM) for 48 h. Cells were stained with anti-ADAM10/PE and with anti-CD38/FITC and CD138/allophycocyanin. ADAM10 expression was analyzed on CD38+/CD138+ cells. Three representative patients (P38, P40, and P41) are shown. Values reported represent the MFI of ADAM10 subtracted from the MFI value of the isotype-control Ig.

Collectively, our results show that ADAM10-dependent MICB release in response to genotoxic stress is associated with a senescent phenotype and requires ROS generation.

Discussion

The release of NKG2DLs from the cell surface in their soluble form was suggested to play an important role in tumor cell escape from NKG2D-dependent NK cell–mediated immunosurveillance; thus, its inhibition could be exploited as a promising strategy to enhance antitumor immunity.

Indeed, soluble forms of NKG2DLs are present in the serum of cancer patients, and their levels correlate with tumor stage and metastasis (12, 20, 21, 38–41), as well as with reduced expression of NKG2D on NK cells and other cytotoxic lymphocytes (11, 21, 42).

In this study, we showed that treatment of MM cells with chemotherapeutic drugs that induce genotoxic stress stimulates the release of MICA and MICB molecules through a mechanism that is primarily dependent on ADAM10 protease, as well as that the combined use of metalloproteinase inhibitors and chemotherapeutic agents strengthens NK cell–mediated killing of MM cells.

We provide novel evidence that genotoxic stress selectively affects the shedding process of MIC molecules that are sensitive to metalloproteinases cleavage (i.e., MICA*019 or MICB), whereas the release of the most frequent allelic variant in the white population, MICA*008 short allele, is not modulated. Our observations are in line with findings by Ashiru et al. (31), who demonstrated that MICA*008 can be anchored to the plasma membrane via GPI and released in association with exosomes or secreted through a mechanism that is dependent on exocytosis. Thus, effective blockade of sMICA release and accumulation in patient sera might require different strategies, depending on the patient’s MICA genotype.

We also provide further insight into the roles played by ADAM10 and ADAM17 in the shedding of endogenous MIC from MM cells and demonstrate, by means of silencing and usage of pharmacological inhibitors, that ADAM10 plays a predominant role in steady-state conditions, as well as in drug-induced MIC release from MM cells. Several studies, primarily performed in different types of tumor cell lines or transfected cells in steady-state conditions, indicate the direct involvement of both ADAM10 and ADAM17 enzymes in MICA and MICB release, but the relative contribution of these enzymes is still controversial. Boutet et al. (32) found that ADAM17 mediates MICB proteolytic cleavage in CV1 epithelial transfectants; however, recent evidence highlights a cell type–specific role for these metalloproteinases (33). Our results also reveal high levels of ADAM10 expression in several MM cell lines and in primary malignant PCs, with enzyme expression not correlating with the disease stage. Previous studies showed that ADAM10 is required for normal PC function (43), and its expression increases during B cell differentiation, reaching higher levels at the PC stage (44). However, no evidence is available for the malignant cell counterpart, although a role for ADAM10 in hematologic malignancies was recently reported (45, 46). Importantly, we provide evidence that ADAM10 expression is enhanced in MM cell lines and in primary PCs in response to genotoxic agents, and its upregulation requires ROS generation and is associated with the acquisition of a senescent phenotype. Similar to our observations, etoposide, a topoisomerase II inhibitor able to activate the response to DNA damage, was demonstrated to enhance ADAM10 expression (47). However, at variance with our results, Kohga et al. (48) observed that MICA shedding upon epirubicin treatment in hepatocarcinoma cells was accompanied by reduced ADAM10 expression. These discrepancies may depend on the different cellular systems and/or experimental conditions used. Our evidence for ROS involvement in ADAM10 upregulation and MICB shedding are in line with the notion that conditions causing cellular stress, including ionizing radiation, chemotherapeutic agents, or ROS, can lead to increased metalloproteinase-mediated release of cell surface molecules (49–51) and suggest that stress stimuli promote the proteolytic process by upregulating metalloproteinase expression and activity. We also are the first to describe, to our knowledge, that genotoxic stress-induced upregulation of ADAM10 expression and sMICB secretion are primarily associated with senescent cells. Accordingly, senescence-induced generation of sIL-6R was mediated by ADAM10-dependent ectodomain shedding (52). Increased expression of several MMPs, including MMP2 and MMP3, was reported previously in cells undergoing senescence that develop a