| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print September 12, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of
* Surgery,
Pathology, and
Cell Biology and Physiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA
1Correspondence: University of Pittsburgh Cancer Institute, Surgery and Bioengineering, Hillman Cancer Center, 5117 Centre Avenue, Room G.21, Pittsburgh, PA 15213, USA. E-mail: lotzemt{at}upmc.edu
ABSTRACT
High mobility group box 1 (HMGB1) is one of the recently defined damage-associated molecular pattern molecules, passively released from necrotic cells and secreted by activated macrophage/monocytes. Whether cytolytic cells induce HMGB1 release from tumor cells is not known. We developed a highly sensitive method for detecting intracellular HMGB1 in tumor cells, allowing analysis of the type of cell death and in particular, necrosis. We induced melanoma cell death with cytolytic lymphokine-activated killing (LAK) cells, tumor-specific cytolytic T lymphocytes, TRAIL, or granzyme B delivery and assessed intracellular HMGB1 retention or release to investigate the mechanism of HMGB1 release by cytolytic cells. HMGB1 release from melanoma cells (451Lu, WM9) was detected within 4 h and 24 h following incubation with IL-2-activated PBMC (LAK activity). HLA-A2 and MART1 or gp100-specific cytolytic T lymphocytes induced HMGB1 release from HLA-A2-positive and MART1-positive melanoma cells (FEM X) or T2 cell-loaded, gp100-specific peptides. TRAIL treatment, however, induced HMGB1 release, and it is interesting that this extrinsic pathway-mediated cell death was blocked with the pancaspase inhibitor N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone. Conversely, granzyme B delivery did not induce HMGB1 release. HMGB1, along with other intracellular factors released from tumor cells induced by cytolysis, may be important components of the disordered tumor microenvironment. This has important implications for the immunotherapy of patients with cancer. Specifically, HMGB1 may promote healing or immune reactivity, depending on the nature of the local inflammatory response and the presence (or absence) of immune effectors.
Key Words: lymphocytes • necrosis • granzyme • death receptor
INTRODUCTION
Type III cell death or necrosis is a type of cell death defined as membrane discontinuity and leakage of intracellular molecules into the extracellular milieu [1 , 2 ]. Necrosis causes inflammation, activating the local endothelium, recruiting inflammatory cells, and thereby driving tissue matrix degradation, promoting functional changes in the remaining epithelia and stroma [3 4 5 ]. Other types of cell death, apoptosis (Type I) and autophagy (Type II), do not induce the same type of inflammation [2 , 6 7 8 ]. One of the key molecules released by necrotic cells is the nuclear protein and molecular superadaptor protein, high mobility group B 1 (HMGB1), which is released following necrosis but not with apoptosis or autophagic programmed cell death. HMGB1 matures dendritic cells (DC) and polarizes Th cells, predominantly to the Th1 pathway [9 ]. HMGB1 induces chronic inflammation, neoangiogenesis, and stromagenesis and protects cells from death, promoting the expression of antiapoptotic factors [4 ].
CTL and NK cells promote tumor cell apoptosis with delivery of perforin and granzymes directly into cells or through the extrinsic pathway, inducing apoptotic death by triggering death receptors by the TNF family of molecules, including TRAIL, TNF-
, Fas ligand (FasL), and lymphotoxin (LT), thereby inducing apoptosis in cells [10
]. Such cytolytic cells could also induce extracellular delivery of intracellular molecules, including HMGB1, ATP, and lactate dehydrogenase (LDH). The 51Cr and more recently, the calcein release assays are representative methods to measure the leakage of intracellular proteins by effector cells during necrotic death [11
]. There have been no previous studies, to our knowledge, assessing whether HMGB1 is released from tumor cells following tumor cytolysis by immune cells. We examined HMGB1 release from tumor cells following coincubation with NK cells with enhanced lymphokine-activated killing (LAK) activity, induced by IL-2 exposure or mediated by melanoma-specific CTL.
