Topoisomerase II Poisoning and Mitochondrial ROS Coupling to Inflammasome Priming
Doxorubicin functions through intercalation into planar genomic base pairs and stabilization of Topoisomerase II-DNA cleavage complexes, impeding religation and inducing double-strand DNA breaks marked by Ser139 phosphorylated histone γ-H2AX [1][2]. Parallel to nuclear genotoxicity, the anthracycline quinone moiety undergoes single-electron reduction via mitochondrial Complex I and NAD(P)H dehydrogenases [1][3]. Subsequent oxidation of the unstable semiquinone radical generates continuous cascades of superoxide anions (O₂^{\bullet-}) and hydrogen peroxide (H₂O₂), promoting mitochondrial outer membrane permeabilization [1]. The resultant loss of inner mitochondrial transmembrane potential leads to cytosolic extrusion of oxidized mitochondrial DNA (mtDNA) and purine catabolites [1][2]. Cytosolic mtDNA engages sensor machineries that stimulate NF-κB transcription, directing the transcriptional synthesis of resting pro-IL-1β, pro-IL-18, and the sensor protein NLRP3 [1][3].
To rigorously benchmark these cellular effects against alternate chemotherapeutic classes, comparative pharmacology profiles are established across standard cytotoxicity, radical generation, and death modalities as outlined in Table 1.
Empirical investigations in human leukemia THP-1 and breast adenocarcinoma MDA-MB-231 models show that doxorubicin-induced reactive oxygen species (ROS) disrupt antioxidant defenses, inactivating Sirt1 and downregulating thioredoxin systems [1][3]. This continuous oxidative tension induces post-translational modifications of NLRP3, comprising dephosphorylation at Ser295 and targeted deubiquitination [1]. These modifications license the sensor for macromolecular assembly. Consequently, Topoisomerase II poisoning acts synergistically with electron transport chain uncoupling, transforming homeostatic cancer cells into immunogenically primed substrates primed for lytic programmed cell death cascades [2][3].
NLRP3-ASC Macromolecular Nucleation and Canonical Caspase-1 Activation
The initiation of canonical pyroptosis requires the assembly of the tripartite NLRP3 inflammasome complex, governed by intracellular ionic perturbations [1][2]. Doxorubicin treatment compromises plasma membrane integrity and alters membrane transport channels, provoking rapid intracellular potassium (K+) efflux [1]. Intracellular K+ drop below the threshold of 90 mM relieves allosteric inhibition across the NLRP3 NACHT domain, exposing its central nucleotide-binding and oligomerization domain [1][2]. In the presence of dATP, NLRP3 undergoes self-oligomerization into an active wheel-like heptameric ring, orienting its pyrin domains (PYD) outward [2].
Exposed PYD motifs nucleate the polymerization of the adaptor protein ASC (apoptosis-associated speck-like protein containing a CARD) via homotypic PYD-PYD interactions [1][2]. Fluorescent biosensor platforms, such as CASPorter assays engineered in HEK-293 reporter systems, demonstrate that doxorubicin initiates rapid nucleocytoplasmic redistribution of ASC into dense, micron-scale perinuclear ASC specks [2]. Pharmacological inhibition using the NLRP3 blocker MCC950 or the caspase-1 inhibitor VX-765 dose-dependently impedes this oligomerization cascade [2].
Once assembled, filamentous ASC filaments present clustered caspase recruitment domains (CARD) that interface directly with the N-terminal CARD of pro-caspase-1 [1][2]. This homotypic CARD-CARD condensation enforces high-density pro-caspase-1 proximity, driving catalytic homodimerization [1][2]. The clustered zymogens undergo auto-proteolytic processing across two defined interdomain linker sites: cleavage at Asp297 releases the catalytic p20 subunit, while secondary cleavage at Asp316 separates the p10 regulatory subunit from the prodomain [1][2]. The resulting heterotetrameric p20-p10 complex constitutes the enzymatically active caspase-1 protease, prepared to cleave downstream inflammatory substrates and execute membrane permeabilization [1][3].
Gasdermin D Proteolysis, Membrane Insertion, and Pyroptotic Pore Assembly
Following catalytic assembly, active caspase-1 targets the executioner gasdermin family, primarily Gasdermin D (GSDMD) [1][3]. Under resting conditions, GSDMD exists in an auto-inhibited state where its C-terminal repressor domain (GSDMD-CT) sterically masks the lethal N-terminal functional domain (GSDMD-NT) [1]. Caspase-1 executes specific proteolytic cleavage within the central interdomain linker at residue Asp275, liberating the 31-kDa GSDMD-NT executioner fragment from its 22-kDa inhibitory C-terminal partner [1][3]. Uncoupled GSDMD-NT exhibits exposed basic patches that display stereospecific affinity for negatively charged phospholipids restricted to the inner plasma membrane leaflet, specifically phosphatidylserine and phosphoinositides [1].
Upon electrostatically docking to the inner leaflet, GSDMD-NT inserts a four-stranded β-hairpin into the lipid bilayer, initiating conformational changes that direct higher-order ring oligomerization [1][3]. Between 16 and 24 GSDMD-NT protomers coalesce into a hollow, symmetrical transmembrane conduit featuring an inner pore diameter of approximately 10 to 18 nm [1]. This hydrophilic pore breaches cellular ionic selectivity, dissipating transmembrane electrical potentials and permitting unconstrained bi-directional trafficking of ions, water, and osmotic solutes below 15 to 20 kDa in mass [1].