SASP, but it primarily was associated with promotion of cell invasion (53). The role of the SASP in tumor progression remains unclear and can be beneficial or deleterious, because senescent cells within a tumor can produce secreted factors with both tumor-promoting and tumor-suppressing activities (54). We suggest that release of sNKG2DLs is a component of tumor cell SASP and contributes to the creation of a microenvironment that is suitable for tumor escape. Because MICA and MICB expression on primary PCs was found in a substantial number of MM patients (70 and 50% of patients analyzed, respectively) (data not shown), strategies aimed at stabilizing MIC ligands on the cell surface of tumor cells look promising to increase NK cell recognition mediated by NKG2D. Indeed, our findings show that the combined use of MEL and marimastat enhances NK cell degranulation triggered by autologous malignant PCs. Similarly, the combination of valproate, which is known to upregulate cell surface MICA/B (25, 55), and metalloproteinase inhibitors substantially stabilized cell surface MICA/B on ovarian carcinoma cells and enhanced the efficacy of immune cell therapy in vivo (56). In addition, the use of ADAM10 and ADAM17 inhibitors was shown to ameliorate the response to chemotherapy treatments in different in vivo models of cancer (57, 58), and ADAM inhibitors were used in clinical trials in breast cancer patients (59). Of interest, combined treatment using chemotherapy and metalloproteinase inhibitors recently was proposed as a therapeutic regimen in MM, because drug-induced ADAM-mediated CD138 release was shown to promote tumor growth (60). Overall, our findings suggest that targeting of metalloproteinases in conjunction with chemotherapy could be exploited for NK cell–based immunotherapeutic approaches, thus contributing to avoid the escape of malignant cells from stress-elicited immune responses.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank A. Ludwig for kindly providing the ADAM10 inhibitor GI254023X, Lewis L. Lanier for MICA*019 cDNA, and Dina Milana for valuable technical assistance.

Footnotes

  • This work was supported by grants from the Italian Association for Cancer Research (AIRC Investigator Grant and AIRC 5x1000), the Italian Ministry of University and Research (Projects of National Interest and Basic Research Investment Fund PRIN/2010NECHBX_004/MC), the “Sapienza” University of Rome, the Center of Excellence for Biology and Molecular Medicine, and the Italian Institute of Technology.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    ADAM
    a disintegrin and metalloproteinase
    Ct
    threshold cycle
    DOX
    doxorubicin
    βGal
    β-galactosidase
    GAM
    goat anti-mouse
    GPI
    glycosylphosphatidylinositol
    MEL
    melphalan
    MIC
    MHC class I–related chain
    MM
    multiple myeloma
    MMP
    matrix metalloproteinase
    NAC
    N-acetyl cysteine
    NKG2D
    NK group 2D
    NKG2DL
    NKG2D ligand
    PC
    plasma cell
    PI
    propidium iodide
    ROS
    reactive oxygen species
    s
    soluble
    SASP
    senescence-associated secretory phenotype
    shRNA
    short hairpin RNA.

  • Received October 17, 2014.
  • Accepted May 6, 2015.

References

  1. ↵
    1. Lanier L. L.
    2005. NK cell recognition. Annu. Rev. Immunol. 23: 225–274.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Dai Z.,
    2. C. J. Turtle,
    3. G. C. Booth,
    4. S. R. Riddell,
    5. T. A. Gooley,
    6. A. M. Stevens,
    7. T. Spies,
    8. V. Groh
    . 2009. Normally occurring NKG2D+CD4+ T cells are immunosuppressive and inversely correlated with disease activity in juvenile-onset lupus. J. Exp. Med. 206: 793–805.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Bryceson Y. T.,
    2. H. G. Ljunggren
    . 2008. Tumor cell recognition by the NK cell activating receptor NKG2D. Eur. J. Immunol. 38: 2957–2961.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Guerra N.,
    2. Y. X. Tan,
    3. N. T. Joncker,
    4. A. Choy,
    5. F. Gallardo,
    6. N. Xiong,
    7. S. Knoblaugh,
    8. D. Cado,
    9. N. M. Greenberg,
    10. D. H. Raulet
    . 2008. NKG2D-deficient mice are defective in tumor surveillance in models of spontaneous malignancy. Immunity 28: 571–580.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Raulet D. H.,
    2. S. Gasser,
    3. B. G. Gowen,
    4. W. Deng,
    5. H. Jung
    . 2013. Regulation of ligands for the NKG2D activating receptor. Annu. Rev. Immunol. 31: 413–441.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Champsaur M.,
    2. L. L. Lanier
    . 2010. Effect of NKG2D ligand expression on host immune responses. Immunol. Rev. 235: 267–285.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Cerboni C.,
    2. M. Ardolino,
    3. A. Santoni,
    4. A. Zingoni
    . 2009. Detuning CD8+ T lymphocytes by down-regulation of the activating receptor NKG2D: role of NKG2D ligands released by activated T cells. Blood 113: 2955–2964.
