The review, conducted by researchers at Hanoi Medical University and collaborating institutions in Vietnam and Japan, followed the PRISMA 2020 guidelines and screened 15,835 records from PubMed, Scopus, Embase, dissertation databases and preprint servers. From that enormous pool, 350 primary experimental studies met the eligibility criteria. The team cataloged 206 distinct molecular entries in adipocyte studies and 109 in skeletal-muscle studies, each assigned to a specific phase of the exocytotic process: signaling, vesicle translocation, tethering, docking, and membrane fusion. The protocol was prospectively registered in PROSPERO, and two reviewers independently performed study selection, data extraction, and quality appraisal.

To make sense of such a heterogeneous literature, the authors applied a prespecified evidence-tier framework. A molecule could earn Tier 1 status, the strongest category, only if perturbation and rescue experiments, interaction or localization data, and a direct GLUT4 trafficking endpoint converged in the same biological context. Lower tiers captured genetic perturbation without full mechanistic rescue, pharmacological evidence with limited target specificity, and purely descriptive or correlative findings. This tiering matters because a molecule’s status is not fixed: one protein might have Tier 1 support in 3T3-L1 adipocytes but only Tier 4 descriptive evidence in human tissue, a distinction the review makes explicit throughout its synthesis.

Under basal conditions, GLUT4 sits sequestered in specialized storage vesicles inside the cell. Insulin triggers the insulin receptor, insulin receptor substrates, and phosphoinositide 3-kinase to activate AKT, and the review found that AKT2 is the isoform with the most consistent experimental support. In adipocytes, AKT2 phosphorylates the Rab-GTPase activating protein TBC1D4, also known as AS160, which acts as a molecular brake. Phosphorylation relieves this brake, freeing Rab GTPases such as RAB10 to drive vesicle mobilization toward the cell cortex. Genetic depletion, rescue experiments, and live trafficking measurements all support this model, although the evidence also shows that TBC1D4-independent routes coexist, meaning no single gatekeeper controls vesicle release.

The review also highlights a parallel branch of cortical regulation built around Rho-family GTPases. In 3T3-L1 adipocytes, the CAP/APS/CBL/CRK/C3G cascade activates TC10, which organizes the actin cortex and recruits vesicles to the cell periphery. RAC1, supported both in cell culture and in adipocyte-specific knockout mice, appears to act as a complementary, context-dependent regulator rather than a redundant duplicate. Downstream, TUG-dependent proteolytic processing releases GLUT4 from Golgi-associated retention compartments, RAB10 recruits the exocyst tethering complex, and kinesin KIF13A drives microtubule-based transport. At the terminal step, functional experiments distinguish clear roles: the t-SNAREs syntaxin-4 and SNAP23 are indispensable for tethering vesicles at the plasma membrane, while the v-SNARE VAMP2 is preferentially required for the final fusion event itself.

Skeletal muscle tells a related but distinct story. Because muscle must respond both to insulin after meals and to contraction during exercise, its trafficking network integrates two different classes of inputs. The review found strong support for an insulin-responsive relay in which AKT2 activates the guanine-nucleotide exchange factor FLJ00068, which in turn switches on RAC1 and cortical actin remodeling. On the contraction side, AMPK and calcium-associated pathways engage the muscle-enriched Rab-GAP TBC1D1, whose Ser231 phosphorylation contributes in a context-dependent manner to AMPK-regulated glucose uptake, while TBC1D4 serves as a convergence node for both insulin and exercise-related sensitization. Electrical-pulse stimulation experiments in C2C12 myotubes even revealed cross-talk, with a CaMKII-Kalirin-RAC1-AKT link connecting contraction signals back to insulin signaling.

The downstream Rab and cytoskeletal machinery also diverges between tissues. Where adipocytes rely most heavily on RAB10, skeletal muscle experiments identify RAB8A and RAB13, acting with myosin Va and the MICAL-L2-alpha-actinin-4 module respectively, as key late-trafficking regulators in L6 and C2C12 systems. ARP2/3-cofilin actin cycling has convergent Tier 1 support in L6 cells, and MYO1C mediates insulin-induced vesicle tethering in both L6 cultures and isolated rat muscle fibers. In vivo, muscle-specific deletion of the exocyst component Exoc5 disrupted GLUT4 exocytosis, providing genetic confirmation that the tethering apparatus matters in living animals, not just in culture dishes. Syntaxin-4 transgenic mice showed enhanced insulin-mediated glucose uptake, and VAMP2 carries functional fusion evidence in L6 cells.

Perhaps the most clinically significant conclusion is that the shared terminal machinery coexists with genuinely different upstream logic. Both tissues converge on SNARE- and exocyst-dependent surface delivery, but adipocytes appear optimized for rapid insulin-driven nutrient storage, while muscle preserves glucose disposal during exercise even when insulin sensitivity is impaired. The authors frame defects in signal-to-trafficking coupling as a complementary mechanistic layer of insulin resistance, operating alongside receptor-proximal signaling failure, lipotoxicity, inflammation, mitochondrial stress, and reduced GLUT4 expression. Human biopsy studies corroborate this framework: in skeletal muscle of people with type 2 diabetes, insulin-stimulated TBC1D4 phosphorylation is reduced despite largely preserved upstream AKT activity, proteomics reveals coordinated losses of endosomal sorting proteins, and adipose GLUT4 mRNA is markedly reduced and correlates with insulin resistance.

The review is candid about its limitations. Most mechanistic evidence derives from differentiated cell lines such as 3T3-L1 adipocytes and L6 or C2C12 myotubes, with rodent models strengthening selected nodes and human-derived material remaining sparse. Risk-of-bias assessment showed that incomplete reporting, rather than overtly flawed methods, was the dominant concern: randomization, blinding, and cell-line authentication were rarely documented. The larger inventory of adipocyte regulators may partly reflect experimental accessibility rather than true biological complexity. The authors argue that future work should prioritize physiologically relevant human systems, combining endogenous protein tagging, genome editing, spatial proteomics, and live-cell imaging to map which trafficking checkpoints are genuinely disrupted in obesity, aging, and physical inactivity. Therapeutically, the framework suggests a new classification scheme: rather than only boosting insulin signaling, future interventions might target vesicle mobilization, cytoskeletal remodeling, tethering, docking, or fusion directly, offering mechanistically distinct routes to restore glucose disposal in metabolic disease.

Subject of Research: Molecular regulation of GLUT4 vesicle trafficking and exocytosis in adipocytes and skeletal muscle

Article Title: Molecular architecture of GLUT4 trafficking and exocytosis in adipocytes and skeletal muscle: A systematic review and integrated conceptual framework

Article References: Nguyen, H. S., Tran, K. T., Ho, M. S., Nguyen, L. M., Tran, T. M., Vu, P. H., Nguyen, L. T., Pham, L. T., Pham, H. N., My Phan, D. T., Pham, M. D., Nguyen, K. D., Bui, H. T. M., Hosaka, T., Tran, N. H., Phan, H. H., & Kim Le, C. T. (2026). Molecular architecture of GLUT4 trafficking and exocytosis in adipocytes and skeletal muscle: A systematic review and integrated conceptual framework. Frontiers in Endocrinology, 17, Article 1932455. https://doi.org/10.3389/fendo.2026.1932455

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1932455

Keywords: GLUT4, insulin signaling, exocytosis, adipocytes, skeletal muscle, SNARE proteins, TBC1D4, AKT2, Rab GTPases, insulin resistance, systematic review, glucose transport