Daily Endocrinology Research Analysis
Three papers stood out today in endocrinology: (1) LEOPARD, a novel AI method to complete missing views in longitudinal multi-omics, robustly outperforms standard imputations and enables better temporal biology in personalized health. (2) A mechanistic Cell Reports study reveals that reduced Notch signaling in hypothalamic endothelial cells drives early high-fat-diet effects on brain glucose transport, highlighting a neurovascular target in obesity. (3) A large Danish cohort shows YKL-40 strongl
Summary
Three papers stood out today in endocrinology: (1) LEOPARD, a novel AI method to complete missing views in longitudinal multi-omics, robustly outperforms standard imputations and enables better temporal biology in personalized health. (2) A mechanistic Cell Reports study reveals that reduced Notch signaling in hypothalamic endothelial cells drives early high-fat-diet effects on brain glucose transport, highlighting a neurovascular target in obesity. (3) A large Danish cohort shows YKL-40 strongly predicts incident liver and bladder cancers in early type 2 diabetes, outperforming CRP for these sites.
Research Themes
- AI-enabled longitudinal multi-omics for endocrinology
- Neurovascular regulation of metabolic homeostasis
- Cancer risk biomarkers in early type 2 diabetes
Selected Articles
1. LEOPARD: missing view completion for multi-timepoint omics data via representation disentanglement and temporal knowledge transfer.
LEOPARD introduces a representation-disentanglement approach to complete missing omics views in longitudinal datasets and consistently outperforms common imputation methods. Validated across four real-world cohorts spanning 3 days to 14 years, it improves detection of age-associated metabolites, eGFR-associated proteins, and CKD prediction.
Impact: This is a first generalized approach for missing-view completion in longitudinal multi-omics, enabling robust temporal analyses central to endocrine and metabolic disease research.
Clinical Implications: While methodological, LEOPARD can enhance biomarker discovery, patient stratification, and trajectory modeling in diabetes, obesity, and endocrine disorders by maximizing use of incomplete longitudinal omics.
Key Findings
- Introduces LEOPARD, which disentangles content and temporal representations to impute missing omics views.
- Validated on four real-world datasets (MGH COVID, KORA) covering 3 days to 14 years, outperforming missForest, PMM, GLMM, and cGAN.
- LEOPARD-imputed data showed highest agreement with observed data for age-related metabolites, eGFR-associated proteins, and CKD prediction.
Methodological Strengths
- Multi-cohort validation across distinct timescales and diseases.
- Head-to-head benchmarking against multiple state-of-the-art imputation methods.
Limitations
- No prospective clinical outcome validation directly tied to imputed features.
- Generalizability to omics modalities beyond those tested remains to be established.
Future Directions: Prospectively validate LEOPARD-driven biomarkers in endocrine cohorts (e.g., T2D, obesity), integrate EHR and imaging, and extend to multimodal (omics-clinical) harmonization.
Longitudinal multi-view omics data offer unique insights into the temporal dynamics of individual-level physiology, which provides opportunities to advance personalized healthcare. However, the common occurrence of incomplete views makes extrapolation tasks difficult, and there is a lack of tailored methods for this critical issue. Here, we introduce LEOPARD, an innovative approach specifically designed to complete missing views in multi-timepoint omics data. By disentangling longitudinal omics data into content and temporal representations, LEOPARD transfers the temporal knowledge to the omics-specific content, thereby completing missing views. The effectiveness of LEOPARD is validated on four real-world omics datasets constructed with data from the MGH COVID study and the KORA cohort, spanning periods from 3 days to 14 years. Compared to conventional imputation methods, such as missForest, PMM, GLMM, and cGAN, LEOPARD yields the most robust results across the benchmark datasets. LEOPARD-imputed data also achieve the highest agreement with observed data in our analyses for age-associated metabolites detection, estimated glomerular filtration rate-associated proteins identification, and chronic kidney disease prediction. Our work takes the first step toward a generalized treatment of missing views in longitudinal omics data, enabling comprehensive exploration of temporal dynamics and providing valuable insights into personalized healthcare.
2. Reduced Notch signaling in hypothalamic endothelial cells mediates obesity-induced alterations in glucose uptake and insulin signaling.
Short-term HFD rapidly suppresses Notch signaling in hypothalamic endothelial cells, reducing GLUT1 and brain glucose uptake. Notch activation rescues GLUT1 expression and glucose uptake in vivo and in cultured BMECs, suggesting endothelial Notch as a neurovascular target in obesity.
Impact: Identifies a mechanistic BBB pathway linking diet-induced obesity to altered central glucose handling, opening avenues for modulating endothelial signaling to restore metabolic homeostasis.
Clinical Implications: While preclinical, targeting endothelial Notch-GLUT1 pathways could inspire therapies to improve central glucose sensing and insulin signaling in obesity and type 2 diabetes.