    OpenUrlAbstract/FREE Full Text
    1. Mincheva-Nilsson L.,
    2. O. Nagaeva,
    3. T. Chen,
    4. U. Stendahl,
    5. J. Antsiferova,
    6. I. Mogren,
    7. J. Hernestål,
    8. V. Baranov
    . 2006. Placenta-derived soluble MHC class I chain-related molecules down-regulate NKG2D receptor on peripheral blood mononuclear cells during human pregnancy: a possible novel immune escape mechanism for fetal survival. J. Immunol. 176: 3585–3592.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Salih H. R.,
    2. H. G. Rammensee,
    3. A. Steinle
    . 2002. Cutting edge: down-regulation of MICA on human tumors by proteolytic shedding. J. Immunol. 169: 4098–4102.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Salih H. R.,
    2. H. Antropius,
    3. F. Gieseke,
    4. S. Z. Lutz,
    5. L. Kanz,
    6. H. G. Rammensee,
    7. A. Steinle
    . 2003. Functional expression and release of ligands for the activating immunoreceptor NKG2D in leukemia. Blood 102: 1389–1396.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Groh V.,
    2. J. Wu,
    3. C. Yee,
    4. T. Spies
    . 2002. Tumour-derived soluble MIC ligands impair expression of NKG2D and T-cell activation. Nature 419: 734–738.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Doubrovina E. S.,
    2. M. M. Doubrovin,
    3. E. Vider,
    4. R. B. Sisson,
    5. R. J. O’Reilly,
    6. B. Dupont,
    7. Y. M. Vyas
    . 2003. Evasion from NK cell immunity by MHC class I chain-related molecules expressing colon adenocarcinoma. J. Immunol. 171: 6891–6899.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Chitadze G.,
    2. J. Bhat,
    3. M. Lettau,
    4. O. Janssen,
    5. D. Kabelitz
    . 2013. Generation of soluble NKG2D ligands: proteolytic cleavage, exosome secretion and functional implications. Scand. J. Immunol. 78: 120–129.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Ashiru O.,
    2. P. Boutet,
    3. L. Fernández-Messina,
    4. S. Agüera-González,
    5. J. N. Skepper,
    6. M. Valés-Gómez,
    7. H. T. Reyburn
    . 2010. Natural killer cell cytotoxicity is suppressed by exposure to the human NKG2D ligand MICA*008 that is shed by tumor cells in exosomes. Cancer Res. 70: 481–489.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Mahindra A.,
    2. T. Hideshima,
    3. K. C. Anderson
    . 2010. Multiple myeloma: biology of the disease. Blood Rev. 24(Suppl. 1): S5–S11.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Godfrey J.,
    2. D. M. Benson Jr..
    2012. The role of natural killer cells in immunity against multiple myeloma. Leuk. Lymphoma 53: 1666–1676.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Carbone E.,
    2. P. Neri,
    3. M. Mesuraca,
    4. M. T. Fulciniti,
    5. T. Otsuki,
    6. D. Pende,
    7. V. Groh,
    8. T. Spies,
    9. G. Pollio,
    10. D. Cosman,
    11. et al
    . 2005. HLA class I, NKG2D, and natural cytotoxicity receptors regulate multiple myeloma cell recognition by natural killer cells. Blood 105: 251–258.
