Purpose: In Hodgkin's lymphoma, constitutive activation of NF-κB promotes tumor cell survival and proliferation. The molecular chaperone heat shock protein 90 (HSP90) has immune regulatory activity and supports the activation of NF-κB in Hodgkin's lymphoma cells.

Experimental Design: We analyzed the effect of HSP90 inhibition on viability and NF-κB activity in Hodgkin's lymphoma cells and the consequences for their recognition and killing through natural killer (NK) cells.

Results: The novel orally administrable HSP90 inhibitor BIIB021 (CNF2024) inhibited Hodgkin's lymphoma cell viability at low nanomolar concentrations in synergy with doxorubicin and gemcitabine. Annexin V/7-aminoactinomycin D binding assay revealed that BIIB021 selectively induced cell death in Hodgkin's lymphoma cells but not in lymphocytes from healthy individuals. We observed that BIIB021 inhibited the constitutive activity of NF-κB and this was independent of IκB mutations. Furthermore, we analyzed the effect of HSP90 inhibition on NK cell–mediated cytotoxicity. BIIB021 induced the expression of ligands for the activating NK cell receptor NKG2D on Hodgkin's lymphoma cells resulting in an increased susceptibility to NK cell–mediated killing. In a xenograft model of Hodgkin's lymphoma, HSP90 inhibition significantly delayed tumor growth.

Conclusions: HSP90 inhibition has direct antitumor activity in Hodgkin's lymphoma in vitro and in vivo. Moreover, HSP90 inhibition may sensitize Hodgkin's lymphoma cells for NK cell–mediated killing via up-regulation of ligands engaging activating NK cell receptors. (Clin Cancer Res 2009;15(16):5108–16)

Translational Relevance

Although to date most Hodgkin's lymphoma patients achieve long-term remission, there are still limited treatment options for patients with primary progressive and multiple relapsed disease. Thus, novel therapeutic strategies are needed. The work presented in this article shows that the novel heat shock protein 90 inhibitor BIIB021 (CNF2024) has direct in vitro and in vivo efficacy in Hodgkin's lymphoma cell lines and might moreover have favorable effects for the antitumor immunity mediated by natural killer cells. Therefore, the data presented here directly provide the basis for a clinical application of heat shock protein 90 inhibitors in Hodgkin's lymphoma and other malignancies.

The ubiquitously expressed molecular chaperone heat shock protein 90 (HSP90) promotes tumor cell survival and proliferation by maintaining conformation, stability, and activity of several key oncogenic client proteins (1). Formation of HSP90 cochaperone complexes is essential for the function of HSP90 and requires ATP hydrolysis by a NH2-terminal ATPase. HSP90 inhibitors, such as the ansamycin 17-N-allylamino-17-demethoxygeldanamycin (17-AAG), inhibit ATPase function resulting in ubiquitination and proteasomal degradation of the client proteins (1, 2). Consequently, inhibition of HSP90 induces cell death in a broad variety of solid and hematologic tumor cell lines in vitro (2). However, 17-AAG has significant clinical limitations, including weak target potency and low aqueous solubility (3). Furthermore, initial clinical trials have suggested low bioavailability and considerable toxicity of 17-AAG, prompting the development of novel small-molecule HSP90 inhibitors such as the orally administrable purine scaffold BIIB021 (CNF2024; refs. 48).

Besides stabilizing oncogenic proteins, HSP90 is critically involved in antitumor immunity (9). As recently reported, HSP90 inhibition decreases the cell surface MHC-I expression and impairs loading of MHC-I with peptides, suggesting an unfavorable effect of HSP90 inhibition on tumor cell recognition by CD8+ T cells (10). Furthermore, inhibition of HSP90 has been shown to decrease antigen presentation and cytokine secretion by dendritic cells (11). However, the effect of HSP90 inhibition on natural killer (NK) cells, innate lymphocytes that detect MHC-I-negative cells, is less well understood.

Current evidence indicates that HSP90 is overexpressed in primary and cultured Hodgkin's lymphoma cells and that 17-AAG-mediated inhibition of HSP90 induces cell cycle arrest and cell death in Hodgkin's lymphoma cell lines in vitro (12). Importantly, HSP90 supports the constitutive activation of the critical survival signal NF-κB in Hodgkin's lymphoma cells by stabilizing the upstream kinase IKK (13, 14). The relevance of the upstream signaling by IKKs (13, 14) is not yet analyzed, although ∼40% of Hodgkin's lymphoma cases have defective IκB proteins (15). Moreover, the in vivo activity of HSP90 inhibition in Hodgkin's lymphoma and possible consequences on the innate immunity, such as NK cell activity, are currently unknown.

Here we investigated the effect of HSP90 inhibition on viability and NF-κB activity in Hodgkin's lymphoma–derived cell lines with functional and defective IκBs and the consequences for the activation of NK cells. We report that the novel orally administrable HSP90 inhibitor BIIB021 selectively targets Hodgkin's lymphoma cells by inhibiting NF-κB independent of the IκB mutation status. Moreover, HSP90 inhibition induces expression of activating NK cell ligands on Hodgkin's lymphoma cells, resulting in an increased susceptibility to NK cell–mediated killing. In line with these findings, we observed in vivo antitumor activity of HSP90 inhibition in a xenograft model of human Hodgkin's lymphoma.

Cell lines, peripheral blood mononuclear cells, and primary NK cells

The Hodgkin's lymphoma cell lines used in this study include DEV (B-cell, nodular lymphocyte predominant Hodgkin's lymphoma, IκB status unknown), KM-H2 (B-cell, mixed cellularity, defective IκBα, wild-type IκBϵ), L428 (B-cell, nodular sclerosis, defective IκBα and IκBϵ), L1236 (B-cell, mixed cellularity, wild-type IκBα and IκBϵ), L591 (EBV-positive, B-cell, nodular sclerosis, wild-type IκBα and IκBϵ), and L540 and L540cy (both T-cell and nodular sclerosis, IκBα wild-type in L540, unknown in L540cy, IκBϵ status unknown in both; refs. 1620). Cells were cultured in RPMI 1640 (PAA) supplemented with 10% heat inactivated FCS (Invitrogen), 50 μg/mL penicillin, 50 μg/mL streptomycin, and 2 mmol/L l-glutamine (Sigma) at 37°C in a 5% CO2 atmosphere.