MATERIALS AND METHODS
Reagents
TRAIL (R&D Systems, Minneapolis, MN), granzyme B (Calbiochem, San Diego, CA), N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (Z-VAD-FMK; R&D Systems), and Con A (Sigma Chemical Co., St. Louis, MO) were purchased from the commercial sources indicated. Recombinant IL-2 was the kind gift of Chiron (Emeryville, CA).
Antibodies
Affinity-purified polyclonal rabbit anti-HMGB1 antibody that recognizes a specific HMGB1 peptide (166–172) was generated for our laboratory on contract (Sigma Antibody Service, St. Louis, MO). Mouse anti-HMGB1 mAb (R&D Systems), mouse anti-histone H1 mAb (Stressgen, San Diego, CA), and rabbit anti-histone H2A antibody (Cell Signaling Technology, Beverly, MA) and anti-ß-actin antibody (Sigma) were purchased and used at various concentrations in assays described below. Secondary antibodies, goat anti-rabbit IgG coupled to Alexa 660 and goat anti-mouse IgG Alexa 546 (Molecular Probes, Eugene, OR), were used for two-color flow cytometry (FC). In some studies, antibodies were labeled directly with Alexa 546, Alexa 647, or Alexa 750 with antibody labeling kits (Molecular Probes) as directed by the manufacturer.
Cell lines
The melanoma cell lines, 451Lu and WM9, were kind gifts from Dr. Meenhard Herlyn (Wistar Institute, Philadelphia, PA). MEL397, a HLA-A2-negative/Melan-A-positive melanoma cell line and CTL1520 were kind gifts of Steven Rosenberg (Surgery Branch, National Cancer Institute, Bethesda, MD). FEM X is a HLA-A2-positive /MART1-positive melanoma cell line available from American Type Culture Collection (ATCC; Manassas, VA). T2 is a HLA-A2-positive cell line purchased from ATCC.
Subcellular fractionation of melanoma cells
Cells (1x106/100 ul) were suspended in fractionation buffer containing 20 mM Tris-HCl (pH 7.5), 2 mM EDTA, 1 mM 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride AEBSF (Sigma), 50 uM leupeptin (Sigma), 1 mM benzamidine (Sigma), and 500 nM aprotinin (Sigma). Cell membranes were mechanically disrupted with a 36G needle and serial passage with a syringe at 4ºC. Cells were spun down at 500 g for 5 min at 4ºC. Supernatants were collected and used as a bulk cytosol fraction. Pellets were washed with fractionation buffer and spun down at 500 g for 5 min. Pellets were resuspended at 1 x 107 cells/ml in lysis buffer containing 50 mM Tris, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.25% deoxycholate, 1 mM AEBSF, 50 uM leupeptin, 1 mM benzamidine, and 500 nM aprotinin and used as a bulk nuclear fraction.
Cell culture
PBMC were collected from the buffy coat using Ficoll-PaqueTM Plus (Amersham Biosciences, Uppsala, Sweden) and stored in 90% FBS with 10% DMSO (FisherBiotech, Fair Lawn, NJ) at –80°C. Cells were thawed and incubated at 106/ml, with or without 6000 U/ml recombinant human IL-2 in IMDM (Mediatech Inc., Herndon, VA), supplemented with 10% FBS for 3 or 4 days. IL-2-activated PBMC (LAK) were harvested and mixed with 1 x 105/ml tumor cells at various E:T ratios in IMDM, supplemented with 2% FBS for 4 h or 24 h. Then, cells and culture supernatant were collected and analyzed. Z-VAD-FMK is used with PBMC or TRAIL treatment to block caspase activation at the concentration of 50 uM.
CTL induction and coincubation
HLA-A2+/MART1+ tumor-specific CTL were induced from HLA-A2-positive CD8+ T cells from healthy donors as we reported previously [12
]. Tumor cells were incubated with 1 µM Cell TraceTM Far Red DDAO-SE (Molecular Probes, Eugene, OR) for 15 min at room temperature and washed with PBS twice. T2 cells were incubated with 1 µg/ml gp100-specific peptide for 1 h and washed with PBS once. Tumor cells were coincubated with CTL for 24 h in round-bottom, 96-well plates at a 10:1 E:T ratio in IMDM supplemented with 2% FBS. Cells and supernatant were then collected and analyzed.
