Rationale:Dysregulated phagocytic clearance within the synovial microenvironment contributes to persistent inflammation and impairs immune homeostasis in rheumatoid arthritis (RA). Because MerTK-positive macrophages are essential for efferocytosis and inflammation resolution, but are functionally impaired in RA, this study aimed to engineer macrophages capable of enhancing tumor necrosis factor-α (TNF-α) clearance and restoring inflammation-resolving macrophage function. Methods:We generated anti-TNFα chimeric antigen receptor macrophages (CAR-M) by replacing the MerTK antigen-binding domain with a TNF-R1 fragment. The phagocytic and degradative capacity of CAR-M toward soluble TNF-α was assessed, together with activation of Rho GTPases Rac1, Cdc42, and RhoA. Macrophage inflammatory mediator production, phenotypic polarization, SOCS1/3, NF-κB, and MAPK signaling were analyzed. The effects of CAR-M cells were further compared with adalimumab and control treatments under synovial fluid stimulation from patients with arthritis and in mice with collagen-induced arthritis (CIA). Results:Anti-TNFα CAR-M showed markedly enhanced phagocytosis and degradation of soluble TNF-α, accompanied by Rac1, Cdc42, and RhoA activation. CAR activation reduced inflammatory mediator production, including soluble TNF-α, interleukin-1 β (IL-1β), and interleukin-6 (IL-6), while increasing interleukin-10 (IL-10) secretion. This functional shift was associated with polarization toward an inflammation-resolving MerTK+CD206+ phenotype, SOCS1/3 upregulation, and NF-κB and MAPK pathway inhibition. Under stimulation with synovial fluid from arthritis patients, anti-TNFα CAR-M decreased TNF-α and IL-6 levels and drove macrophages toward an M2-like anti-inflammatory state. In CIA mice, local in situ CAR-M cell treatment produced stronger therapeutic effects than adalimumab or other controls, significantly alleviating inflammatory activation, clinical symptoms, and joint pathologies. Conclusions:Anti-TNFα CAR-M exert therapeutic effects via a dual mechanism: direct TNF-α clearance and synovial macrophage reprogramming toward a reparative phenotype. This synthetic biology strategy highlights the potential of precise cell engineering to correct phagocytic deficits and resolve chronic inflammation in RA.
Rationale:The blood-brain barrier (BBB) remains a major obstacle to the delivery of therapeutics for central nervous system (CNS) diseases. Although several BBB-penetrating or BBB-bypassing strategies have been investigated, there remains a need for sustained, controllable delivery routes that can support prolonged CNS exposure of BBB-impermeable agents. Recent anatomical studies have shown direct connections between skull bone marrow and brain, suggesting that the skull may serve as an alternative access route to the brain. Here, we assessed the feasibility of sustained intracalvariosseous infusion (ICO) as skull-to-brain delivery approach for BBB-impermeable molecular and nanoscale agents in mice and rabbits. Methods:Four-week ICO was performed using a species-adapted nano-flow pump-cannula configuration that positioned the cannula terminus within the skull diploic space. A positive comparator was established by positioning the terminus beyond the inner skull cortex. Placement was verified by conventional CT and micro-CT. Brain-associated exposure of BBB-impermeable paclitaxel, antisense oligonucleotides and gold nanoparticles was assessed, along with systemic toxicity, bone marrow-derived immune activation, and neuroinflammation. Results:ICO produced measurable brain-associated exposure of all three model agents in both species. Relative to the positive comparator, ICO achieved the highest relative brain-associated exposure for ASO, reaching approximately 24-28% of comparator levels, whereas PTX and AuNP showed lower but detectable exposure. Four-week ICO was not associated with overt hematological, biochemical, histopathological, or neuroinflammatory abnormalities under the present experimental conditions. Conclusions:These findings support the technical feasibility of ICO-based sustained infusion as a skull-to-brain delivery platform for controlled and prolonged CNS exposure of BBB-impermeable molecular and nanoscale agents through an extracerebral skull compartment.
Rationale:Non-small cell lung cancer (NSCLC) develops a high GSH/GPX4 antioxidant phenotype under persistent oxidative pressure, which suppresses membrane lipid peroxidation and ferroptosis, constituting a core mechanism underlying chemotherapy resistance and suboptimal therapeutic efficacy. Breaking this resistance barrier demands not simply attacking the tumor but disarming the antioxidant defense to intensify oxidative damage and awaken durable antitumor immunity. Methods:A self-assembled nano-prodrug, E-R@ISSL, was engineered by co-assembling a carbamate-linked RGD-modified etoposide (ETP) prodrug and a disulfide-bonded indole derivative conjugated to linoleic acid (LA) at an optimal 1:2 molar ratio. The system thus possesses dual αvβ5/mitochondria targeting with CES2/GSH cascade-responsive drug release. Antitumor activity and mechanisms were evaluated in vitro and in both subcutaneous and orthotopic NSCLC mouse models. Results:E-R@ISSL exhibited efficient αvβ5-mediated internalization and mitochondria-targeted delivery. Upon sequential CES2/GSH-triggered disassembly, E-R@ISSL co-released ETP and LA, reduced the available intracellular GSH pool, and functionally dampened the GSH-dependent GPX4 antioxidant defense. ETP induced DNA double-strand breaks and elevated reactive oxygen species (ROS), while LA expanded the oxidizable lipid pool, synergistically driving lipid peroxidation and ferroptosis. Concurrently, the system activated immunogenic cell death (ICD), promoting dendritic cell maturation and enhancing CD8+ T cell infiltration. In orthotopic NSCLC models, E-R@ISSL significantly suppressed tumor progression, prolonged survival, and demonstrated a favorable safety profile with reduced systemic toxicity compared with free ETP. Conclusions:Together, these findings demonstrate that E-R@ISSL achieved coordinated DNA damage, ferroptosis, and ICD induction through its dual-targeted and sequentially activated co-delivery of ETP and LA, concurrently dismantling the GSH/GPX4 antioxidant defense and intensifying oxidative injury. This nano-prodrug thus represents a viable strategy for overcoming treatment resistance and extending durable antitumor responses in NSCLC.