Peripheral blood mononuclear cells were isolated by Ficoll-Paque density gradient centrifugation from healthy human donor buffy coats using Leucosep columns (Greiner Bio-One). NK cells were isolated by depleting non-NK cells with the NK Cell Isolation Kit and VarioMACS (Miltenyi) as described previously (21). NK cells were cultured in MEM-α (Sigma) supplemented with 50 μg/mL penicillin, 50 μg/mL streptomycin, 20% FCS, and 10 units/mL recombinant human interleukin-2 (R&D Systems).

Viability assay and combination experiments

2,3-Bis[2-methoxy-4-nitro-S-sulfophenyl]H-tetrazolium-5-carboxanilide inner salt (XTT) viability assays following 48 h incubation were done as described previously (22). XTT was purchased from Sigma.

The IC50 values were calculated by linear regression as the mean ± SD of at least three independent experiments.

For combination experiments, cells were incubated for 48 h in a 96-well XTT assay with varying ratios of 17-AAG (Invivogen) or BIIB0215

5Lundgren K, Zhang H, Brekken J, et al. BIIB021, an orally available, fully synthetic small molecule inhibitor of the heat shock protein, Hsp90, submitted for publication 2008.

(Biogen Idec) and with doxorubicin (Medac) or gemcitabine (Lilly). Doxorubicin and gemcitabine were applied in concentrations determined by three independent experiments corresponding to the IC50, 1/2 × IC50, and 2 × IC50 as recommended by Chou and Talalay (23). The IC50 were 0.17 and 0.08 μmol/L for doxorubicin and 0.14 and 0.12 μmol/L for gemcitabine in L428 and L540cy, respectively. Confirmation experiments using 17-AAG as HSP90 inhibitor were done accordingly (data not shown). The combination effect was analyzed with Calcusyn software (Biosoft) according to the manufacturer's instructions based on the median effect method (23). Combination indices 1, <1, and >1 indicate additive effects, synergy, and antagonism, respectively.

ELISA-based NF-κB assay

NF-κB activity was measured using the immunosorbent p65-TransAM assay from Active Motif according to the manufacturer's instructions.

Briefly, nuclear extracts were prepared after 12 h incubation with 1 μmol/L BIIB021 or vehicle (DMSO, control), 20 μg nuclear protein extract was loaded on plates coated with an oligonucleotide containing a NF-κB consensus site (5′-GGGACTTTCC-3′), and DNA-bound NF-κB subunit p65 was detected with anti-p65 antibody and horseradish peroxidase–conjugated secondary antibody (Active Motif). The blank absorbance was subtracted and NF-κB activity was calculated as the mean ± SD of duplicate samples. To monitor the specificity of the assay, wild-type and mutated consensus oligonucleotides were used as competitors for NF-κB binding (Active Motif).

Western blot analysis

Western blot analysis of whole-cell lysates was done as described previously (24). Anti–poly(ADP-ribose) polymerase antibody was obtained from Cell Signaling Technology (9542) and the anti–X-linked inhibitor of apoptosis antibody (clone 48) was from Becton Dickinson. Anti-HSP90 (clone AC88) and anti-HSP70 (clone C92F3A-5) antibodies were purchased from Stressgen.

Flow cytometry

For flow cytometric analysis, cells were pretreated for 12 to 48 h with HSP90 inhibitors (0.8 μmol/L BIIB021 or 6 μmol/L 17-AAG) or vehicle control (DMSO) and incubated for 1 h with 10 μg/mL primary antibody or Fc construct, if appropriate, followed by a 30-min incubation with 10 μg/mL FITC-conjugated goat anti-mouse secondary antibody (Jackson Immuno Research) or FITC-conjugated rabbit anti-human Fc (Dianova). The analysis was done on a FACSCalibur cytometer (Becton Dickinson) as described previously (21). Lymphocytes from healthy donor–derived peripheral blood mononuclear cells were gated by light-scattering characteristics. The following monoclonal antibodies were used: FITC-conjugated CD95/FAS (clone DX2), FITC-conjugated IgG1κ isotype (clone MOPC-21), and MICA/MICB (clone 6D4) were obtained from Becton Dickinson and NKp30-Fc, NKp46-Fc, and NKG2D-Fc constructs were purchased from R&D Systems. CD30-Fc control construct and mouse BH1 isotype-matched control antibody have been described previously (25).

Fluorescence-activated cell sorting analysis of Annexin V/7-aminoactinomycin D–stained cells (Becton Dickinson) was done according to the manufacturer's instructions. For the positive control, cells were incubated 30 min with 37% formaldehyde (see “formalin” in Fig. 4; Sigma).

For cell cycle analysis, cells were fixed for 4 h at 4°C with 90% ethanol and incubated for 4 h with 100 μg/mL RNase (Sigma). After 3 h of staining with 50 μg/mL propidium iodide (Sigma) at 4°C, flow cytometric analysis was done excluding doublettes using CellQuest Pro software (Becton Dickinson).

Cytotoxicity assay

NK cell–mediated cytotoxicity was analyzed by a standard 3 h europium release assay in a 96-well microtiter plate as described previously (21). Briefly, target L428 cells were preincubated 12 h in plain culture medium (data not shown), vehicle (DMSO, control), 0.8 μmol/L BIIB021, or 6 μmol/L 17-AAG.

NK effector cells were mixed with 5 × 103 target cells previously labeled with europium chloride (Fluka) at different ratios. Supernatants were assayed for europium release after 3 h in a gamma counter. Spontaneous release and maximal release were determined by incubation of the target cells in the absence of effector cells and by target cells lysed with 1% Triton X-100, respectively.

The spontaneous release did not exceed 25% of the maximum release. The percentage of specific lysis was calculated as 100 × [(experimental release × spontaneous release) / (maximal release × spontaneous release)].

Xenograft model of human Hodgkin's lymphoma

Subcutaneous solid L540cy tumors were established by injection of L540cy cells into the flank of severe combined immunodeficient mice (Taconic) as described previously (19). Tumor development was measured every 3 days and tumor volume was determined using the formula: (length × width × height) / 2. Animals with established tumors of ∼100 mm3 were assigned randomly to four groups consisting of five animals each. Mice received a HSP90 inhibitor in a dosage that corresponded to 20% of the maximum tolerable dose5,6

6Unpublished data kindly provided by Biogen Idec.

or vector as control: (a) 0.1 N HCl (vehicle, control) or 120 mg/kg BIIB021 via oral gavage or (2) phospholipon/sucrose emulsion (kindly provided by Biogen Idec, vehicle control) or 60 mg/kg 17-AAG (1.5 mg/mouse) as intraperitoneal injection every 3 days (as 17-AAG cannot be administered orally because of physicochemical limitations).