    OpenUrlAbstract/FREE Full Text
    1. El-Sherbiny Y. M.,
    2. J. L. Meade,
    3. T. D. Holmes,
    4. D. McGonagle,
    5. S. L. Mackie,
    6. A. W. Morgan,
    7. G. Cook,
    8. S. Feyler,
    9. S. J. Richards,
    10. F. E. Davies,
    11. et al
    . 2007. The requirement for DNAM-1, NKG2D, and NKp46 in the natural killer cell-mediated killing of myeloma cells. Cancer Res. 67: 8444–8449.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Lu X.,
    2. A. Zhu,
    3. X. Cai,
    4. Z. Jia,
    5. W. Han,
    6. L. Ma,
    7. M. Zhou,
    8. K. Qian,
    9. L. Cen,
    10. B. Chen
    . 2012. Role of NKG2D in cytokine-induced killer cells against multiple myeloma cells. Cancer Biol. Ther. 13: 623–629.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Rebmann V.,
    2. P. Schütt,
    3. D. Brandhorst,
    4. B. Opalka,
    5. T. Moritz,
    6. M. R. Nowrousian,
    7. H. Grosse-Wilde
    . 2007. Soluble MICA as an independent prognostic factor for the overall survival and progression-free survival of multiple myeloma patients. Clin. Immunol. 123: 114–120.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Jinushi M.,
    2. M. Vanneman,
    3. N. C. Munshi,
    4. Y. T. Tai,
    5. R. H. Prabhala,
    6. J. Ritz,
    7. D. Neuberg,
    8. K. C. Anderson,
    9. D. R. Carrasco,
    10. G. Dranoff
    . 2008. MHC class I chain-related protein A antibodies and shedding are associated with the progression of multiple myeloma. Proc. Natl. Acad. Sci. USA 105: 1285–1290.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Soriani A.,
    2. A. Zingoni,
    3. C. Cerboni,
    4. M. L. Iannitto,
    5. M. R. Ricciardi,
    6. V. Di Gialleonardo,
    7. M. Cippitelli,
    8. C. Fionda,
    9. M. T. Petrucci,
    10. A. Guarini,
    11. et al
    . 2009. ATM-ATR-dependent up-regulation of DNAM-1 and NKG2D ligands on multiple myeloma cells by therapeutic agents results in enhanced NK-cell susceptibility and is associated with a senescent phenotype. Blood 113: 3503–3511.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Fionda C.,
    2. A. Soriani,
    3. G. Malgarini,
    4. M. L. Iannitto,
    5. A. Santoni,
    6. M. Cippitelli
    . 2009. Heat shock protein-90 inhibitors increase MHC class I-related chain A and B ligand expression on multiple myeloma cells and their ability to trigger NK cell degranulation. J. Immunol. 183: 4385–4394.
    OpenUrlAbstract/FREE Full Text
    1. Fionda C.,
    2. G. Malgarini,
    3. A. Soriani,
    4. A. Zingoni,
    5. F. Cecere,
    6. M. L. Iannitto,
    7. M. R. Ricciardi,
    8. V. Federico,
    9. M. T. Petrucci,
    10. A. Santoni,
    11. M. Cippitelli
    . 2013. Inhibition of glycogen synthase kinase-3 increases NKG2D ligand MICA expression and sensitivity to NK cell-mediated cytotoxicity in multiple myeloma cells: role of STAT3. J. Immunol. 190: 6662–6672.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Wu X.,
    2. Y. Tao,
    3. J. Hou,
    4. X. Meng,
    5. J. Shi
    . 2012. Valproic acid upregulates NKG2D ligand expression through an ERK-dependent mechanism and potentially enhances NK cell-mediated lysis of myeloma. Neoplasia 14: 1178–1189.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Soriani A.,
    2. M. L. Iannitto,
    3. B. Ricci,
    4. C. Fionda,
    5. G. Malgarini,
    6. S. Morrone,
    7. G. Peruzzi,
    8. M. R. Ricciardi,
    9. M. T. Petrucci,
    10. M. Cippitelli,
    11. A. Santoni
    . 2014. Reactive oxygen species- and DNA damage response-dependent NK cell activating ligand upregulation occurs at transcriptional levels and requires the transcriptional factor E2F1. J. Immunol. 193: 950–960.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Molfetta R.,
    2. L. Quatrini,
    3. C. Capuano,
    4. F. Gasparrini,
    5. B. Zitti,
    6. A. Zingoni,
    7. R. Galandrini,
    8. A. Santoni,
    9. R. Paolini
