Daily Endocrinology Research Analysis
Three studies advance precision endocrinology across thyroid, adrenal, and type 1 diabetes. A transcriptomic classifier plus CA12 targeting predicts and mitigates invasion in RAS-mutant thyroid tumors; plasma proteomics accurately distinguishes unilateral primary aldosteronism and tracks post-adrenalectomy change; and a meta-analysis clarifies which tests best predict progression to stage 3 type 1 diabetes in autoantibody-positive youth.
Summary
Three studies advance precision endocrinology across thyroid, adrenal, and type 1 diabetes. A transcriptomic classifier plus CA12 targeting predicts and mitigates invasion in RAS-mutant thyroid tumors; plasma proteomics accurately distinguishes unilateral primary aldosteronism and tracks post-adrenalectomy change; and a meta-analysis clarifies which tests best predict progression to stage 3 type 1 diabetes in autoantibody-positive youth.
Research Themes
- Omics-driven precision diagnostics in endocrine diseases
- Translational targets linking mechanisms to therapy (CA12 in RAS-mutant thyroid cancer)
- Risk prediction tools for early-stage type 1 diabetes progression
Selected Articles
1. Alterations in gene expression associated with invasion of RAS-mutant thyroid tumors and their potential diagnostic and therapeutic utility.
RNA-seq of 48 RAS-mutant thyroid tumors revealed distinct expression profiles between invasive and non-invasive lesions. A 6-gene panel (CA12, CD44, LRP4, ECM1, FN1, CRABP1) plus nodule size predicted invasion in RAS-mutant FNA samples with 95% sensitivity and 89% specificity. Targeting CA12 reduced invasion in vitro and arrested growth in RAS-mutant xenografts.
Impact: This study links a clinically actionable transcriptomic classifier with a druggable target (CA12) in RAS-mutant thyroid tumors, enabling both improved preoperative risk stratification and a plausible therapeutic avenue.
Clinical Implications: Preoperative FNA-based testing using the 6-gene panel could identify invasive RAS-mutant nodules to guide extent of surgery and surveillance. CA12 inhibition represents a potential targeted therapy for invasive RAS-mutant thyroid tumors.
Key Findings
- Invasive vs non-invasive RAS-mutant tumors exhibited distinct RNA-seq expression profiles.
- A 6-gene panel (CA12, CD44, LRP4, ECM1, FN1, CRABP1) plus nodule size predicted invasion in FNA samples (95% sensitivity, 89% specificity).
- siRNA and chemical inhibition of CA12 reduced invasion in RAS-mutant thyroid cells; CA12 inhibitors arrested growth in RAS-mutant xenografts.
Methodological Strengths
- Multi-stage validation: discovery by RNA-seq, qRT-PCR confirmation, independent FNA cohort prediction.
- Mechanistic corroboration with in vitro knockdown/inhibition and in vivo xenograft efficacy.
Limitations
- Moderate sample size from a single research network; external multicenter validation is needed.
- Classifier limited to RAS-mutant nodules; applicability to non-RAS tumors is unknown.
Future Directions: Prospective multicenter validation of the FNA classifier, pharmacologic optimization of CA12 inhibitors, and evaluation of combination strategies in RAS-mutant thyroid cancer.
INTRODUCTION: Mutations of RAS genes are detected in a spectrum of follicular-patterned thyroid tumors. Preoperative prediction of invasive cancers based on the presence of RAS mutation alone is challenging because non-invasive and invasive tumors tend to have similar sonographic and cytologic features. The aim of this study was to perform clinicopathologic and molecular analyses of RAS-mutant tumors, identify molecular and clinical markers associated with invasiveness, and determine their diagnostic and thera
2. Reactivity-based metabolomics reveal cysteine has glyoxalase 1-like and glyoxalase 2-like activities.
Using symmetric isotope-labeled reactivity-based metabolomics, the authors identified >200 MG/LGSH adducts in living cells, including abundant lactoylated amino acids. Cysteine rapidly forms D-Lac-Cys from LGSH and L-Lac-Cys from MG, revealing cysteine possesses glyoxalase 1-like and glyoxalase 2-like activities. These adducts are dynamically regulated by cellular cysteine/MG and increase in diabetes, suggesting biomarker potential and two additional nonenzymatic protein lactoylation pathways.