Key Findings
- Short-term high-fat diet rapidly downregulates Notch signaling in brain microvascular endothelial cells.
- Notch activation restores GLUT1 expression and glycolysis in cultured BMECs exposed to HFD-fed mouse serum.
- Endothelial Notch intracellular domain expression prevents HFD-induced reduction of GLUT1 and hypothalamic glucose uptake in vivo.
- Caveolin-1 expression in BMECs increases with short-term HFD feeding.
Methodological Strengths
- Combines in vivo endothelial-specific genetic activation with in vitro BMEC assays.
- Direct functional readouts of GLUT1 expression and hypothalamic glucose uptake.
Limitations
- Abstract truncation limits detailed understanding of downstream pathways (e.g., Cav-1 role with Notch activation).
- Short-term HFD model; translational relevance to chronic obesity requires further validation.
Future Directions: Define how endothelial Notch interfaces with caveolar trafficking and insulin signaling at the BBB, and test pharmacologic modulation in chronic metabolic disease models.
Short-term transition to high-fat diet (HFD) feeding causes rapid changes in the molecular architecture of the blood-brain barrier (BBB), BBB permeability, and brain glucose uptake. However, the precise mechanisms responsible for these changes remain elusive. Here, we detect a rapid downregulation of Notch signaling after short-term HFD feeding. Conversely, Notch activation restores HFD-fed mouse serum-induced reduction of Glut1 expression and glycolysis in cultured brain microvascular endothelial cells (BMECs). Selective, inducible expression of the Notch intracellular domain (IC) in BMECs prevents HFD-induced reduction of Glut1 expression and hypothalamic glucose uptake. Caveolin (Cav)-1 expression in BMECs is increased upon short-term HFD feeding. However, NotchIC
3. YKL-40 and risk of incident cancer in early type 2 diabetes: a Danish cohort study.
In 11,346 individuals with early T2D followed up to 14 years, higher YKL-40 strongly predicted liver (HR 44.2) and bladder (HR 4.2) cancers and was comparable to CRP for obesity-related and GI cancers. YKL-40 outperformed CRP for liver/bladder cancers, indicating distinct biomarker roles.
Impact: Provides robust evidence that YKL-40 improves cancer risk stratification in early T2D, particularly for liver cancer, with potential to guide surveillance strategies.
Clinical Implications: In early T2D, YKL-40 measurement could inform targeted liver (and possibly bladder) cancer surveillance beyond CRP; validation and cost-effectiveness analyses are warranted.
Key Findings
- In a cohort of 11,346 new T2D patients, highest vs lowest YKL-40 category associated with HRs: 2.4 (obesity-related), 2.6 (GI), 44.2 (liver), 4.2 (bladder) cancers.
- YKL-40 outperformed CRP for predicting liver and bladder cancers; CRP performed better for lung, colorectal, and ovarian cancers.
- No significant associations were found for other cancer sites.
Methodological Strengths
- Large, well-defined incident T2D cohort with up to 14-year follow-up and Cox modeling.
- Direct comparison of YKL-40 with CRP across multiple cancer sites.
Limitations
- Observational design limits causal inference and residual confounding cannot be excluded.
- Generalizability outside the Danish healthcare context requires evaluation.
Future Directions: External validation across diverse T2D populations, integration into multivariable risk scores, and trials to assess surveillance benefits and cost-effectiveness.
BACKGROUND: We examined the association of serum YKL-40, an inflammatory biomarker, with incident cancer risk in early type 2 diabetes. METHODS: A cohort of 11,346 individuals newly diagnosed with type 2 diabetes was followed for up to 14 years. YKL-40 levels (n = 9010) were categorised into five percentiles (0-33%, 34-66%, 67-90%, 91-95%, and 96-100%), and baseline YKL-40 and CRP (n = 9644) were analyzed continuously (per 1 SD log increment) for comparison. Cox regression models assessed associations with obesity-related, gastrointestinal, liver, pancreatic, colorectal, bladder and lung cancers, as well as cancers of reproductive organs. RESULTS: Adjusted HRs (95% CIs) for the highest versus lowest YKL-40 category were 2.4 (1.6-3.7) for obesity-related, 2.6 (1.7-4.1) for gastrointestinal, 44.2 (12.8-153.4) for liver, and 4.2 (1.3-14.1) for bladder cancers. No associations were found for other cancers. YKL-40 and CRP had similar prognostic abilities for obesity-related and gastrointestinal cancers, but YKL-40 outperformed CRP for liver and bladder cancers. Conversely, CRP was a stronger predictor for lung, colorectal, and ovarian cancers. DISCUSSION: YKL-40 was associated with the risks of liver and bladder cancers, clearly outperforming CRP for these cancers. This suggests distinct prognostic roles for YKL-40 and CRP, and highlights YKL-40 as a promising biomarker for liver cancer.