    . 2014. c-Cbl regulates MICA- but not ULBP2-induced NKG2D down-modulation in human NK cells. Eur. J. Immunol. 44: 2761–2770.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Zingoni A.,
    2. G. Palmieri,
    3. S. Morrone,
    4. M. Carretero,
    5. M. Lopez-Botel,
    6. M. Piccoli,
    7. L. Frati,
    8. A. Santoni
    . 2000. CD69-triggered ERK activation and functions are negatively regulated by CD94/NKG2-A inhibitory receptor. Eur. J. Immunol. 30: 644–651.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Cerboni C.,
    2. A. Zingoni,
    3. M. Cippitelli,
    4. M. Piccoli,
    5. L. Frati,
    6. A. Santoni
    . 2007. Antigen-activated human T lymphocytes express cell-surface NKG2D ligands via an ATM/ATR-dependent mechanism and become susceptible to autologous NK- cell lysis. Blood 110: 606–615.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Fürst D.,
    2. G. Solgi,
    3. K. Recker,
    4. D. Mytilineos,
    5. H. Schrezenmeier,
    6. J. Mytilineos
    . 2011. Sequence-based typing of major histocompatibility complex class I chain-related gene A alleles by use of exons 2-5 information. Tissue Antigens 77: 201–205.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Ashiru O.,
    2. S. López-Cobo,
    3. L. Fernández-Messina,
    4. S. Pontes-Quero,
    5. R. Pandolfi,
    6. H. T. Reyburn,
    7. M. Valés-Gómez
    . 2013. A GPI anchor explains the unique biological features of the common NKG2D-ligand allele MICA*008. Biochem. J. 454: 295–302.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Boutet P.,
    2. S. Agüera-González,
    3. S. Atkinson,
    4. C. J. Pennington,
    5. D. R. Edwards,
    6. G. Murphy,
    7. H. T. Reyburn,
    8. M. Valés-Gómez
    . 2009. Cutting edge: the metalloproteinase ADAM17/TNF-alpha-converting enzyme regulates proteolytic shedding of the MHC class I-related chain B protein. J. Immunol. 182: 49–53.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Chitadze G.,
    2. M. Lettau,
    3. J. Bhat,
    4. D. Wesch,
    5. A. Steinle,
    6. D. Fürst,
    7. J. Mytilineos,
    8. H. Kalthoff,
    9. O. Janssen,
    10. H. H. Oberg,
    11. D. Kabelitz
    . 2013. Shedding of endogenous MHC class I-related chain molecules A and B from different human tumor entities: heterogeneous involvement of the “a disintegrin and metalloproteases” 10 and 17. Int. J. Cancer 133: 1557–1566.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Waldhauer I.,
    2. D. Goehlsdorf,
    3. F. Gieseke,
    4. T. Weinschenk,
    5. M. Wittenbrink,
    6. A. Ludwig,
    7. S. Stevanovic,
    8. H. G. Rammensee,
    9. A. Steinle
    . 2008. Tumor-associated MICA is shed by ADAM proteases. Cancer Res. 68: 6368–6376.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Ramani V. C.,
    2. A. Purushothaman,
    3. M. D. Stewart,
    4. C. A. Thompson,
    5. I. Vlodavsky,
    6. J. L. Au,
    7. R. D. Sanderson
    . 2013. The heparanase/syndecan-1 axis in cancer: mechanisms and therapies. FEBS J. 280: 2294–2306.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Hoettecke N.,
    2. A. Ludwig,
    3. S. Foro,
    4. B. Schmidt
    . 2010. Improved synthesis of ADAM10 inhibitor GI254023X. Neurodegener. Dis. 7: 232–238.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Rodier F.,
    2. J. Campisi
    . 2011. Four faces of cellular senescence. J. Cell Biol. 192: 547–556.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Wu J. D.,
    2. L. M. Higgins,
    3. A. Steinle,
    4. D. Cosman,
    5. K. Haugk,
    6. S. R. Plymate
    . 2004. Prevalent expression of the immunostimulatory MHC class I chain-related molecule is counteracted by shedding in prostate cancer. J. Clin. Invest. 114: 560–568.
    OpenUrlCrossRefPubMed
    1. Holdenrieder S.,
    2. P. Stieber,
    3. A. Peterfi,
    4. D. Nagel,
    5. A. Steinle,
    6. H. R. Salih
    . 2006. Soluble MICB in malignant diseases: analysis of diagnostic significance and correlation with soluble MICA. Cancer Immunol. Immunother. 55: 1584–1589.