Mice were killed when the mean tumor diameter exceeded 1,200 mm3 or after 30 days. Statistical analysis was done using the GraphPad Prism software applying the unpaired t test. The experiment was conducted following guidelines published by the U.S. Office of Laboratory Animal Welfare and the European Federation of European Laboratory Animal Science and approved by local authorities (authorization no. 50.203.2-K 17, 37/05).

BIIB021 inhibits Hodgkin's lymphoma cell viability in vitro synergistically with conventional chemotherapy

BIIB021 is a novel orally administrable HSP90 inhibitor that is currently being tested in phase I and II clinical trials in solid tumors and chronic lymphocytic leukemia (NCT006187035, NCT00618319, NCT00412412, NCT00344786, and NCT00345189). We tested the effect of BIIB021-mediated HSP90 inhibition in a panel of Hodgkin's lymphoma cell lines reflecting different origin (B and T cells), histology, and EBV status in a 48 h XTT assay. BIIB021 effectively inhibited Hodgkin's lymphoma cell viability in all cell lines tested at low concentrations, with IC50 values of 0.24 to 0.80 μmol/L. In line with previous studies showing a higher target efficacy of BIIB021,5 we found a drastically higher efficacy for BIIB021 compared with 17-AAG, which showed up to 7.5-fold lower IC50 values (Fig. 1A and B).

Fig. 1.

BIIB021 inhibits Hodgkin's lymphoma cell viability in vitro synergistically with doxorubicin and gemcitabine. A, Hodgkin's lymphoma–derived cell lines were cultured for 48 h with vehicle (DMSO, control) or varying concentrations of BIIB021 (black columns) in 96-well XTT assays. Mean ± SD viability of at least three independent experiments (X axis in log scale). B, IC50 of 17-AAG (gray columns) and BIIB021 (black columns) in the indicated Hodgkin's lymphoma cell lines was calculated by linear regression in 48 h XTT assays. Mean ± SD IC50 of at least three independent experiments. L428 (C) and L540cy (D) Hodgkin's lymphoma cells were incubated for 48 h in a 96-well XTT assay with concentrations of BIIB021 and doxorubicin (○) or gemcitabine (□) corresponding to the IC50, 1/2 × IC50, and 2 × IC50. IC50 values were 0.17 and 0.08 μmol/L for doxorubicin and 0.14 and 0.12 μmol/L for gemcitabine in L428 and L540cy, respectively. Dose-effect calculations were done using the Calcusyn software based on the median effect method (24). Combination indices (Y axis) 1, <1, and >1 indicate additive effects, synergy, and antagonism at different concentrations (fractions of the IC50; X axis), respectively.

Fig. 1.

BIIB021 inhibits Hodgkin's lymphoma cell viability in vitro synergistically with doxorubicin and gemcitabine. A, Hodgkin's lymphoma–derived cell lines were cultured for 48 h with vehicle (DMSO, control) or varying concentrations of BIIB021 (black columns) in 96-well XTT assays. Mean ± SD viability of at least three independent experiments (X axis in log scale). B, IC50 of 17-AAG (gray columns) and BIIB021 (black columns) in the indicated Hodgkin's lymphoma cell lines was calculated by linear regression in 48 h XTT assays. Mean ± SD IC50 of at least three independent experiments. L428 (C) and L540cy (D) Hodgkin's lymphoma cells were incubated for 48 h in a 96-well XTT assay with concentrations of BIIB021 and doxorubicin (○) or gemcitabine (□) corresponding to the IC50, 1/2 × IC50, and 2 × IC50. IC50 values were 0.17 and 0.08 μmol/L for doxorubicin and 0.14 and 0.12 μmol/L for gemcitabine in L428 and L540cy, respectively. Dose-effect calculations were done using the Calcusyn software based on the median effect method (24). Combination indices (Y axis) 1, <1, and >1 indicate additive effects, synergy, and antagonism at different concentrations (fractions of the IC50; X axis), respectively.

Close modal

To test the effect of BIIB021 in combination with conventional chemotherapy commonly used in the first- and second-line treatment of Hodgkin's lymphoma, such as doxorubicin and gemcitabine (26), we applied the median effect method using Calcusyn software (23). The combination of BIIB021 with doxorubicin or gemcitabine resulted in additive or even synergistic effects in L428 and L540cy cells (Fig. 1C and D). However, we observed a high variability in higher dosages, which could be due to saturation effects.

HSP90 inhibitors selectively target Hodgkin's lymphoma cells and inhibit constitutionally active NF-κB despite defective IκB

Different Hodgkin's lymphoma cell lines were cultured for 48 h with vehicle (DMSO, control) or 0.8 μmol/L BIIB021 and apoptosis induction was monitored by detection of poly(ADP-ribose) polymerase cleavage. The induction of apoptosis in response to HSP90 inhibition was observed in all tested Hodgkin's lymphoma cell lines (Fig. 2A) and was associated with increased expression levels of HSP70 and HSP90 as published previously (ref. 12; data not shown). Interestingly, a reduced cellular level of X-linked inhibitor of apoptosis, which is known as a crucial mediator of apoptosis resistance in Hodgkin's lymphoma cells (27), was observed in BIIB021-treated L428 cells (Fig. 2B) and L540cy cells (data not shown).

Fig. 2.

BIIB021 selectively targets Hodgkin's lymphoma cells and inhibits constitutionally active NF-κB independent of functional IκB. A, Hodgkin's lymphoma–derived cell lines were cultured for 48 h with vehicle (DMSO; -) or 0.8 μmol/L BIIB021 (+). Whole-cell extracts were prepared and tested in Western blot analysis using anti–poly(ADP-ribose) polymerase and anti-β-actin antibodies. B, L428 Hodgkin's lymphoma cells were cultured for up to 72 h with vehicle (DMSO) or 0.8 μmol/L BIIB021. Whole-cell extracts were prepared and tested in Western blot analysis using anti–X-linked inhibitor of apoptosis (XIAP) and anti-β-actin antibodies. Equivalent results were observed in L540cy (data not shown). C, healthy donor–derived peripheral blood mononuclear cells and L428 Hodgkin's lymphoma cells were cultured for 48 h with vehicle (DMSO) or 0.8 μmol/L BIIB021. Cells were stained with Annexin V-FITC and 7-aminoactinomycin D and analyzed by flow cytometry. Lymphocytes were gated using light-scattering characteristics. Mean ± SD percentage of Annexin V and 7-aminoactinomycin D–positive cells from three independent experiments. D, Hodgkin's lymphoma–derived cell lines were cultured for 12 h with vehicle (DMSO, control; white columns) or 0.8 μmol/L BIIB021 (black columns). NF-κB activity in nuclear extracts was analyzed by ELISA-based p65-TransAM NF-κB assay. Wild-type (DEV+WT) and mutated (DEV+Mut) consensus oligonucleotides were used as competitors for NF-κB binding as controls. Columns, mean of a duplicate; bars, SD. *, P < 0.05, unpaired t test.

