Naimi, A. | Search bioRxiv

bioRxiv2026

Discovery of a Cell-Permeable Gi-Signaling Inhibitor

Heterotrimeric G proteins regulate diverse physiological processes, yet selective small molecule modulators for the Gi subfamily remain scarce. Here, we employed native metabolomics to screen a fungal culture collection for G protein binders, which led to the discovery of N-hydroxyapiosporamide (NHAP), a fungal specialized metabolite that functionally inhibits Gi-signaling. NHAP selectively binds Gi1 and diminishes Gi1-mediated GTP turnover in biochemical assays. In primary ventricular cardiomyocytes, NHAP largely reversed acetylcholine-induced negative inotropy, demonstrating functional Gi-signaling blockade in a physiological context without acute cytotoxicity. This work establishes NHAP as a first-in-class Gi-selective inhibitor, providing a cell-permeable chemical scaffold for the optimization of next-generation agents to control Gi-signaling in cell-based systems and, ultimately, to target Gi1-driven pathologies.

biochemistry↗

bioRxiv2026

Model-Dependent Renal Phenotypes in Diabetic Kidney Disease: Comparative Histopathological Characterization of Commonly Used Animal Models

BackgroundDiabetic kidney disease (DKD) remains a leading cause of end-stage renal disease worldwide, characterized by progressive structural and metabolic alterations secondary to chronic hyperglycemia. While numerous type 1 and type 2 rodent models have been developed to study the pathophysiology of DKD, no single model perfectly recapitulates the full clinical spectrum of human disease. The selection of an optimal model depends deeply on the specific research objective, as phenotypic expression and histopathological severity vary significantly across different strains and induction methods. The present study provides a comparative analysis of the renal histological of three widely utilized murine models: the chemically induced streptozotocin (STZ) model and the genetic Akita (type 1) and db/db (type 2) models. MethodsMale STZ-induced (28 weeks post-induction), heterozygous Akita (28 weeks old), and db/db mice at two different age intervals (18-21 and 16-24 weeks old) were assessed. Renal injury was quantified using four light-microscopic parameters: glomerulomegaly, mesangial hypercellularity, tubular vacuolization and arteriolar hyalinosis. Due to observed discrepancies between metabolic and structural findings in the db/db strain, transmission electron microscopy (TEM) was employed for subcellular characterization. ResultsAll models exhibited significant hyperglycemia and albuminuria. At the light-microscopic level, STZ and Akita mice demonstrated consistent and pronounced renal lesions. In contrast, db/db mice despite increasing albuminuria and obesity, light microscopy revealed heterogeneous and inconsistent histopathological changes. However, TEM analysis of db/db mice kidneys successfully captured early ultrastructural injury, including irregular glomerular basement membrane (GBM) thickening and focal podocyte foot process effacement, which were undetectable by light microscopy. ConclusionsOur findings indicate that the Akita and STZ-induced models exhibit prominent structural alterations detectable by conventional light microscopy, whereas the db/db model requires ultrastructural evaluation by TEM to reliably confirm renal injury. This study underscores the limitation of routine histology in certain type 2 diabetes models and highlights the complementary value of TEM for accurate histopathological characterization. Collectively, the alternative histopathological markers identified herein offer sensitive and readily accessible indices for monitoring early-to-moderate DKD progression, providing a more robust framework for preclinical model selection and therapeutic evaluation in future studies.

pathology↗

bioRxiv2017

Stacked Generalization: An Introduction to Super Learning

Stacked generalization is an ensemble method that allows researchers to combine several different prediction algorithms into one. Since its introduction in the early 1990s, the method has evolved several times into what is now known as \"Super Learner\". Super Learner uses V -fold cross-validation to build the optimal weighted combination of predictions from a library of candidate algorithms. Optimality is defined by a user-specified objective function, such as minimizing mean squared error or maximizing the area under the receiver operating characteristic curve. Although relatively simple in nature, use of the Super Learner by epidemiologists has been hampered by limitations in understanding conceptual and technical details. We work step-by-step through two examples to illustrate concepts and address common concerns.

epidemiology↗