    OpenUrlCrossRefPubMed
    1. Holdenrieder S.,
    2. P. Stieber,
    3. A. Peterfi,
    4. D. Nagel,
    5. A. Steinle,
    6. H. R. Salih
    . 2006. Soluble MICA in malignant diseases. Int. J. Cancer 118: 684–687.
    OpenUrlCrossRefPubMed
  41. ↵
    1. Baragaño Raneros A.,
    2. B. Suarez-Álvarez,
    3. C. López-Larrea
    . 2014. Secretory pathways generating immunosuppressive NKG2D ligands: New targets for therapeutic intervention. OncoImmunology 3: e28497.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Arreygue-Garcia N. A.,
    2. A. Daneri-Navarro,
    3. A. del Toro-Arreola,
    4. A. Cid-Arregui,
    5. O. Gonzalez-Ramella,
    6. L. F. Jave-Suarez,
    7. A. Aguilar-Lemarroy,
    8. R. Troyo-Sanroman,
    9. A. Bravo-Cuellar,
    10. V. Delgado-Rizo,
    11. et al
    . 2008. Augmented serum level of major histocompatibility complex class I-related chain A (MICA) protein and reduced NKG2D expression on NK and T cells in patients with cervical cancer and precursor lesions. BMC Cancer 8: 16.
    OpenUrlCrossRefPubMed
  43. ↵
    1. Chaimowitz N. S.,
    2. D. J. Kang,
    3. L. M. Dean,
    4. D. H. Conrad
    . 2012. ADAM10 regulates transcription factor expression required for plasma cell function. PLoS ONE 7: e42694.
    OpenUrlCrossRefPubMed
  44. ↵
    Bret, C., D. Hose, T. Reme, A. Kassambara, A. Seckinger, T. Meissner, J. F. Schved, T. Kanouni, H. Goldschmidt, and B. Klein. 2011. Gene expression profile of ADAMs and ADAMTSs metalloproteinases in normal and malignant plasma cells and in the bone marrow environment. Exp. Hematol. 39: 546–557.e8.
  45. ↵
    1. Armanious H.,
    2. P. Gelebart,
    3. M. Anand,
    4. A. Belch,
    5. R. Lai
    . 2011. Constitutive activation of metalloproteinase ADAM10 in mantle cell lymphoma promotes cell growth and activates the TNFα/NFκB pathway. Blood 117: 6237–6246.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Zocchi M. R.,
    2. S. Catellani,
    3. P. Canevali,
    4. S. Tavella,
    5. A. Garuti,
    6. B. Villaggio,
    7. A. Zunino,
    8. M. Gobbi,
    9. G. Fraternali-Orcioni,
    10. A. Kunkl,
    11. et al
    . 2012. High ERp5/ADAM10 expression in lymph node microenvironment and impaired NKG2D ligands recognition in Hodgkin lymphomas. Blood 119: 1479–1489.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Lizama C.,
    2. D. Rojas-Benitez,
    3. M. Antonelli,
    4. A. Ludwig,
    5. R. D. Moreno
    . 2012. Involvement of TACE/ADAM17 and ADAM10 in etoposide-induced apoptosis of germ cells in rat spermatogenesis. J. Cell. Physiol. 227: 829–838.
    OpenUrlCrossRefPubMed
  48. ↵
    1. Kohga K.,
    2. T. Takehara,
    3. T. Tatsumi,
    4. T. Miyagi,
    5. H. Ishida,
    6. K. Ohkawa,
    7. T. Kanto,
    8. N. Hiramatsu,
    9. N. Hayashi
    . 2009. Anticancer chemotherapy inhibits MHC class I-related chain a ectodomain shedding by downregulating ADAM10 expression in hepatocellular carcinoma. Cancer Res. 69: 8050–8057.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Fischer O. M.,
    2. S. Hart,
    3. A. Gschwind,
    4. N. Prenzel,
    5. A. Ullrich
    . 2004. Oxidative and osmotic stress signaling in tumor cells is mediated by ADAM proteases and heparin-binding epidermal growth factor. Mol. Cell. Biol. 24: 5172–5183.