TRAIL treatment
Tumor cells/ml (1x105) were incubated with TRAIL (50 ng/ml) for 24 h, with or without Z-VAD-FMK (Sigma-Aldrich, St. Louis, MO).
Granzyme B treatment
Tumor cells (1x106/ml) were incubated with nonreplicating adenovirus and granzyme B (10–20 µg/ml) for 5 h, as we have reported previously [13
]. Following incubation, cells and supernatant were harvested and delivered for assessment of cell death and release of intracellular contents. The replication-incompetent adenovirus enters vesicles along with granzyme B and disrupts them intracellularly, allowing its release into the cytosol.
Western blotting analysis
Tumor cells were incubated with or without PBMC or LAK for 4 h at 5 x 106/ml with 100 µl/well. Culture supernatant was collected, and cell pellets were lysed in lysis buffer. Criterion XT sample buffer was added and boiled at 100ºC for 5 min. Samples were stored at –80ºC. Samples were analyzed by Western blotting. Briefly, rabbit anti-HMGB1 antibody and HRP-conjugated goat anti-rabbit IgG were used for the first and secondary antibody. Rabbit anti-histone H2A antibody and HRP-conjugated goat anti-rabbit IgG were used for the detection of histone H2A within the nuclei. Mouse anti-ß-actin antibody and HRP-conjugated donkey anti-mouse IgG were used for the detection of ß-actin. When performing reblotting of other proteins on the same membrane, antibodies were stripped by incubating in 100 mM 2-ME, 2% SDS, and 62.5 mM Tris-HCl (pH 6.7) at 50ºC for 30 min.
Intracellular staining of HMGB1 and histone H1
Before mixing tumor cells with PBMC or tumor-specific CTL, DDAO-SE dye at a concentration of 1 µM was added and incubated for 15 min at room temperature and then washed twice with PBS. Target cells were treated with effector cells or compounds for 4 h or 24 h. Cultured cells were harvested and fixed with 2% paraformaldehyde (PFA) for 15 min at room temperature. Cells were then washed with FC buffer, which contains 1% FBS and 0.1% sodium azide in PBS once and then permeabilized with 0.2% Triton X-100 with 300 mM sucrose in PBS for 5 min. Cells were washed, and anti-HMGB1 and anti-histone H1 antibodies were added and incubated for 1 h at room temperature. Cells were washed, and secondary Alexa dye-conjugated secondary antibodies were added and incubated for an additional hour at room temperature. Cells were washed and resuspended with 200 uL FC buffer and analyzed by FC.
Immunocytochemistry
Cells (2x104) were placed in 96-well, flat-bottom plates and incubated for 2 h. Cells were treated with or without LAK cells for 24 h at an E:T ratio of 50:1. Cells were fixed with 2% PFA and permeabilized with 0.2% Triton X-100-containing buffer. Cells were stained with rabbit anti-HMGB1 antibody, mouse anti-ß-actin antibody, and Hoechst 33342 and viewed on a confocal-scanning fluorescence microscope (Olympus Fluoview 1000, Malvern, NY).
Annexin V/Sytox Orange staining
Cells (1x105) were resuspended in 50 µl annexin-binding buffer containing 100 mM HEPES, 100 mM NaCl, and 2.5 mM CaCl2. Cells were incubated with 2.5 ul Annexin V-Alexa 647 (Invitrogen, Eugene, OR) and 50 nM Sytox Orange (Molecular Probes) for 15 min.
SR-VAD-FMK staining
Sulforhodamine B-labeled (SR)-VAD-FMK staining is performed following the manufacturer’s instruction. Cells (1x105/200 µl) were added 30x to fluorescence-labeled inhibitor of caspase (FLICA) solution and incubated for 1 h. Cells were washed with FC buffer twice and resuspended in FC buffer.
