The success of small-molecule inhibitors targeting the BCR-ABL fusion kinase has been difficult to replicate in more common types of cancers with heterogeneous molecular abnormalities, such as diffuse large B-cell lymphoma (DLBCL). The original classification of DLBCL into germinal center B-cell (GCB) and activated B-cell (ABC) subtypes, based on gene expression profiling, still has a physiological basis but has been supplanted by newer schemes with more categories,1 based on diverse but recurrent abnormalities. However, specific inhibitors targeting certain of these abnormalities or related pathways have not yet achieved Food and Drug Administration approval in DLBCL, even for patients displaying the targeted abnormality. Might an empiric approach be a complementary alternative for finding (or repurposing) effective drugs? Could 1 drug target different types of DLBCL in different ways? In this issue of Blood, Schmitt and colleagues2 provide affirmative answers to both questions.

Dimethyl fumarate (DMF), a simple compound that is already approved for 2 diseases mediated by activated lymphocytes (multiple sclerosis and psoriasis), was tested in vitro against diverse cell lines. Efficacy was frequent against lines representing both ABC and GCB subtypes of DLBCL, and against mantle cell lymphoma (MCL), but not against myeloid, carcinoma, or melanoma types. The authors then did extensive work on the mechanism of action of DMF in DLBCL lines and found differences between the subtypes (see figure).

DMF profoundly depleted glutathione in GCB-DLBCL and MCL lines, apparently through direct succinylation. Because glutathione is critical for cellular redox balance and neutralization of ROS, its depletion led to ferroptosis, a caspase-independent form of programmed cell death in which iron-dependent lipid peroxidation causes loss of plasma membrane integrity.3  Molecular features of ferroptosis were observed with DMF treatment, and multiple inhibitors of ferroptosis protected GCB-DLBCL lines from DMF, confirming its causal role. DMF-induced ferroptosis was facilitated by the ROS-generating, iron-containing enzyme arachidonate 5-LOX, expressed in the nucleus (as required for its activity) in GCB-DLBCL cell lines and 50% of primary tumors.

DMF has different mechanisms of action in different subtypes of DLBCL. In the GCB subtype (left), direct succinylation of glutathione (GSH) by DMF leads to GSH depletion, impairing the ability of glutathione peroxidase 4 (GPX4) to detoxify polyunsaturated fatty acids (PUFAs) that have undergone peroxidation. Such damage to plasma membrane lipids results from reactive oxygen species (ROS), produced by enzymes such as arachidonate 5-lipoxygenase (5-LOX), and causes ferroptosis, a caspase-independent form of programmed cell death. DMF does not cause ferroptosis in the ABC subtype of DLBCL (right), but instead inhibits 2 of its distinctive prosurvival signaling pathways. Transcription of target genes by canonical nuclear factor κB (NF-κB), one of the key consequences of “chronic active” signaling by the B-cell receptor (BCR) in ABC-DLBCL, depends on activation by the IKK2 kinase; similarly, activation of the STAT3 transcription factor depends on the kinase JAK1, downstream of autocrine signaling by interleukin-6 (IL-6) and IL-10. DMF inactivates both of these kinases by direct succinylation of critical cysteine residues. See the visual abstract in the article by Schmitt et al that begins on page XXX.

DMF has different mechanisms of action in different subtypes of DLBCL. In the GCB subtype (left), direct succinylation of glutathione (GSH) by DMF leads to GSH depletion, impairing the ability of glutathione peroxidase 4 (GPX4) to detoxify polyunsaturated fatty acids (PUFAs) that have undergone peroxidation. Such damage to plasma membrane lipids results from reactive oxygen species (ROS), produced by enzymes such as arachidonate 5-lipoxygenase (5-LOX), and causes ferroptosis, a caspase-independent form of programmed cell death. DMF does not cause ferroptosis in the ABC subtype of DLBCL (right), but instead inhibits 2 of its distinctive prosurvival signaling pathways. Transcription of target genes by canonical nuclear factor κB (NF-κB), one of the key consequences of “chronic active” signaling by the B-cell receptor (BCR) in ABC-DLBCL, depends on activation by the IKK2 kinase; similarly, activation of the STAT3 transcription factor depends on the kinase JAK1, downstream of autocrine signaling by interleukin-6 (IL-6) and IL-10. DMF inactivates both of these kinases by direct succinylation of critical cysteine residues. See the visual abstract in the article by Schmitt et al that begins on page XXX.

In contrast, ABC-DLBCL lines were relatively resistant to DMF-induced ferroptosis, attributable to 2 effects of its “chronic active” form of B-cell receptor signaling: suppression of 5-LOX transcription and constitutive activity of the canonical NF-κB pathway. However, DMF was toxic to ABC-DLBCL lines by other mechanisms, including at least the inhibition of NF-κB and JAK/STAT3, another of the essential signaling pathways that are distinctive of ABC-DLBCL. Extensive analyses pinpointed the cause of these inhibitory effects to direct succinylation of critical cysteine residues of enzymes that activate these pathways, C179 of IKK-β and C257 of JAK1.

Further testing of DMF efficacy included showing that it reduced tumor formation by GCB-DLBCL lines in zebrafish embryos. In vitro combination testing showed synergism between DMF and an inhibitor of the glutathione-independent ferroptosis suppressor protein 1, as well as with the BCL2 inhibitor ABT-199 in ABC-DLBCL lines. Finally, DMF and ABT-199 were more effective when used together in vivo, substantially reducing the growth in mice of an ABC-DLBCL line (HBL-1) and a patient-derived xenograft (VFN-D1). However, whether the same molecular effects of DMF observed in vitro were also active in these in vivo settings was not examined.