    OpenUrlAbstract/FREE Full Text
    1. Vahdat A. M.,
    2. K. S. Reiners,
    3. V. L. Simhadri,
    4. D. A. Eichenauer,
    5. B. Böll,
    6. A. Chalaris,
    7. V. R. Simhadri,
    8. K. Wiegmann,
    9. H. W. Krell,
    10. S. Rose-John,
    11. et al
    . 2010. TNF-alpha-converting enzyme (TACE/ADAM17)-dependent loss of CD30 induced by proteasome inhibition through reactive oxygen species. Leukemia 24: 51–57.
    OpenUrlCrossRefPubMed
  51. ↵
    1. Kyula J. N.,
    2. S. Van Schaeybroeck,
    3. J. Doherty,
    4. C. S. Fenning,
    5. D. B. Longley,
    6. P. G. Johnston
    . 2010. Chemotherapy-induced activation of ADAM-17: a novel mechanism of drug resistance in colorectal cancer. Clin. Cancer Res. 16: 3378–3389.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Garbers C.,
    2. F. Kuck,
    3. S. Aparicio-Siegmund,
    4. K. Konzak,
    5. M. Kessenbrock,
    6. A. Sommerfeld,
    7. D. Häussinger,
    8. P. A. Lang,
    9. D. Brenner,
    10. T. W. Mak,
    11. et al
    . 2013. Cellular senescence or EGFR signaling induces Interleukin 6 (IL-6) receptor expression controlled by mammalian target of rapamycin (mTOR). Cell Cycle 12: 3421–3432.
    OpenUrlCrossRefPubMed
  53. ↵
    1. Coppé J. P.,
    2. P. Y. Desprez,
    3. A. Krtolica,
    4. J. Campisi
    . 2010. The senescence-associated secretory phenotype: the dark side of tumor suppression. Annu. Rev. Pathol. 5: 99–118.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Kuilman T.,
    2. D. S. Peeper
    . 2009. Senescence-messaging secretome: SMS-ing cellular stress. Nat. Rev. Cancer 9: 81–94.
    OpenUrlCrossRefPubMed
  55. ↵
    1. Armeanu S.,
    2. M. Bitzer,
    3. U. M. Lauer,
    4. S. Venturelli,
    5. A. Pathil,
    6. M. Krusch,
    7. S. Kaiser,
    8. J. Jobst,
    9. I. Smirnow,
    10. A. Wagner,
    11. et al
    . 2005. Natural killer cell-mediated lysis of hepatoma cells via specific induction of NKG2D ligands by the histone deacetylase inhibitor sodium valproate. Cancer Res. 65: 6321–6329.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    1. Huang B.,
    2. R. Sikorski,
    3. P. Sampath,
    4. S. H. Thorne
    . 2011. Modulation of NKG2D-ligand cell surface expression enhances immune cell therapy of cancer. J. Immunother. 34: 289–296.
    OpenUrlCrossRef
  57. ↵
    Fridman, J. S., E. Caulder, M. Hansbury, X. Liu, G. Yang, Q. Wang, Y. Lo, B. B. Zhou, M. Pan, S. M. Thomas, et al. 2007. Selective inhibition of ADAM metalloproteases as a novel approach for modulating ErbB pathways in cancer. Clin. Cancer Res. 13: 1892–1902.
  58. ↵
    1. Zhou B. B.,
    2. M. Peyton,
    3. B. He,
    4. C. Liu,
    5. L. Girard,
    6. E. Caudler,
    7. Y. Lo,
    8. F. Baribaud,
    9. I. Mikami,
    10. N. Reguart,
    11. et al
    . 2006. Targeting ADAM-mediated ligand cleavage to inhibit HER3 and EGFR pathways in non-small cell lung cancer. Cancer Cell 10: 39–50.
    OpenUrlCrossRefPubMed
  59. ↵
    1. Duffy M. J.,
    2. M. Mullooly,
    3. N. O’Donovan,
    4. S. Sukor,
    5. J. Crown,
    6. A. Pierce,
    7. P. M. McGowan
    . 2011. The ADAMs family of proteases: new biomarkers and therapeutic targets for cancer? Clin. Proteomics 8: 9.
    OpenUrlCrossRefPubMed
  60. ↵
    1. Ramani V. C.,
    2. R. D. Sanderson
    . 2014. Chemotherapy stimulates syndecan-1 shedding: a potentially negative effect of treatment that may promote tumor relapse. Matrix Biol. 35: 215222.
    OpenUrlCrossRefPubMed
View Abstract
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section