Impact: This study revises fundamental reactive carbonyl detoxification biology by assigning dual glyoxalase-like activities to cysteine, with direct implications for metabolic disease biomarkers and protein post-translational modification pathways.
Clinical Implications: D-/L-lactoyl-cysteine adducts may serve as biomarkers of glyoxal/metabolic stress in diabetes and aging; targeting MG/LGSH–cysteine chemistry may modulate nonenzymatic lactoylation implicated in complications.
Key Findings
- Reactivity-based metabolomics in living cells identified >200 MG/LGSH adducts, with abundant lactoylated amino acids.
- Cysteine forms D-Lac-Cys from LGSH and L-Lac-Cys from MG, conferring glyoxalase 1-like and 2-like activities.
- D-/L-Lac-Cys levels are dynamically regulated by intracellular cysteine and MG and are elevated in diabetes; cysteine amides undergo lactoylation, suggesting two additional nonenzymatic protein lactoylation pathways.
Methodological Strengths
- Symmetric isotope-labeled, reactivity-based metabolomics enabling confident adduct assignment in living cells.
- Convergence of chemical, cellular, and disease-relevant measurements linking findings to diabetes.
Limitations
- Primarily mechanistic preclinical evidence; clinical validation of adducts as biomarkers is pending.
- Quantitative flux and tissue distribution across organs were not comprehensively characterized.
Future Directions: Validate D-/L-lactoyl-cysteine in human cohorts as prognostic biomarkers, map organ/tissue distribution, and test interventions modulating MG/LGSH–cysteine chemistry.
Methylglyoxal (MG) is a reactive metabolite involved in diabetes and aging through the formation of protein adducts. Less is known about the extent that MG and its metabolic product S-D-lactoylglutathione (LGSH) form adducts with cell metabolites. Using a 'symmetric' isotope-labeled and reactivity-based metabolomics approach in living cells, we found over 200 adducts and, surprisingly, discovered that 10 of the most abundant are lactoylated amino acids mainly derived from LGSH. The most abundant adduct D
3. Proteomic signatures to detect unilateral primary aldosteronism in hypertensive patients.
A plasma proteomic model using six peptide features (HBB, FIBA, Complement C7, ALBU, C4BPA, A2AP) distinguished unilateral primary aldosteronism from essential hypertension with sensitivity/specificity ≈81–83% and AUC 0.92. Risk scores decreased significantly after unilateral adrenalectomy, indicating potential for diagnosis and monitoring.
Impact: Provides a non-invasive, omics-based adjunct to detect unilateral PA and to assess surgical response, potentially reducing reliance on invasive adrenal venous sampling.
Clinical Implications: Proteomic risk scoring could triage hypertensive patients for confirmatory testing and guide surgical candidacy; post-operative decreases in scores offer a monitoring tool for cure assessment.
Key Findings
- Six peptide features yielded a risk score differentiating unilateral PA from essential hypertension with sensitivity/specificity ~81–83% and AUC 0.92.
- Risk scores were significantly higher in PA than EH and decreased after unilateral adrenalectomy.
- Model performance was consistent across training and validation cohorts.
Methodological Strengths
- Independent training/validation cohorts with pre- and post-operative sampling.
- Deep proteomic profiling with feature selection and risk score generation.
Limitations
- Modest sample size from specialized centers; external multicenter validation required.
- Focused on unilateral PA; applicability to bilateral disease not addressed.
Future Directions: Prospective multicenter validation, comparison with adrenal venous sampling, and cost-effectiveness analyses for clinical implementation.
CONTEXT: Primary aldosteronism (PA) is a major cause of hypertension and cardiovascular disease; however, diagnosing PA remains challenging. OBJECTIVE: We investigated whether deep proteomic analyses could be used to diagnose unilateral PA in hypertensive patients. METHODS: We enrolled 52 patients with unilateral PA and 46 with essential hypertension (EH) and divided them into training and validation cohorts. Plasma samples were collected at baseline from all patients and again from PA patients after ad