Background:Three-dimensional (3D) visualization of the renal vasculature plays a crucial role in the pathological assessment and mechanistic investigation of kidney diseases. However, existing 3D imaging approaches remain unable to rapidly and accurately achieve continuous visualization spanning from millimeter-scale arteries to micrometer-scale glomerular capillaries. Here, we developed an innovative strategy, termed EBDE clearing, which enables continuous 3D visualization from the renal artery to the glomerulus, systematically characterizes the spatiotemporal architecture of the entire renal arterial tree and glomeruli in unilateral ureteral obstruction (UUO) and diabetic nephropathy (DN) mouse models. Methods:We combined the vascular tracers of Evans blue and dextran with Ethanol-ECi tissue clearing (EBDE clearing) to achieve the improved labeling of all renal arteries and glomeruli. By further integrating light-sheet microscopy-based 3D imaging with reconstruction and segmentation, we performed high-resolution 3D visualization and quantitative analysis of both renal vascular and glomeruli in normal and diseased kidneys. Furthermore, we investigated, classified, and statistically analyzed the spatiotemporal architecture of the arteries and glomeruli in the kidneys of mice with UUO and DN. Results:Firstly, through quantitative analysis of renal artery diameter, vascular branching, and vascular segment straightness, we found that the arteries of both UUO and DN mice exhibited varying degrees of damage. The renal vascular damage in UUO mice was more severe, manifested by a significant reduction in vessel diameter and the number of vascular branches, as well as an increased in the segment straightness of some arteries; in contrast, the pathological changes in the renal arteries of DN mice were mainly concentrated in vessel diameter and the number of vascular branches. We further classified the glomeruli into three distinct types (M1, M2, and M3) based on their origins, and conducted a quantitative analysis of their 3D spatial distribution within the kidneys of UUO and DN mice, and discovered that in UUO mice, there was a severe reduction in the number and the average volume of the both total and the individual M1-M3 glomeruli. However, in the DN mice, apart from no changes in the total and M1 glomerular number, the other relevant pathological parameters of glomeruli also showed a decrease, although the extent of this decrease was much milder compared to that in the UUO mice. Conclusion:EBDE clearing enables continuous, high-resolution 3D visualization and quantitative analysis from the renal artery to the glomerulus in both normal and diseased intact kidneys. The study not only provides important insights into the renal pathophysiology of UUO and DN, but also establishes a rapid and comprehensive strategy for visualizing structural alterations during disease progression in the kidney and potentially other organs.
Radiopharmaceutical therapy (RPT) plays an important role in modern precision oncology. However, treatment activity is still prescribed by fixed-dose protocols rather than adjusted to planned patient-specific absorbed dose (AD). Individualized dosimetry holds promise for improving treatment planning but is not yet standard. One main reason is that present dosimetry workflows depend on repeated imaging at multiple time points and require kinetic modeling and computational modeling. This review covers recent approaches to make dosimetry more practical and easier to use in everyday RPT workflows. We introduce a theoretical framework in which simplification methods, which help ease the clinical burden of dosimetry, are classified into three related categories. The first two categories simplify the process through direct reduction in the burdens of either the acquisition or processing of data: (1) Methodological simplifications through reduction in the imaging burden, such as reduced- or single-time point (STP) imaging; and (2) Computational automation via artificial intelligence (AI) to automate dosimetry, including image processing, segmentation, registration, and AD estimation. Advanced modeling strategies in the third category improve efficiency through inference, thereby enabling estimation of AD from sparsely sampled data. Reduced imaging played a major role in defining these areas. While all of these have their promise, every category of methods has certain drawbacks. STP techniques are less reliable for heterogeneous tumors, black box models have poor transferability and explainability, and digital twin/physiologically-based pharmacokinetic (DT/PBPK) models lack clinical validation. Fundamental dosimetry, with appropriate validity conditions, could allow for reduced reliance on imaging skills and expertise while still retaining sufficient precision for its intended purposes. Standardization of imaging and multicenter analysis is essential for wider adoption of RPTs. A detailed description of models and their effectiveness would be helpful in the clinical application of dosimetry.