This study challenges us to consider what “targeted therapy” really means. All drugs have targets, whether we know what they are or not; what is important is whether those targets are critical to tumor cells, more than to normal cells. DMF is not a targeted therapy in the usual sense; it is likely, although not investigated in this study, that numerous sulfhydryl-containing proteins and perhaps other molecules (like glutathione) are succinylated by DMF. In effect, empiric DMF treatment coupled with subsequent target identification in this study functioned as a screen; from a hypothetical “library” of succinylated molecular species that was created in cells, those that are critical targets, at least within the context of DMF treatment overall, were identified. Functional specificity of DMF’s effects was shown by finding only a few critical targets in DLBCL lines, differing between subtypes, and apparently none in other cell types.

For 1 drug to have different mechanisms in different DLBCL subtypes has been seen before, with an inhibitor of the NEDD8-activating enzyme (MLN4924, Pevonedistat), a drug that has a precise direct molecular target that affects multiple proteins.4  Similarly, lenalidomide and related immunomodulatory imide drugs have precise direct targets but pleiotropic and cell-specific effects on multiple proteins, affecting tumor cells (including ABC-DLBCL5 ) and immune cells in different but therapeutically beneficial ways.

The most important question is whether these findings can be successfully translated into therapy for DLBCL. At first glance, that DMF is already an approved and well-tolerated drug for certain nonneoplastic diseases should make this more plausible; however, there are many unknowns. Achieving sufficient DMF levels to treat DLBCL is likely to require injection, according to the authors, whereas DMF is currently only approved for oral use; therefore, dosing and administration will need to be determined. That bears on a second and larger concern: especially at levels needed for DLBCL, will the beneficial tumor cell-intrinsic effects of DMF be outweighed by its immunosuppressive effects,6  which coincide with ones generally found in tumor microenvironments? This is countered by other, more encouraging evidence: ferroptosis induction is important for tumor-cell killing by T cells,7  at least during treatment with checkpoint inhibitors, and could be enhanced by DMF; furthermore, ferroptosis makes tumor cells immunogenic,8  which also suggests that short treatments with DMF might be more effective than longer ones. Uncertain for predicting DMF’s efficacy is that it has been found in immune cells to suppress aerobic glycolysis,9  essential for effector T-cell function, as well as in tumor cells.10 

Further evaluation of these promising results, in immunocompetent models or clinical trials, will be necessary to determine whether “succinylation therapy” will become a new front in the war on DLBCL.

Conflict-of-interest disclosure: The author declares no competing financial interests.

REFERENCES

1.
Wright
GW
,
Huang
DW
,
Phelan
JD
, et al
.
A probabilistic classification tool for genetic subtypes of diffuse large B cell lymphoma with therapeutic implications
.
Cancer Cell.
2020
;
37
(
4
):
551
-
568
.
2.
Schmitt
A
,
Xu
W
,
Bucher
P
, et al
.
Dimethyl fumarate induces ferroptosis and impairs NF-κB/STAT3 signaling in DLBCL
.
Blood.
2021
;
138
(
10
):
871
-
884
.
3.
Stockwell
BR
,
Jiang
X
,
Gu
W.
Emerging mechanisms and disease relevance of ferroptosis
.
Trends Cell Biol.
2020
;
30
(
6
):
478
-
490
.
4.
Milhollen
MA
,
Traore
T
,
Adams-Duffy
J
, et al
.
MLN4924, a NEDD8-activating enzyme inhibitor, is active in diffuse large B-cell lymphoma models: rationale for treatment of NF-κB-dependent lymphoma
.
Blood.
2010
;
116
(
9
):
1515
-
1523
.
5.
Yang
Y
,
Shaffer
AL
III
,
Emre
NC
, et al
.
Exploiting synthetic lethality for the therapy of ABC diffuse large B cell lymphoma
.
Cancer Cell.
2012
;
21
(
6
):
723
-
737
.
6.
Mills
EA
,
Ogrodnik
MA
,
Plave
A
,
Mao-Draayer
Y.
Emerging understanding of the mechanism of action for dimethyl fumarate in the treatment of multiple sclerosis
.
Front Neurol.
2018
;
9
:
5
.
7.
Wang
W
,
Green
M
,
Choi
JE
, et al
.
CD8+ T cells regulate tumour ferroptosis during cancer immunotherapy
.
Nature.
2019
;
569
(
7755
):
270
-
274
.
8.
Efimova
I
,
Catanzaro
E
,
Van der Meeren
L
, et al
.
Vaccination with early ferroptotic cancer cells induces efficient antitumor immunity
.
J Immunother Cancer.
2020
;
8
(
2
):
e001369
.
9.
Kornberg
MD
,
Bhargava
P
,
Kim
PM
, et al
.
Dimethyl fumarate targets GAPDH and aerobic glycolysis to modulate immunity
.
Science.
2018
;
360
(
6387
):
449
-
453
.
10.
Chen
K
,
Wu
S
,
Ye
S
, et al
.
Dimethyl fumarate induces metabolic crisie to suppress pancreatic carcinoma
.
Front Pharmacol.
2021
;
12
:
617714
.