Fig. 2.

BIIB021 selectively targets Hodgkin's lymphoma cells and inhibits constitutionally active NF-κB independent of functional IκB. A, Hodgkin's lymphoma–derived cell lines were cultured for 48 h with vehicle (DMSO; -) or 0.8 μmol/L BIIB021 (+). Whole-cell extracts were prepared and tested in Western blot analysis using anti–poly(ADP-ribose) polymerase and anti-β-actin antibodies. B, L428 Hodgkin's lymphoma cells were cultured for up to 72 h with vehicle (DMSO) or 0.8 μmol/L BIIB021. Whole-cell extracts were prepared and tested in Western blot analysis using anti–X-linked inhibitor of apoptosis (XIAP) and anti-β-actin antibodies. Equivalent results were observed in L540cy (data not shown). C, healthy donor–derived peripheral blood mononuclear cells and L428 Hodgkin's lymphoma cells were cultured for 48 h with vehicle (DMSO) or 0.8 μmol/L BIIB021. Cells were stained with Annexin V-FITC and 7-aminoactinomycin D and analyzed by flow cytometry. Lymphocytes were gated using light-scattering characteristics. Mean ± SD percentage of Annexin V and 7-aminoactinomycin D–positive cells from three independent experiments. D, Hodgkin's lymphoma–derived cell lines were cultured for 12 h with vehicle (DMSO, control; white columns) or 0.8 μmol/L BIIB021 (black columns). NF-κB activity in nuclear extracts was analyzed by ELISA-based p65-TransAM NF-κB assay. Wild-type (DEV+WT) and mutated (DEV+Mut) consensus oligonucleotides were used as competitors for NF-κB binding as controls. Columns, mean of a duplicate; bars, SD. *, P < 0.05, unpaired t test.

Close modal

Although HSP90 is present in abundance in both malignant and normal cells, 17-AAG has been shown previously to selectively target malignant cells (28). To determine whether BIIB021-mediated HSP90 inhibition selectively kills Hodgkin's lymphoma cells, we compared the effects on L428 cells and healthy donor–derived peripheral blood mononuclear cells. Annexin V-FITC and 7-aminoactinomycin D staining revealed that both agents, BIIB021 and 17-AAG, induced cell death in L428 cells, whereas healthy donor lymphocytes remained unaffected (Fig. 2C).

Previous reports have suggested that the effect of HSP90 inhibition in Hodgkin's lymphoma cells is dependent on functional IκB proteins (14). Because in ∼40% of Hodgkin's lymphoma cases Hodgkin's lymphoma cells harbor mutated nonfunctional IκB proteins, we tested the activity of NF-κB in response to HSP90 inhibition in Hodgkin's lymphoma cells with mutated IκB (L428 and KM-H2) or functional IκB (all other Hodgkin's lymphoma cell lines tested; ref. 29). Cells were preincubated for 12 h with 1 μmol/L BIIB021 and nuclear extracts were tested for DNA-bound p65 by a highly sensitive ELISA-based NF-κB assay. Basal NF-κB activity varied among the tested cell lines; however, inhibition of HSP90 by CNF2024 decreased the constitutive NF-κB activity in all tested cell lines by 14% to 70% and was significant for L591, L540, L540cy, L428, and DEV (*, P < 0.05, unpaired t test). Importantly, BIIB021-induced NF-κB inhibition was not dependent on the presence of functional IκB nor on the NF-κB inducing effect (Fig. 2D). 17-AAG in equipotent concentrations led to similar results (data not shown).

BIIB021 induces expression of NK cell activating NKG2D ligands on Hodgkin's lymphoma cells

Recent findings indicate that HSP90 is critically involved in tumor cell recognition by dendritic cells and T cells (10, 11). However, the effect of HSP90 inhibition on target cell lysis by NK cells is currently unknown.

To examine the effect of HSP90 inhibition on the expression of activating NK cell ligands on Hodgkin's lymphoma cells, L428 cells were pretreated for 12 h with 0.8 μmol/L BIIB021 (Fig. 3A and B, open histograms, left column), 6 μmol/L 17-AAG (Fig. 3A and B, open histograms, right column), or vehicle only (Fig. 3A and B, DMSO, filled histograms). Cells were analyzed for surface expression of the indicated NK cell ligands by flow cytometry after staining with recombinant Fc constructs (NKP30-Fc, NKP46-Fc, NKG2D-Fc, and CD30-Fc as control) and secondary FITC-labeled anti-Fc antibody (Fig. 3A) or antibodies directed against ligands (MICA/B, ULBP2, CD95-FITC, and isotype controls) and FITC-conjugated secondary anti-mouse antibody (Fig. 3B), if appropriate.

Fig. 3.

HSP90 inhibition induces NKG2D ligand expression on L428. A and B, L428 Hodgkin's lymphoma cells were pretreated for 12 h with 0.8 μmol/L BIIB021 (open histograms) or 6 μmol/L 17-AAG (open histograms) or vehicle only (DMSO; filled histograms). Cells were harvested and incubated for 1 h with 10 μg/mL primary antibody, if appropriate, followed by 30 min incubation with 10 μg/mL FITC-conjugated goat anti-mouse antibody or with 50 μg/mL of the indicated Fc-fusion constructs followed by incubation with 10 μg/mL anti-human Fc antibody. Cells were washed and analyzed on FACSCalibur cytometer (Becton Dickinson). C, mean fluorescence intensity calculated as difference from isotype control.

Fig. 3.

HSP90 inhibition induces NKG2D ligand expression on L428. A and B, L428 Hodgkin's lymphoma cells were pretreated for 12 h with 0.8 μmol/L BIIB021 (open histograms) or 6 μmol/L 17-AAG (open histograms) or vehicle only (DMSO; filled histograms). Cells were harvested and incubated for 1 h with 10 μg/mL primary antibody, if appropriate, followed by 30 min incubation with 10 μg/mL FITC-conjugated goat anti-mouse antibody or with 50 μg/mL of the indicated Fc-fusion constructs followed by incubation with 10 μg/mL anti-human Fc antibody. Cells were washed and analyzed on FACSCalibur cytometer (Becton Dickinson). C, mean fluorescence intensity calculated as difference from isotype control.

