Nanosized spinel-type ferrites have gained recognition as a unique class of engineered nanomaterials with promising applications, but their safety profiles remain insufficiently explored. Although iron (Fe), zinc (Zn), and manganese (Mn) are biologically relevant elements, the use of Zn- and Mn-containing ferrite nanoparticles in biomedical contexts demands careful (geno)toxicity evaluation. In this study, three ferrite nanoparticles – γFe2O3 (FeNPs), Zn0.7Fe2.3O4 (ZnNPs), and Mn0.4Fe2.6O4 (MnNPs) – synthesised through a microwave-assisted polyol route, functionalized with citric acid to improve colloidal stability, were evaluated for their potential (geno)toxic effects in an advanced in vitro 3D cell model, HepG2 spheroids. Cellular stress responses upon exposure to the particle were assessed using toxicogenomic analysis.This approach allows the identification of early molecular events that may precede overt toxicity, supporting a mechanistic understanding of adverse outcomes and facilitating the development of predictive biomarkers for hazard assessment. In the present study, the expression of selected DNA damage-responsive genes (TP53, MDM2, GADD45a, CDKN1A, OGG1, and JUNB), apoptosis-related genes (BCL2 and BAX) and oxidative stress response genes (SOD1, CAT, GPX1, GCLC, and GSR) was evaluated. The expression of the selected genes after exposure to the tested nanoparticles was analysed by qPCR primer assays (Applied Biosystems, USA) and One 48.48 Dynamic Array IFC for Gene Expression (Fluidigm, USA). After 24 and 96 hours of exposure, the spheroids were collected, and total RNA was isolated using the RNeasy Mini Kit from Qiagen (Qiagen, Germany) according to the manufacturer's instructions. 10 µg/mL etoposide served as athe positive control for the toxicogenomic analysis. RNA concentration and purity were assessed using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific) by measuring absorbance at 260/280 nm and gele efectrophoresis (Figure 1). Reverse transcription of 1 µg total RNA per sample was performed with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, MA, USA) on a BIO-RAD T100 thermal cycler under conditions listed in Table 3. For preamplification, 4 µL of each of the 24 selected TaqMan assays (SM2) were pooled into a primer mix. The reaction mixture was prepared using TATAA PreAmp GrandMasterMix (Tataa Biocenter, Sweden), the primer pool, and nuclease-free water, following manufacturer instructions. Negative controls (NTC for preamplification and NTCq for qPCR) were included. Each reaction contained 8 µL of mix and 2 µL of 5× diluted cDNA, processed in a 96-deep well plate, sealed, vortexed, and centrifuged (1000 g, 1 min). Preamplification was carried out on a BIO-RAD T100 thermal cycler under conditions in Table 4. Gene expression analysis used TaqMan Universal PCR Master Mix and TaqMan Gene Expression Assays (Table 6). Preamplified samples were diluted 10× with nuclease-free water. Assays were prepared by mixing equal volumes (6 µL) of each assay with Fluidigm Assay Loading Reagent Kit – 10IFCS. The reaction premix combined DNA Sample Loading Reagent and Fast Probe Master Mix (Biotium/Roche) and was added to each diluted cDNA sample. qPCR was performed on 48.48 Dynamic Array™ IFC chips using the Fluidigm BioMark™ HD System under conditions in Table 5. Data were analysed with Fluidigm Gene Expression Analysis Software and quantGenious. Fold changes >1.5 or <0.66 were considered biologically relevant. Statistical significance between NP-exposed cells and solvent controls was assessed using ANOVA and Dunnett’s test in GraphPad Prism v9 (GraphPad Software, CA, USA).