Catherine C. L. Wong

Current understanding of T-cell immunity largely focuses on protein factors responsible for immune surveillance and regulation, with limited exploration of the roles played by other molecular types. While proteins such as perforin and granzymes are well-established mediators of CD8⁺ T cell cytotoxicity ^1-5 , our study uncovers the cone-shaped phosphatidylethanolamine (PE) lipids as a new class of effector molecules with broad cytotoxic functions against tumor and infected cells. Upon activation, CD8⁺ T cells upregulated the de novo biosynthesis pathway to produce substantial amounts of PE lipids carrying polyunsaturated or monounsaturated fatty acid sidechains. The newly synthesized PE lipids were packaged into vesicular compartments distinct from Lamp1⁺ granules, rapidly polarized to immunological synapses, and secreted through the membrane fusion protein Munc13-4, generating PE⁺ extracellular vesicles enriched in synaptic lipids and proteins. CD8-derived PE lipids were efficiently taken up by target cells and transported to the endoplasmic reticulum via Esyt1, where excessive PE loading induced ER stress and triggered Chop-dependent cell death within a few hours. The PE pathway functioned independently of conventional cytotoxic mechanisms and remained effective even when those pathways were absent. Manipulation of the PE-Esyt1-ER stress axis markedly altered T cell cytotoxicity, and ethanolamine pretreatment of T cells enhanced the efficacy of adoptive cell therapy. These findings introduce a new lipid-based mechanism of cytotoxic immunity and underscore the importance of non-protein molecules as immune effectors.

Background: Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune disease that causes substantial disability worldwide. Early recognition of RA is critical to prevent irreversible joint damage and improve long-term prognosis. However, current diagnostic strategies rely largely on individual serological biomarkers, primarily anti-citrullinated protein antibodies (ACPA) and rheumatoid factor (RF), which fail to identify 20–40% of patients with RA. We therefore sought to develop a robust serum molecular fingerprinting strategy using miniaturized ion mobility spectrometry (IMS) coupled with machine learning for RA screening and to explore the molecular basis of the discriminatory fingerprints using tandem mass spectrometry (MS).Methods: We recruited a total of 321 participants, including 100 patients with RA (age range: 49–65 years; mean: 62.6; SD: 2.0), 100 with osteoarthritis (OA; age range: 40–85 years; mean: 61.5; SD: 9.5), 10 with systemic lupus erythematosus (SLE; age range: 48–68 years; mean: 60.4; SD: 6.7), 11 with Sjögren's syndrome (SS; age range: 49–69 years; mean: 60.4; SD: 6.1), and 100 healthy controls (HC; age range: 44–65 years; mean: 59.6; SD: 5.4) for RA-specific pattern profiling and validation. Serum samples (20 μL) were processed using a high-throughput preparation workflow and analyzed with a direct-spray miniaturized IMS platform to rapidly generate serum molecular fingerprints. Machine learning algorithms, including regularized logistic regression (RLR) and multilayer perceptron (MLP) models, were used to identify characteristic IMS-derived features and distinguish RA-specific molecular patterns from non-RA groups. Diagnostic performance was assessed by receiver operating characteristic (ROC) analysis, including sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), and F1 score. High-resolution trapped ion mobility spectrometry time-of-flight mass spectrometry (tIMS-TOF/MS) and de novo proteome analysis were used to determine the molecular composition and biological pathways underlying the discriminatory fingerprints.Findings: IMS-derived serum fingerprints showed distinct clustering of RA, OA, and HC samples, with high intragroup similarity and clear intergroup separation. Twenty standardized drift-time peaks ranging from 5.58 ms to 21.48 ms were retained as fingerprint features following integration of the RLR and MLP models, of which eight peaks were the principal contributors to RA discrimination. In validation analyses, the IMS-based model achieved excellent discrimination of RA from OA and HC, with an area under the ROC curve (AUC) of 1.0000 for both comparisons. Although discrimination between RA and SLE or SS was lower than for OA and healthy controls, the model maintained informative diagnostic performance, with AUCs of 0.7471 and 0.8222, respectively. For RA versus the combined non-RA group, the model achieved an overall AUC of 0.9995, with an accuracy of 0.969, sensitivity of 0.927, specificity of 0.986, PPV of 0.967, NPV of 0.969, and an F1 score of 0.947, indicating strong diagnostic capability for RA. High-resolution tIMS-TOF/MS identified the molecular composition underlying the discriminatory IMS fingerprints, including immunopeptides, bile acids, and lipid species, while de novo peptide sequencing and proteome profiling showed that these molecular signatures converged on immune regulation, inflammatory responses, complement and coagulation cascades, and lipid metabolic pathways, providing biological interpretation of the RA-associated serum fingerprints.Interpretation: The machine learning-guided miniaturized IMS platform enables rapid and accurate discrimination of RA from non-RA groups (OA, SS, SLE, and HC) by capturing multidimensional serum molecular fingerprints rather than relying on a single biomarker. In addition to its excellent diagnostic performance, high-resolution molecular annotation provides biological interpretability and suggests that these characteristic signatures reflect coordinated alterations across proteins, peptides, lipids, and metabolites. Together, these features make this strategy well suited for clinical translation, with the potential to facilitate rapid molecular triage in routine rheumatology practice.

During autophagy, the contents enclosed within autophagosomes are degraded, while the outer membrane components are recycled from autolysosomes by the recycler complex through the recently discovered autophagosomal components recycling (ACR) process. This recycling is essential for maintaining autophagic activity. However, the molecular machinery and upstream regulatory mechanisms driving this recycling process remain poorly understood. Here, we identify SNX16 as a key component of the recycler complex, which localizes to autolysosomes and is required for ACR. SNX16 functions in ACR by regulating recycler complex formation, facilitating cargo recognition, and mediating the connection between STX17–SNX4–SNX5 and dynein–dynactin complexes. In addition, SNX16–cargo interactions are regulated by two ACR-related small GTPases, Rab32 and Rab38. Importantly, mTORC1 phosphorylates SNX16 to regulate ACR by inhibiting its interactions with STX17 and other recycler components, thus preventing recycler complex formation. Taken together, our findings identify SNX16 as a recycler component and establish a link between mTORC1 and ACR.

Meiotic spindle is an intricate structure and required for chromosome segregation and the proper meiotic progression during oocyte maturation, and its function is regulated by a complex network of proteins located at spindle and its peripheral region. However, proteome of meiotic spindle remains poorly characterized. Here, we acquired the proteomic profile of spindles isolated from metaphase I (MI) and metaphase II (MII) mouse oocytes. In particular, we identified Ccdc69 as a novel regulator of spindle assembly in mouse oocytes. Although deletion of Ccdc69 did not affect female fertility, the MI spindles were elongated in Ccdc69 knockout oocytes. Overexpression of Ccdc69 induced spindle defects by reducing microtubule formation and disturbing acentriolar microtubule organization centers (aMTOCs) distribution. Furthermore, Ccdc69 overexpression impaired kinetochore-microtubule (K-MT) attachment and delayed meiotic progression by abnormal activation of spindle assembly checkpoint (SAC). Taken together, our study depicts the proteome of spindles during mouse oocyte maturation and demonstrates that Ccdc69 regulates spindle assembly and meiotic progression the way similar to “The Tightening Spell of Sun Wukong’s Golden Headband” in the famous Chinese Classic Journey to the West.