Driven by intracellular oncotic pressure, massive water influx produces rapid cellular swelling characterized by balloon-like, spherical phase-lucent bubbles [1]. Unlike the membrane blebbing and apoptotic vesicle generation driven by actomyosin contractility during apoptosis, pyroptotic tumefaction concludes in physical membrane tearing and complete lytic collapse [1][4]. Quantification in murine C1498 myeloid cells and primary patient-derived blasts reveals that GSDMD pore formation initiates profound lactate dehydrogenase (LDH) release within 12 to 24 hours of exposure [4][5]. This lytic collapse exposes intracellular components directly to the extracellular environment [1][5].
Bifunctional DAMP Emission: Convergence of Canonical ICD with IL-1β and IL-18 Secretion
The catastrophic disruption of the plasma membrane by GSDMD channels establishes a dual immunogenic secretome, integrating active cytokine release with danger-associated molecular pattern (DAMP) extrusion [1][4]. Concurrently with GSDMD cleavage, catalytic caspase-1 processes cytosolic pro-IL-1β and pro-IL-18 at specific aspartate cleavage residues, generating mature, bioactive 17-kDa IL-1β and 18-kDa IL-18 cytokines [1][3]. Because these mature cytokines lack classic endoplasmic reticulum-Golgi secretory leader sequences, they rely on the dilated conduit of GSDMD pores for passive release into the interstitial tumor microenvironment [1][4].
Simultaneously, anthracycline-induced endoplasmic reticulum stress promotes the translocation of the chaperone calreticulin (CRT) to the outer cell membrane, presenting an early "eat-me" signal for phagocytic internalization [5][6]. As the plasma membrane ruptures under osmotic swelling, high-mobility group box 1 (HMGB1) translocates from chromatin to the extracellular space, while intracellular ATP rapidly discharges through open pores [4][5]. Extracellular ATP acts via purinergic P2RX7 receptors to recruit inflammatory phagocytes, whereas HMGB1 ligates Toll-like receptor 4 (TLR4) on local immune populations [4][5].
Murine models utilizing B16F10 melanoma and breast carcinoma show that this simultaneous release of inflammatory cytokines and DAMPs converts immunologically cold microenvironments into highly active hubs [4][7]. Infiltrating conventional type I dendritic cells (cDC1), identified by CD103+ and XCR1+ surface phenotypes, take up tumor antigens and undergo maturation through IL-1β-driven licensing [4][5]. These mature cDC1s traffic efficiently to draining lymph nodes and drive robust cross-presentation to naïve CD8+ T cells [4]. Furthermore, doxorubicin elevates tertiary lymphoid structure (TLS) formation, marked by peripheral node addressin-positive high endothelial venules (PNAd+ HEV) [7]. The resulting effector cytotoxic CD8+ T cells demonstrate elevated tumor infiltration and sustained cytotoxic interferon-gamma production, producing protective systemic antitumor immunity [4][6][7].
Cross-Regulation Between Apoptotic Caspase-3/8 Cascades and Gasdermin Processing
The therapeutic outcome of doxorubicin administration is dictated by an intricate balance between classic apoptotic execution and lytic pyroptosis [4][8]. Exposure to escalating anthracycline concentrations reveals a concentration-dependent threshold where cell fate bifurcates between apoptotic caspase-3/8 cascades and caspase-1-driven pyroptosis [8]. When the apoptotic program predominates, activated caspase-3 cleaves full-length GSDMD at residue Asp87 within its functional N-terminal domain [8]. This secondary processing releases an inactive 13-kDa fragment, dismantling the pore-forming capacity of GSDMD and serving as a molecular damper on pyroptotic lysis [8].
Conversely, tumor cells exhibiting low or mutated Caspase-1/GSDMD expression frequently channel doxorubicin-induced stress toward Gasdermin E (GSDME, also known as DFNA5) [8]. Activated caspase-3 directly hydrolyzes GSDME at Asp270, liberating a 34-kDa GSDME N-terminal fragment (GSDME-NT) [8]. Similar to GSDMD, GSDME-NT oligomerizes in the plasma membrane to assemble lytic pores, executing secondary pyroptosis [8]. In MCF-7 breast cancer cells and multiple myeloma models, doxorubicin promotes both apoptotic marker expression and GSDME-mediated membrane perforation, proving that death pathways can operate cooperatively [3][8].
Modulation of survival proteins further modifies this crosstalk. Doxorubicin treatment diminishes levels of c-FLIP and Mcl-1 in triple-negative BT-549 cells, enhancing susceptibility to extrinsic death signals and perforin-dependent lysis [9]. In addition, while doxorubicin stimulates robust cytotoxic T-cell recruitment, prolonged inflammatory signaling induces adaptive immune resistance by upregulating PD-L1 on non-leukocytic tumor cells and PD-1 on CD8+ T cells [7]. Consequently, combination approaches coupling liposomal doxorubicin with immune checkpoint blockers such as anti-PD-1 or anti-CTLA-4 reverse T-cell exhaustion and prevent tumor recurrence [6][7]. Interrogating the balance between apoptotic silencing and pyroptotic amplification remains essential to optimizing anthracycline immunogenicity without triggering off-target inflammatory toxicities [1][4][8].