Close modal

Cell surface analysis by fluorescence-activated cell sorting revealed enhanced expression of the NKG2D ligands MICA/B and ULBP2. Quantitative analysis of mean fluorescence showed >100% increase in mean fluorescence for cells stained with NKG2D-Fc as a consequence of HSP90 inhibitor treatment (Fig. 3C). A comparable increase in NKG2D ligand expression in response to HSP90 inhibition was detected in L540cy (data not shown). No significant increase was detected for ligands of NKP30 and NKP46. Concordantly, analysis of mean fluorescence intensity showed no cell surface expression of NKp30 using NKp30Fc but a slight increase in NKp46Fc binding to Hodgkin's lymphoma cells in response to HSP90 inhibition. However, recent data indicate that ligands for NKp30 (30) as well as NKp46 (31) can be released from target cells and might therefore not be detectable on the cell surface under certain conditions.

HSP90 inhibition targets Hodgkin's lymphoma cells for NK cell–mediated lysis

In contrast to the effects observed after 48 h of incubation (Figs. 1A and 2A and B), XTT analysis and Annexin V-FITC/7-aminoactinomycin D/fluorescence-activated cell sorting analysis revealed that 12 h of incubation with the 17-AAG and BIIB021 did not inhibit Hodgkin's lymphoma cell viability nor induce cell death (data not shown). Staining with propidium iodide and fluorescence-activated cell sorting analysis of cell cycle distribution revealed a slight shift of BIIB021-treated L428 cells toward G2 phase as an early effect of HSP90 inhibition (Fig. 4A and B).

Fig. 4.

HSP90 inhibition sensitizes L428 Hodgkin's lymphoma cells for NK cell–mediated cytotoxicity. L428 Hodgkin's lymphoma cells were pretreated for 12 h with varying concentrations of BIIB021 or 17-AAG or vehicle (DMSO, control) and tested for cell cycle distribution or exposed to NK cell–mediated cytolysis in varying E:T ratios. Measurements were done in triplicates. One of three independent experiments. A and B, L428 Hodgkin's lymphoma cells treated with vehicle (DMSO; C) or 0.8 μmol/L BIIB021 (D) were stained with propidium iodide and analyzed for cell cycle distribution by flow cytometry. One of three independent experiments. C, NK cell–mediated lysis of L428 Hodgkin's lymphoma cells pretreated with 0.8 μmol/L (▴, dotted line) or 0.4 μmol/L (▾, dotted line) BIIB021 or vehicle (control; ▪, solid line). D, NK cell–mediated lysis of L428 Hodgkin's lymphoma cells pretreated with 6 μmol/L (▴, dotted line) or 3 μmol/L (▾, dotted line) 17-AAG or vehicle (control; ▪, solid line).

Fig. 4.

HSP90 inhibition sensitizes L428 Hodgkin's lymphoma cells for NK cell–mediated cytotoxicity. L428 Hodgkin's lymphoma cells were pretreated for 12 h with varying concentrations of BIIB021 or 17-AAG or vehicle (DMSO, control) and tested for cell cycle distribution or exposed to NK cell–mediated cytolysis in varying E:T ratios. Measurements were done in triplicates. One of three independent experiments. A and B, L428 Hodgkin's lymphoma cells treated with vehicle (DMSO; C) or 0.8 μmol/L BIIB021 (D) were stained with propidium iodide and analyzed for cell cycle distribution by flow cytometry. One of three independent experiments. C, NK cell–mediated lysis of L428 Hodgkin's lymphoma cells pretreated with 0.8 μmol/L (▴, dotted line) or 0.4 μmol/L (▾, dotted line) BIIB021 or vehicle (control; ▪, solid line). D, NK cell–mediated lysis of L428 Hodgkin's lymphoma cells pretreated with 6 μmol/L (▴, dotted line) or 3 μmol/L (▾, dotted line) 17-AAG or vehicle (control; ▪, solid line).

Close modal

Next, we analyzed the effect of the observed increase in NKG2D ligand expression in response to HSP90 inhibition on the susceptibility of L428 cells to NK cell–mediated killing. L428 cells were incubated for 12 h as described above and nonspecific effects of DMSO were tested by including L428 incubated for 12 h in plain culture medium (data not shown).

Cells were labeled with europium chloride and subjected to a 3 h cytotoxicity assay with MACS-purified healthy donor–derived NK cells. Vehicle only–treated L428 cells were killed in high E:T ratios at a maximum of 30% (Fig. 4C and D). Importantly, treatment with both HSP90 inhibitors BIIB021 (Fig. 4C) and 17-AAG (Fig. 4D) led to a >2-fold increase in specific lysis. These results suggest that the induced expression of NKG2D ligands in response to HSP90 inhibition is associated with an increased susceptibility to NK cell–mediated killing.

HSP90 inhibition inhibits Hodgkin's lymphoma growth in vivo

Current knowledge indicates that 17-AAG-mediated HSP90 inhibition effectively induces cell death in Hodgkin's lymphoma cell lines in vitro (12). To analyze in vivo effects, we established subcutaneous solid Hodgkin's lymphoma tumors by injecting L540cy cells into the flanks of severe combined immunodeficient mice and monitored tumor development. Animals with established tumors of ∼100 mm3 were assigned randomly into four groups consisting of five animals each and treated once every 3 days with an equivalently toxic dose of a HSP90 inhibitor (17-AAG or BIIB021, 1/5 maximum tolerated dose)7

7Communication from Biogen Idec.

or vehicle as control (phenylmethylsulfonyl fluoride or 0.1 N HCl, respectively).

In mice receiving 17-AAG intraperitoneally every 3 days (60 mg/kg body weight), tumor growth was reduced compared with the control group (Fig. 5A), indicating in vivo activity of HSP90 inhibitors in Hodgkin's lymphoma. However, due to the small number of animals per treatment group and the relatively small differences in tumor size, the differences in tumor growth between 17-AAG-treated mice and phenylmethylsulfonyl fluoride–treated control mice were not statistically significant (P = 0.10, unpaired t test).

Fig. 5.

HSP90 inhibition impairs Hodgkin's lymphoma tumor growth in vivo. Severe combined immunodeficient mice with established subcutaneous solid L540cy tumors of ∼100 mm3 were divided randomly into four different groups and treated with 17-AAG or BIIB021 or the respective vehicle (control) every 3 d. Mice were killed when the mean tumor diameter exceeded 1,200 mm3 or after 30 d. A, mice received either 60 mg/kg body weight 17-AAG in phospholipon/sucrose emulsion (1.5 mg/mouse) or phospholipon/sucrose emulsion only (vector control) per oral gavage every 3 d. Mean ± SE tumor volume in each group. B, mice received either 120 mg/kg body weight BIIB021 in 0.1 N HCl (3 mg/mouse) or 0.1 N HCl only (vector control) per intraperitoneal injection every 3 d. Mean ± SE tumor volume in each group. C, differences in mean tumor size on day 23 (last day of treatment) between BIIB021-treated mice and the respective control mice were tested for statistical significance using the unpaired t test (GraphPad Prism).

