Dimethyl fumarate restores apoptosis sensitivity and inhibits tumor growth and metastasis in CTCL by targeting NF-κB

Abstract

Despite intensive efforts in recent years, a curative therapy for cutaneous T-cell lymphoma (CTCL) has not yet been developed. Therefore, the establishment of new therapeutic approaches with higher efficacy rates and milder side effects is strongly desired. A characteristic feature of the malignant T-cell population in CTCL is resistance toward cell death resulting from constitutive NF-κB activation. Therefore, NF-κB-dependent cell death resistance represents an interesting therapeutic target in CTCL because an NF-κB-directed therapy would leave bystander T cells widely unaffected. We investigated the effects of dimethyl fumarate (DMF) on CTCL cells in vitro and in vivo. DMF induced cell death in primary patient-derived CD4(+) cells and CTCL cell lines, but hardly in T cells from healthy donors. DMF-induced cell death was linked specifically to NF-κB inhibition. To study the impact of DMF in vivo, we developed 2 CTCL xenograft mouse models with different cutaneous localizations of the T-cell infiltrate. DMF treatment delayed the growth of CTCL tumors and prevented formation of distant metastases. In addition, DMF induced increased cell death in primary CTCL tumors and in liver metastases. In summary, DMF treatment represents a remarkable therapeutic option in CTCL because it restores CTCL apoptosis in vitro and in preclinical models in vivo and prevents spreading of the disease to distant sites. DMF treatment is of particular promise in CTCL because DMF is already in successful clinical use in the treatment of psoriasis and multiple sclerosis allowing fast translation into clinical studies in CTCL.

Figure 1

DMF causes cell death in primary CTCL cells and CTCL cell lines. (A) Specific cell death in primary CD4⁺ cells isolated from 10 healthy volunteers (control) and 10 patients with Sézary syndrome (patient) upon treatment with 30 µM DMF (left) or 30 µM MMF (right) for 48 hours. Triangles, single samples; gray bars, median. (B) Specific cell death rates in primary CD4⁺ cells isolated from 10 healthy volunteers (control) and 10 patients with Sézary syndrome (patient) upon treatment with different concentrations of DMF for 48 hours. (C) Specific cell death rates in J16, HH, and SeAx cells upon treatment with different concentrations of DMF solubilized in dimethyl sulfoxide for 24 hours (n = 4, each). (D) Decrease in tetramethylrhodamine ethyl ester (TMRE) mean fluorescence intensity in J16, HH, and SeAx cells upon treatment with either 30 µM DMF or MMF or 10 µM CCCP for 24 hours (n = 4, each). *P < .05. (E) Western blot analysis of caspase 3 cleavage in HH (left) and SeAx (right) cells after treatment with 50 µM DMF for the indicated time points. (F) Single cell gel electrophoresis of HH (left panels) and SeAx (right panels) cell upon treatment with either vehicle or 50 µM DMF for 8 hours.

Figure 2

DMF inhibits NF-κB activity in CTCL cells. (A) Specific luminescence of the NF-κB luciferase assay in J16 and HH cells 24 hours after transfection without treatment (n = 3, each). (B) Specific luminescence of the NF-κB luciferase assay in J16 and HH cells 24 hours after transfection with and without treatment with 50 µM DMF for 24 hours. The specific NF-κB activity of untreated J16 cells was normalized to 1 and all other activities were normalized to that value (n = 3 each). (C) Relative IκBα expression measured by qRT-PCR in SeAx (upper) and HH (lower) cells 1 to 3 hours after treatment with 50 µM of either DMF or MMF, normalized to GAPDH (n = 3 each). (D) TNF-α concentration in supernatants of HH cells either untreated or treated with 50 µM DMF for 12 hours (n = 3 each). (E) Normalized nuclear p65 activity measured by binding ELISA in HH and SeAx cells upon treatment with different concentrations of DMF and MMF for 1 hour (upper) and 16 hours (lower), (n = 3 each). DMSO, dimethyl sulfoxide. (F) Relative changes in expression levels of NF-κB target genes measured by qRT-PCR array after 3 hours of treatment with DMF and MMF (30 µM each). The quantitative values are (expression [DMF treatment])/(expression(MMF treatment)). (n = 3, each) *P < .05.

