Your concern is valid GAPDH is not a reliable housekeeping gene in 3D cultures or under hypoxic conditions, as its expression can be significantly influenced by metabolic reprogramming and oxygen tension, both of which are common in tumor spheroid models. In spheroids, especially under hypoxia, glycolytic genes including GAPDH and ACTB often show variable expression, leading to normalization errors in qPCR.

For more stable reference genes in 3D or hypoxic conditions, several studies suggest evaluating and validating candidates such as HPRT1, RPLP0, TBP, YWHAZ, or PPIA. Among them, RPLP0 and TBP have shown relatively stable expression across 2D vs 3D systems and in oxygen-variable environments in various cancer cell models.

To ensure accurate normalization, it is strongly recommended to use geNorm, NormFinder, or BestKeeper to validate reference gene stability in your specific 3D spheroid conditions. Ideally, use two or more validated reference genes for normalization (per MIQE guidelines), particularly when analyzing subtle differences in cancer stem cell (CSC) marker expression.

Thank you for for your answer