The research, conducted by Rachel Giles, Chrisna Durandt, Melvin A. Ambele and Michael S. Pepper at the University of Pretoria, tackles a bottleneck that has long constrained obesity research. Most laboratory models of adipogenesis—the process by which undifferentiated cells mature into lipid-filled fat cells—rely on murine cell lines such as the immortalized 3T3-L1 preadipocyte line derived from mouse embryos. Although these models are well established, human cells offer greater physiological compatibility and relevance to human disease. The South African team therefore turned to adipose-derived stromal/stem cells, or ASCs, which are isolated from adipose tissue and possess a strong natural capacity to differentiate into adipocytes. Because these cells are of human origin, findings obtained with them are more likely to translate reliably into the clinical setting, making them an excellent platform for studying human fat cell formation in a dish.

Adipogenesis itself is a tightly choreographed biological program. Multipotent mesenchymal stromal/stem cells first commit to the preadipocyte lineage and then accumulate intracellular lipid droplets as they mature into adipocytes. Researchers typically quantify this process in two complementary ways: by measuring the accumulation of lipid droplets inside the cells, a morphological hallmark of differentiation, or by tracking the expression of genes associated with adipogenesis. Both approaches have historically been performed in large 6-well or 12-well culture plates, formats that consume substantial quantities of cells, reagents and time. Screening a library of 100 potential anti-obesity compounds in triplicate under such conditions quickly becomes prohibitively expensive, which is precisely the problem the Pretoria group set out to solve by asking whether adipogenic differentiation could be reliably monitored in much smaller wells.

To answer that question, the team isolated ASCs from adipose tissue donated by patients undergoing liposuction, with informed consent and approval from the university’s Research Ethics Committee. The cells were expanded in complete growth medium and characterized by flow cytometry, showing the expected surface marker profile: positive for CD44, CD73 and CD90, and negative for CD34, CD45 and CD105. The absence of CD105 deviates slightly from the minimal criteria recommended by the International Society for Cell and Gene Therapy and the International Federation for Adipose Therapeutics and Science, but the researchers note that CD105 expression is known to vary with culture conditions, and previous work has shown that CD105-negative mesenchymal stromal cells retain full differentiation potential across all three mesodermal lineages. Consistent with that literature, the cells in this study differentiated robustly into adipocytes when induced.

The experimental design was elegantly systematic. ASCs were seeded into 6-, 12-, 24-, 48- and 96-well plates, with plating densities adjusted to each well size, and driven to differentiate over 21 days using a cocktail of adipogenic inducers: the phosphodiesterase inhibitor IBMX, insulin, the cyclooxygenase inhibitor indomethacin and the glucocorticoid dexamethasone. Non-induced cultures served as undifferentiated controls. Lipid accumulation was then assessed on days 0, 14 and 21 using four independent readouts: fluorescence microscopy, flow cytometry, reverse transcription quantitative polymerase chain reaction (RT-qPCR) and spectrophotometry. Under the microscope, induced cultures in every well size displayed cells studded with multiple intracellular lipid droplets stained by the fluorescent dye Nile Red, while non-induced controls showed only low-level background fluorescence attributable to endogenous cellular fluorophores such as flavins and flavoproteins. By day 21, some cells harbored enlarged droplets, hinting at the droplet merging that characterizes fully mature adipocytes in living tissue.

Flow cytometry provided quantitative confirmation. The proportion of Nile Red-positive cells was highest in the standard 6-well format, reaching 41.85 percent on day 14 and 38.38 percent on day 21, but substantial differentiation was also detected in every smaller format, including roughly 30 percent of cells in the 96-well plates. Differences between days 14 and 21 were not statistically significant, and the only significant differences among plate sizes on day 14 were between the 6-well plate and the 24- and 48-well plates. Spectrophotometry, which measures the optical density of the lipid dye Oil Red O extracted from stained cells and normalized to cell count, likewise showed elevated lipid content in induced cultures across all well sizes, with no significant differences between formats on either measurement day. One technical wrinkle emerged: unbound Oil Red O tended to stick to the plastic of smaller wells, inflating optical density readings in non-induced controls as well size decreased.

The gene expression data told a similar story, with an important caveat. Induced cultures upregulated the key adipogenic transcription factors PPARG and CEBPA—PPARγ being the master regulator of fat cell formation—together with the PPARγ-responsive genes CD36 and FABP4, which mark terminally differentiated adipocytes. Expression levels were statistically indistinguishable across the 6-, 12- and 24-well formats. However, the smaller wells yielded too few cells to extract sufficient RNA consistently, forcing the researchers to pool wells for RT-qPCR and leaving the 48- and 96-well gene expression data incomplete. Low RNA yield from the smallest formats thus stands as the principal limitation for transcript-based readouts, and the authors suggest that the RT-qPCR protocol will need optimization for low cell numbers before gene expression can serve as a truly high-throughput endpoint.

To compare the different plate sizes and methods rigorously, the team employed Bland-Altman multiple comparison plots, a statistical technique that assesses agreement between measurement approaches by examining bias and the limits of agreement. Using the 6-well plate as the reference standard, the 24-well format showed the least variability for flow cytometry and for both PPARG and FABP4 gene expression, while the 96-well plate consistently showed the highest variability. When methods were compared against each other with flow cytometry as the reference, spectrophotometry produced the tightest agreement across all well sizes on both days 14 and 21. The overall conclusion was encouraging: all well sizes were broadly comparable, meaning the choice of assay can be guided by reliability and cost rather than by fundamental incompatibility between formats.

Cost analysis drove the practical payoff home. Assuming a standard 21-day induction and including positive, negative and non-induced controls in triplicate, screening 100 compounds by spectrophotometry in 48-well plates would cost approximately R30,998, or about $1,658, making it the cheapest option by a clear margin. Flow cytometry in the same format would cost roughly $2,099, while RT-qPCR in 12-well plates—the smallest format that reliably yielded sufficient mRNA—would run to about $5,223. The researchers conclude that adipogenic differentiation of human ASCs can be quantified with confidence in 24- and 48-well plates, offering a higher-throughput, more time-efficient and more affordable approach than the traditional 6-well standard. Spectrophotometry emerges as an ideal first-pass screen for anti-adipogenic activity, with flow cytometry providing rich single-cell quantification and RT-qPCR illuminating gene-level effects at later stages of a screening pipeline. The critical next step will be validating the platform with established anti-adipogenic compounds such as PPARγ antagonists, a demonstration that would cement the assay’s utility in the hunt for the next generation of obesity therapies.

Subject of Research: Development of a high-throughput in vitro assay for monitoring adipogenesis of human adipose-derived stromal/stem cells for anti-obesity drug screening

Article Title: Development of a high-throughput assay for monitoring adipogenesis in vitro

Article References: Giles, R., Durandt, C., Ambele, M. A., & Pepper, M. S. (2026). Development of a high-throughput assay for monitoring adipogenesis in vitro. Current Research in Biotechnology, Article 100419. https://doi.org/10.1016/j.crbiot.2026.100419

Image Credits: AI Generated

DOI: 10.1016/j.crbiot.2026.100419

Keywords: adipogenesis, obesity, adipose-derived stromal/stem cells, high-throughput screening, flow cytometry, spectrophotometry, RT-qPCR, Nile Red, Oil Red O, PPARG, drug discovery, cell culture