Figure 3

DMF treatment inhibits CTCL cell and tumor growth in subcutaneous and orthotopic xenograft mouse models. NSG mice were xenografted with HH cells intradermally and treated once daily with either 30 mg/kg bodyweight of DMF or PBS by IP injection. (n = 37 each). In the subcutaneous model, the NSG mice were xenografted with HH cells subcutaneously and treated once daily with either 20 mg/kg bodyweight of DMF or PBS orally by gavage (n = 20 each). (A) Macroscopic pictures of a representative primary HH tumor of the respective mouse from either a PBS-treated (upper) or a DMF-treated (lower) animal in the intradermal model. (B) Survival curves of NSG mice. Decrease in survival is either caused by spontaneous death or by reaching critical tumor size of 1.5 cm at the largest diameter. (C) Percent survival rate of orthotopically xenografted mice treated with PBS or DMF at day 29, the end of the treatment phase. (D) Median time of tumor growth in PBS- and DMF-treated intradermal CTCL xenograft mice from first detection of a tumor to either death or the end of the experiment. (E) Median tumor volume of PBS- and DMF-treated intradermal CTCL xenograft mice over time. (F) Median tumor volume of subcutaneous HH xenografts of mice treated with either 20 mg/kg bodyweight DMF or PBS by oral gavage (n = 20, each) over time. (G) Median tumor volume in PBS- and DMF-treated subcutaneous CTCL xenografts at day 20. *P < .05.

Figure 4

DMF treatment induces massive cell death specifically within CTCL tumors in vivo. HH xenografted tumors grown intradermally in NSG mice that were treated once daily IP with either 30 mg/kg bodyweight DMF or PBS. (A) Representative hematoxylin and eosin–stained specimens of primary HH tumors (upper panels, 20×; lower panels, 200×). (B) Representative pictures of immunofluorescent stainings of primary HH tumors stained for cleaved caspase 3, CD3, and with Hoechst dye, the latter to counterstain nuclei. (C) Semiquantitative score of necrosis areas in the primary xenograft tumors (0% to 25% tumor area covered by necrosis = 0; 25% to 50% tumor area covered by necrosis = 1; 50% to 75% tumor area covered by necrosis = 2; 75% to 100% tumor area covered by necrosis = 3). (D) Quantification of cleaved caspase 3 (left) and cleaved caspase 8 (right) mean fluorescence intensity in HH tumors of either the PBS or DMF treatment group (n = 3 each). (E) Representative immunofluorescent pictures of primary HH xenograft tumors of a PBS- and a DMF-treated mouse stained for cl Casp3, p65, and with Hoechst dye. *P < .05.

Figure 5

DMF treatment inhibits metastasis of CTCL tumors in vivo. Livers were collected from orthotopically HH xenografted tumor bearers that were treated once daily IP with either 30 mg/kg bodyweight DMF or PBS as control (n = 37, each). (A) Representative macroscopic pictures of an explanted liver of the respective mouse from either a PBS-treated (upper) or a DMF-treated (lower) animal in the intradermal model. (B) Representative pictures of hematoxylin and eosin–stained liver specimens (upper panels, 20×; lower panels, 200×). (C) Semiquantitative score of HH T-cell infiltrate areas within the mouse livers (0% to 25% liver area covered by T cells = 0; 25% to 50% liver area covered by T cells = 1; 50% to 75% liver area covered by T cells = 2; 75% to 100% liver area covered by T cells = 3). (D) Median of the semiquantitative score of HH-T cell infiltrates in liver of DMF- and PBS-treated HH xenograft tumor bearers. (E) Representative picture of liver specimens derived from DMF- and PBS-treated HH xenograft tumor bearers immunohistochemically stained for cleaved caspase 3. Note the positive staining in T-cell foci of DMF-treated tumor bearers, but not in surrounding host tissue. *P < .05.