Fig. 5.

HSP90 inhibition impairs Hodgkin's lymphoma tumor growth in vivo. Severe combined immunodeficient mice with established subcutaneous solid L540cy tumors of ∼100 mm3 were divided randomly into four different groups and treated with 17-AAG or BIIB021 or the respective vehicle (control) every 3 d. Mice were killed when the mean tumor diameter exceeded 1,200 mm3 or after 30 d. A, mice received either 60 mg/kg body weight 17-AAG in phospholipon/sucrose emulsion (1.5 mg/mouse) or phospholipon/sucrose emulsion only (vector control) per oral gavage every 3 d. Mean ± SE tumor volume in each group. B, mice received either 120 mg/kg body weight BIIB021 in 0.1 N HCl (3 mg/mouse) or 0.1 N HCl only (vector control) per intraperitoneal injection every 3 d. Mean ± SE tumor volume in each group. C, differences in mean tumor size on day 23 (last day of treatment) between BIIB021-treated mice and the respective control mice were tested for statistical significance using the unpaired t test (GraphPad Prism).

Close modal

BIIB021 administered every 3 days orally (120 mg/kg body weight) effectively inhibited growth of subcutaneous L540cy tumors (Fig. 5B). Reduction in growth was statistically significant as tested with the unpaired t test at the end of treatment at day 23 (P = 0.029; Fig. 5C). Importantly, no dose-limiting toxicity of the HSP90 inhibitors was observed in either group.

Current knowledge shows that HSP90 is highly expressed in primary and cultured Hodgkin's lymphoma cells and contributes to Hodgkin's lymphoma cell survival by supporting the constitutive activation of NF-κB (1214). The aim of this study was (a) to determine the effect of 17-AAG- and BIIB021-mediated HSP90 inhibition on Hodgkin's lymphoma cell viability, (b) to test the consequence of HSP90 inhibition on constitutive NF-κB activation in Hodgkin's lymphoma cells with functional or defective IκB proteins, (c) to analyze the consequences of HSP90 inhibition in Hodgkin's lymphoma cells for NK cell–mediated cytotoxicity, and (d) to test the in vivo efficacy of 17-AAG and BIIB021 in a xenograft model of human Hodgkin's lymphoma.

In accordance with previous studies showing high HSP90 affinity of BIIB021,5 we observed BIIB021-mediated inhibition of Hodgkin's lymphoma cell viability at low nanomolar concentrations and additive or even synergistic effects in combination with doxorubicin and gemcitabine. HSP90 inhibition was associated with reduced cellular levels of X-linked inhibitor of apoptosis, a central mediator of apoptosis resistance in Hodgkin's lymphoma cells (27). Notably, our results showed that HSP90 inhibition effectively induced apoptosis in Hodgkin's lymphoma cells but not in healthy donor lymphocytes. These findings are supported by previous reports showing a selective effect of HSP90 inhibition in malignant cells, although to date no comprehensive explanation has been presented for this selectivity (32). A higher affinity of hydroquinone-ansamycins than of quinone-ansamycins toward HSP90 has been proposed (33); however, this explanation does not extend to the tumor selectivity of the purine scaffold BIIB021. Kamal et. al reported that in tumor cells HSP90 is present entirely as constitutively active multichaperone complex, whereas nonmalignant cells harbor HSP90 in a latent uncomplexed state (28). This latter finding could form the basis of the antitumor selectivity of HSP90 inhibitors in Hodgkin's lymphoma.

Recent findings suggested that HSP90 inhibition depletes constitutively active NF-κB in Hodgkin's lymphoma cells with functional IκBs but less effectively in cells with defective IκBs (KM-H2 and L428; refs. 13, 14). Using a highly sensitive ELISA-based NF-κB assay, we detected a decrease in NF-κB activity in response to HSP90 inhibition in all Hodgkin's lymphoma cell lines tested irrespective of IκB defects. This is strongly supported by previous studies from Thomas et al. who reported 15-deoxy-Δ12,14-prostaglandin J2–mediated NF-κB inhibition and induction of apoptosis despite the absence of functional IκBα and IκBϵ in L428 (34). The inhibition of constitutively active NF-κB in presence of defective IκBα or IκBϵ could be mediated by reduced phosphorylation of other IκB proteins, such as IκBβ, substituting the defective IκBs in the inhibition of NF-κB (34). Our findings indicate that targeting NF-κB by inhibition of HSP90 in Hodgkin's lymphoma is feasible in spite of defective mutations in the IκBs, which have been reported to be present in ∼40% of Hodgkin's lymphoma cases (15, 35).

HSP90 plays an essential role in the immune response against malignant cells and inhibition of HSP90 by 17-AAG and geldanamycin has been suggested to have unfavorable effects on host immunity via modulating APC activity (9). However, nothing is known about the effects of HSP90 inhibitors on the susceptibility of the tumor cells against a NK cell attack. Interestingly, we observed that NKG2D-specific ligands were induced in response to 17-AAG- and BIIB021-mediated HSP90 inhibition on Hodgkin's lymphoma cells. This was a rapid effect, which was independent of cell death and growth arrest but accompanied by G2 cell cycle arrest. This is in line with results from Gasser et al., who found an up-regulation of NKG2D ligands in response to cell cycle arrest and stalled DNA replication, which was initiated by ATM (ataxia telangiectasia, mutated) and ATR (ATM and Rad3 related) protein kinases (36). MICA/MICB expression may also be regulated by microRNAs, which maintain NKG2D ligand expression low and enable acute up-regulation of MICA and MICB during cellular stress (37). Whether the up-regulation of NKG2D ligands in response to HSP90 inhibition is induced by transcriptional activation or posttranslational regulation needs to be further investigated.

The induced expression of NKG2D ligands MICA/B/ULBP2 was associated with a dramatic dose-dependent increase in the NK cell–mediated cytotoxicity of L428 Hodgkin's lymphoma cells. Given the documented role of the NKG2D/NKG2DL system for immune surveillance (21, 3840) the induction of NKG2D ligands could have a profound effect on the antitumor response and may counteract the inhibitory effects of HSP90 inhibition on other immune cells. Although Hodgkin's lymphoma cells frequently down-regulate surface expression of MHC-I and should therefore be susceptible to NK cell–mediated immunity, deficiencies in number and function of NK cells have been observed in the blood of Hodgkin's lymphoma patients and in Hodgkin's lymphoma tumors (4144). In accordance, a recent familial genotyping study revealed that individuals with certain alleles of activating KIR NK cell receptors were less likely to develop Hodgkin's lymphoma (45), further strengthening that Hodgkin's lymphoma patients might benefit from therapies that support NK cell activation.