Flow cytometric analysis
Up to 1 x 104-gated cells were acquired on a FACSArray flow cytometer (BD PharMingen, San Diego, CA). Data were analyzed using FACSDiva software (BD PharMingen). Tumor cells were gated and identified using prestained DDAO-SE, assessing side-scatter and forward-scatter. Gating between HMGB1-positive and -negative cells was determined using a lower oblique line of an untreated living tumor cell population in the HMGB1/histone H1 two-color display as shown.
51Cr release assay
Standard 51Cr release assay was performed at individual E:T ratios (20:1–2.5:1) for 4 h or 24 h. Gamma radiation is detected with a microplate scintillation counter (TopCount, Packard Instrument Co., Boston, MA).
LDH release assay
CytoTox assay kit (Promega, Madison, WI) was used for enzymatic assessment of LDH release following the manufacturer’s instructions. Fluorescence emission at 590 nm is measured with a Safire plate reader (Tecan, Switzerland).
Calcein release assay
Calcein AM (5 µM, Molecular Probes) is added to the tumor cells and incubated for 30 min. Tumor cells were washed with PBS twice and incubated with effector cells for 4 h and 24 h. Release of calcein in the supernatant is measured with a Safire plate reader (Tecan).
Statistical analysis
Percent Cr release, percent LDH release, and percent calcein release were calculated using the following formula: Percent release = 100 x (release of sample–spontaneous release)/(maximal release–spontaneous release). Differences between individual groups were calculated with a Student’s t-test. We considered P < 0.05 as statistically significant.
RESULTS
HMGB1 is located primarily within the nucleus
Immunostaining of melanoma cell lines demonstrates intracellular localization of HMGB1, which is found within the nucleus and only found in the cytosol of mitotic cells (white arrows; Fig. 1A
). We also evaluated subcellular localization of HMGB1 by separating bulk nuclei and cytosol fractions (Fig. 1B)
. HMGB1 was detected in nuclei of FEM X and 451Lu, the nuclei and cytosolic fraction of MEL397 and WM9. Histone H2A is located in only nucleus, and ß-actin is located mainly in the cytosol. The amount of histone H2A content is different in those four different cell lines. The results of Western blotting should be considered with caution, as HMGB1 is readily released from the nucleus following cellular disruption.
|
|
|
|
|
|
|
|
|
How cells die is of considerable importance, not only for the careful husbanding of cellular resources following apoptotic or autophagic death but also for the consequences of necrotic cell death associated with inflammation and reactive angiogenesis and stromagenesis. Here, we show that the nuclear protein, HMGB1, is released following melanoma cell death induced by cytolytic lymphoid cells. It is interesting that HMGB1 was not released following intracellular delivery of granzyme B facilitated by adenovirus infection. Lieberman and colleagues [14 ] reported that HMGB1 was not cleaved with granzyme B treatment. Granzyme A destroys the nuclear envelope by targeting lamins and opens up DNA for degradation by targeting histones [15 ]. It is interesting that HMGB1 is not degraded by granzymes and is released from tumor cells following cytolysis by activated T/NK cells, which kill tumor cells, not only by delivery of perforin/granzyme but also by inducing cell death by triggering the so-called extrinsic, apoptotic pathway, delivering ligands/counter-receptors (FasL, TRAIL, TNF, LT) to tumor cell-expressed cognate death receptors [10 ]. TRAIL is demonstrated here to mediate HMGB1 release from melanoma cell lines. HMGB1 release as well as LDH release were induced by LAK cells. This HMGB1 release was blocked partially by Z-VAD-FMK with one melanoma cell line (451Lu) but not with WM9. The effect of Z-VAD-FMK was comparable with Sytox Orange staining and LDH release. Annexin V staining of 451Lu and WM9 was blocked by Z-VAD-FMK, which also blocked HMGB1 release by TRAIL. LAK-induced HMGB1 release is detected within 4 h, and TRAIL-induced HMGB1 release is observed only after 24 h (data not shown). These results suggest that HMGB1 release is caused by not only necrotic cell death but also as a result of caspase-dependent apoptosis and following hypermembrane permeability. Indeed, release of HMGB1 following apoptosis might be a slow process and only found late during aponecrosis when phagocytic cells are not available. Whether the acute and subacute release of HMGB1 has any difference on the host immune response is still unknown. Very recently, Qin and Tracey, et al. blocked apoptosis with Z-VAD-FMK and decreased serum HMGB1 levels in a mouse septic-shock model [16 ]. Blocking apoptotic pathways could thus be a possible way to limit HMGB1 release.