In accordance with the in vitro effects of HSP90 inhibition on Hodgkin's lymphoma cells, we observed a potent antitumor activity in a subcutaneous xenograft Hodgkin's lymphoma model.

Thus, our data show effective direct in vitro and in vivo efficacy of HSP90 inhibition for Hodgkin's lymphoma. The findings indicate that HSP90 inhibition is a promising target in Hodgkin's lymphoma because HSP90 inhibition directly targets Hodgkin's lymphoma cells and also supports the NK cell–mediated antitumor response.

F.J. Burrows and K. Lundgren are shareholders of Biogen Idec. and K. Lundgren is an employee of Biogen Idec. All other authors declare no competing financial interests.

1
Whitesell
L
,
Lindquist
SL
. 
HSP90 and the chaperoning of cancer
.
Nat Rev Cancer
2005
;
5
:
761
72
.
2
Neckers
L
. 
Heat shock protein 90: the cancer chaperone
.
J Biosci
2007
;
32
:
517
30
.
3
Powers
MV
,
Workman
P
. 
Targeting of multiple signalling pathways by heat shock protein 90 molecular chaperone inhibitors
.
Endocr Relat Cancer
2006
;
13 Suppl 1
:
S125
35
.
4
Bagatell
R
,
Gore
L
,
Egorin
MJ
, et al
. 
Phase I pharmacokinetic and pharmacodynamic study of 17-N-allylamino-17-demethoxygeldanamycin in pediatric patients with recurrent or refractory solid tumors: a Pediatric Oncology Experimental Therapeutics Investigators Consortium Study
.
Clin Cancer Res
2007
;
13
:
1783
8
.
5
Banerji
U
,
O'Donnell
A
,
Scurr
M
, et al
. 
Phase I pharmacokinetic and pharmacodynamic study of 17-allylamino, 17-demethoxygeldanamycin in patients with advanced malignancies
.
J Clin Oncol
2005
;
23
:
4152
61
.
6
Goetz
MP
,
Toft
D
,
Reid
J
, et al
. 
Phase I trial of 17-allylamino-17-demethoxygeldanamycin in patients with advanced cancer
.
J Clin Oncol
2005
;
23
:
1078
87
.
7
Grem
JL
,
Morrison
G
,
Guo
XD
, et al
. 
Phase I and pharmacologic study of 17-(allylamino)-17-demethoxygeldanamycin in adult patients with solid tumors
.
J Clin Oncol
2005
;
23
:
1885
93
.
8
Ramanathan
RK
,
Egorin
MJ
,
Eiseman
JL
, et al
. 
Phase I and pharmacodynamic study of 17-(allylamino)-17-demethoxygeldanamycin in adult patients with refractory advanced cancers
.
Clin Cancer Res
2007
;
13
:
1769
74
.
9
Schmitt
E
,
Gehrmann
M
,
Brunet
M
,
Multhoff
G
,
Garrido
C
. 
Intracellular and extracellular functions of heat shock proteins: repercussions in cancer therapy
.
J Leukoc Biol
2007
;
81
:
15
27
.
10
Callahan
MK
,
Garg
M
,
Srivastava
PK
. 
Heat-shock protein 90 associates with N-terminal extended peptides and is required for direct and indirect antigen presentation
.
Proc Natl Acad Sci U S A
2008
;
105
:
1662
7
.
11
Bae
J
,
Mitsiades
C
,
Tai
YT
, et al
. 
Phenotypic and functional effects of heat shock protein 90 inhibition on dendritic cell
.
J Immunol
2007
;
178
:
7730
7
.
12
Georgakis
GV
,
Li
Y
,
Rassidakis
GZ
,
Martinez-Valdez
H
,
Medeiros
LJ
,
Younes
A
. 
Inhibition of heat shock protein 90 function by 17-allylamino-17-demethoxy-geldanamycin in Hodgkin's lymphoma cells down-regulates Akt kinase, dephosphorylates extracellular signal-regulated kinase, and induces cell cycle arrest and cell death
.
Clin Cancer Res
2006
;
12
:
584
90
.
13
Broemer
M
,
Krappmann
D
,
Scheidereit
C
. 
Requirement of Hsp90 activity for IκB kinase (IKK) biosynthesis and for constitutive and inducible IKK and NF-κB activation
.
Oncogene
2004
;
23
:
5378
86
.
14
Janz
M
,
Stuhmer
T
,
Vassilev
LT
,
Bargou
RC
. 
Pharmacologic activation of p53-dependent and p53-independent apoptotic pathways in Hodgkin/Reed-Sternberg cells
.
Leukemia
2007
;
21
:
772
9
.
15
Thomas
RK
,
Re
D
,
Wolf
J
,
Diehl
V
. 
Part I: Hodgkin's lymphoma-molecular biology of Hodgkin and Reed-Sternberg cells
.
Lancet Oncol
2004
;
5
:
11
8
.
16
von Tresckow
B
,
von Strandmann
EP
,
Sasse
S
,
Tawadros
S
,
Engert
A
,
Hansen
HP
. 
Simvastatin-dependent apoptosis in Hodgkin's lymphoma-cells and growth impairment of human Hodgkin's tumors in vivo
.
Haematologica
2007
;
92
:
682
5
.
17
Berglund
M
,
Flordal
E
,
Gullander
J
, et al
. 
Molecular cytogenetic characterization of four commonly used cell lines derived from Hodgkin lymphoma
.
Cancer Genet Cytogenet
2003
;
141
:
43
8
.
18
Emmerich
F
,
Meiser
M
,
Hummel
M
, et al
. 
Overexpression of IκBα without inhibition of NF-κB activity and mutations in the IκBα gene in Reed-Sternberg cells
.
Blood
1999
;
94
:
3129
34
.
19
Emmerich
F
,
Theurich
S
,
Hummel
M
, et al
. 
Inactivating IκBϵ mutations in Hodgkin/Reed-Sternberg cells
.
J Pathol
2003
;
201
:
413
20
.
20
Krappmann
D
,
Emmerich
F
,
Kordes
U
,
Scharschmidt
E
,
Dorken
B
,
Scheidereit
C
. 
Molecular mechanisms of constitutive NF-κB/Rel activation in Hodgkin/Reed-Sternberg cells
.
Oncogene
1999
;
18
:
943
53
.
21
von Strandmann
EP
,
Hansen
HP
,
Reiners
KS
, et al
. 
A novel bispecific protein (ULBP2-4) targeting the NKG2D receptor on natural killer (NK) cells and CD138 activates NK cells and has potent antitumor activity against human multiple myeloma in vitro and in vivo
.
Blood
2006
;
107
:
1955
62
.
22
Heuck
F
,
Ellermann
J
,
Borchmann
P
, et al
. 
Combination of the human anti-CD30 antibody 5F11 with cytostatic drugs enhances its antitumor activity against Hodgkin and anaplastic large cell lymphoma cell lines
.
J Immunother
2004
;
27
:
347
53
.
23
Chou
TC
,
Talalay
P
. 
Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors
.
Adv Enzyme Regul
1984
;
22
:
27
55
.
24
Boll
B
,
Hansen
H
,
Heuck
F
, et al
. 
The fully human anti-CD30 antibody 5F11 activates NF-κB and sensitizes lymphoma cells to bortezomib-induced apoptosis
.
Blood
2005
;
106
:
1839
42
.
25
Eichenauer
DA
,
Simhadri
VL
,
von Strandmann
EP
, et al
. 
ADAM10 inhibition of human CD30 shedding increases specificity of targeted immunotherapy in vitro
.
Cancer Res
2007
;
67
:
332
8
.
26
Fuchs
M
,
Diehl
V
,
Re
D
. 
Current strategies and new approaches in the treatment of Hodgkin's lymphoma
.
Pathobiology
2006
;
73
:
126
40
.
27
Kashkar
H
,
Haefs
C
,
Shin
H
, et al
. 
XIAP-mediated caspase inhibition in Hodgkin's lymphoma-derived B cells
.
J Exp Med
2003
;
198
:
341
7
.
28
Kamal
A
,
Thao
L
,
Sensintaffar
J
, et al
. 
A high-affinity conformation of Hsp90 confers tumour selectivity on Hsp90 inhibitors
.
Nature
2003
;
425
:
407
10
.
29
Re
D
,
Kuppers
R
,
Diehl
V
. 
Molecular pathogenesis of Hodgkin's lymphoma
.
J Clin Oncol
2005
;
23
:
6379
86
.
30
Pogge von Strandmann
E
,
Simhadri
VR
,
von Tresckow
B
, et al
. 
Human leukocyte antigen-B-associated transcript 3 is released from tumor cells and engages the NKp30 receptor on natural killer cells
.
Immunity
2007
;
27
:
965
74
.
31
Scott
GB
,
Meade
JL
,
Cook
GP
. 
Profiling killers; unravelling the pathways of human natural killer cell function
.
Brief Funct Genomic Proteomic
2008
;
7
:
8
16
.
32
Chiosis
G
,
Neckers
L
. 
Tumor selectivity of Hsp90 inhibitors: the explanation remains elusive
.
ACS Chem Biol
2006
;
1
:
279
84
.
33
Maroney
AC
,
Marugan
JJ
,
Mezzasalma
TM
, et al
. 
Dihydroquinone ansamycins: toward resolving the conflict between low in vitro affinity and high cellular potency of geldanamycin derivatives
.
Biochemistry
2006
;
45
:
5678
85
.
34
Thomas
RK
,
Sos
ML
,
Zander
T
, et al
. 
Inhibition of nuclear translocation of nuclear factor-κB despite lack of functional IκBα protein overcomes multiple defects in apoptosis signaling in human B-cell malignancies
.
Clin Cancer Res
2005
;
11
:
8186
94
.
35
Re
D
,
Thomas
RK
,
Behringer
K
,
Diehl
V
. 
From Hodgkin disease to Hodgkin lymphoma: biologic insights and therapeutic potential
.
Blood
2005
;
105
:
4553
60
.
36
Gasser
S
,
Orsulic
S
,
Brown
EJ
,
Raulet
DH
. 
The DNA damage pathway regulates innate immune system ligands of the NKG2D receptor
.
Nature
2005
;
436
:
1186
90
.
37
Diefenbach
A
,
Jensen
ER
,
Jamieson
AM
,
Raulet
DH
. 
Rae1 and H60 ligands of the NKG2D receptor stimulate tumour immunity
.
Nature
2001
;
413
:
165
71
.
38
Strid
J
,
Roberts
SJ
,
Filler
RB
, et al
. 
Acute upregulation of an NKG2D ligand promotes rapid reorganization of a local immune compartment with pleiotropic effects on carcinogenesis
.
Nat Immunol
2008
;
9
:
146
54
.
39
Cerwenka
A
,
Baron
JL
,
Lanier
LL
. 
Ectopic expression of retinoic acid early inducible-1 gene (RAE-1) permits natural killer cell-mediated rejection of a MHC class I-bearing tumor in vivo
.
Proc Natl Acad Sci U S A
2001
;
98
:
11521
6
.
40
Friese
MA
,
Platten
M
,
Lutz
SZ
, et al
. 
MICA/NKG2D-mediated immunogene therapy of experimental gliomas
.
Cancer Res
2003
;
63
:
8996
9006
.
41
Frydecka
I
. 
Natural killer cell activity during the course of disease in patients with Hodgkin's disease
.
Cancer
1985
;
56
:
2799
803
.
42
Frydecka
I
,
Bocko
D
,
Kosmaczewska
A
,
Ciszak
L
,
Morilla
R
. 
The effect of peripheral blood lymphocyte stimulation on ζ chain expression and IL-2 production in Hodgkin's disease
.
Br J Cancer
2001
;
84
:
1339
43
.
43
Konjevic
G
,
Jurisic
V
,
Banicevic
B
,
Spuzic
I
. 
The difference in NK-cell activity between patients with non-Hodgkin's lymphomas and Hodgkin's disease
.
Br J Haematol
1999
;
104
:
144
51
.
44
Alvaro-Naranjo
T
,
Lejeune
M
,
Salvado-Usach
MT
, et al
. 
Tumor-infiltrating cells as a prognostic factor in Hodgkin's lymphoma: a quantitative tissue microarray study in a large retrospective cohort of 267 patients
.
Leuk Lymphoma
2005
;
46
:
1581
91
.
45
Besson
C
,
Roetynck
S
,
Williams
F
, et al
. 
Association of killer cell immunoglobulin-like receptor genes with Hodgkin's lymphoma in a familial study
.
PLoS ONE
2007
;
2
:
e406
.

Competing Interests

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.