Known receptors of HMGB1-delivered ligands include TLR2 and TLR4 as well as the receptor for advanced glycation end products (RAGE) [3 ]. Suppression of TLR4 by small interfering RNA (siRNA) induces suppression of tumor growth in various mouse models [17 ]. The serum HMGB1 level is detected in many cancer patients (M. T. Lotze, R. A. DeMarco, H. J. Zeh III, unpublished data). Tumor tissues express a high level of HMGB1 and the receptor RAGE [18 , 19 ]. HMGB1 released from tumor cells, by modifying the activity of other immune cells such as plasmacytoid DC and tissue macrophages, may mediate local immune suppression (Petar Popovic et al., submitted).
Blocking HMGB1 release or signaling may be an important, new strategy for cancer therapy [20 ]. Possible methods to block HMGB1 signaling include: preventing HMGB1 release by promoting HMGB1 cleavage within the cell; protecting cell membrane or nuclear membrane from attack using Type II enzyme inducers such as ethyl pyruvate [21 ]; and sequestering HMGB1 within the nucleus by apoptosis inducers such as gamma or ultraviolet irradiation [3 , 22 ] or through creating GpG and GpA cross-links with platinum [23 ], serving as nucleating sites for HMGB1 sequestration. Alternatively, extracellular HMGB1 signals could be blocked by using anti-HMGB1 antibodies [24 ] or limiting access to nominal receptors, including provision of antibodies to TLR2, TLR4, or RAGE [25 ]. Interference with HMGB1 signaling could also use the HMGB1 antagonist: A box [26 ], siRNA preventing HMGB1 production, or possibly creation and delivery of soluble TLRs [18 ].
HMGB1 release from tumor cells was induced by CTL or NK cells or delivery of exogenous TRAIL but not transfection with granzyme B. Extrinsic, pathway-induced release may elicit local inflammatory responses and be important for the emergence of some tumors arising as the consequence of chronic inflammation and mediate, paradoxically, emergent systemic and local immune suppression.
ACKNOWLEDGEMENTS
This work was supported by 1 PO1 CA 101944-01 (M. T. L.), Integrating NK and DC into Cancer Therapy, and 1 R21 CA115059-01 (M. T. L. and David L. Bartlett), Isolated Hepatic Perfusion with Oxaliplatin. We appreciated helpful discussions with Dr. Ramin Lotfi, Ms. Elisa Latorre, and Ms. Katie Horvath.
Received March 5, 2006; revised August 1, 2006; accepted August 2, 2006.
REFERENCES
-Type-1 polarized dendritic cells: a novel immunization tool with optimized CTL-inducing activity Cancer Res. 64,5934-5937This article has been cited by other articles:
|
|
 |
|
  W. Jiang, C. W. Bell, and D. S. Pisetsky The Relationship between Apoptosis and High-Mobility Group Protein 1 Release from Murine Macrophages Stimulated with Lipopolysaccharide or Polyinosinic-Polycytidylic Acid J. Immunol., May 15, 2007; 178(10): 6495 - 6503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Bianchi DAMPs, PAMPs and alarmins: all we need to know about danger J. Leukoc. Biol., January 1, 2007; 81(1): 1 - 5. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
