Hierarchy of Nanoparticles Toxicity Factors Significance as Extracted from NanoCommons Knowledge Base: Influence of Compound, Cell line and Particle Size on Cell Viability Authors: 1. Michal Macko1 2. František Božek1 3. Diana Kmetkova2 4. Jan Antoš1 5. Ivo Ku itkař 1 Affiliations: 1) Tomas Bata University in Zlin, Centre of polymer systems, t ída Tomášeř Bati 5678, 760 01 Zlín, 760 01 Zlín, Czech Republic 2) Institute of Economic Studies, Charles University, Opletalova 26, Prague, Czech Republic Corresponding authors: Ivo Ku itka, Michal Mackoř Address: Tomas Bata University in Zlin, Centre of polymer systems, t ídař Tomáše Bati 5678, 760 01 Zlín, 760 01 Zlín, Czech Republic E-mail: kuritka@utb.cz; mmacko@utb.cz Abstract The objective of the paper was to conduct a thorough statistical analysis of nanoparticles of publicly available data found in the NanoCommons database (“BioXMTM Knowledge Portal." NanoCommons,” 2024), specifically, how various parameters influence cell viability value (CVV). The analysis was performed on nine hundred measurements, where required parameters (compound, dose, exposure time, cell line, nanoparticle size) were available. The data represented a collection out of seven datasets, which contained five cell lines and six compounds. Other parameters included a constant exposure time of 24h, particle diameters from 8 to 150 nm, and exposure doses ranging from 0.3 to 250 [µg/ml]. The effects of variables were explored via two main models - first model was a Multiple Ordinary Least Squares (OLS) Model without interactions (Model 1), aimed at exploring the significance of certain variables and their effect on cell viability without considering any interactions. The second was a Multiple OLS Model with interactions (Model 2), which aimed to explore the significance of certain variables on cell viability while also considering interactions between some of the regressors. The findings emphasize the importance of analyzing compound- and cell-specific interactions, as well as the interplay between particle size and nanoparticle reactivity. Graphical abstract Keywords Nanoparticles, toxicity, statistical analysis, particle size Introduction Nanoparticles (NPs) are defined as particles with at least one dimension ranging between 1 – 100 nm (Bezza et al., 2020). Their properties emerging due to their small size and high surface-to-volume ratio are explored in many industrial fields, such as medicine (Islam et al., 2022), electronics (Chakraborty et al., 2022), agriculture (ESKİN, 2022; Pansambal et al., 2023) or environmental science (Arun et al., 2023; Samuel et al., 2022). Besides their positive properties, they also have adverse toxicological effects on biological systems (Gupta and Xie, 2018). Therefore, the mechanisms of nanoparticle-induced toxicity are intensively researched (Gudkov et al., 2021; Naz et al., 2020; Nie et al., 2023; Thakur et al., 2024). The toxicity mechanisms depend on numerous factors. Among them, the most acknowledged are elemental composition, size, shape, or surface charge (Abbasi et al., 2023). A large amount of knowledge has been gathered but remains dispersed among individual studies. To obtain the bird's-eye view on the multiple sources of various data, an H2020 Infrastructures project, NanoCommons was executed creating a community framework and infrastructure for reproducible science, and in particular for in silico workflows for nanomaterials safety assessment and beyond (“NanoCommons: The European Nanotechnology Knowledge Infrastructure,” 2025). The NanoCommons database is a knowledge base collecting and sharing of self-consistent and reliable data on nanoparticle characterization and toxicity. Among many others, the toxicological behaviour and toxicity mechanisms of six common compounds available as NPs are considered, aiming at the utilisation of the data from the NanoCommons database in this study: (i) oxide materials, namely titanium dioxide (TiO ), zinc oxide (ZnO), cerium₂ dioxide (CeO ), silicon dioxide (SiO ), and (ii) metals, namely silver (Ag),₂ ₂ copper (Cu). Titanium dioxide (TiO ) is one of the most widely produced semiconducting₂ nanomaterials. TiO NPs possess UV-blocking and photocatalytic₂ properties, high stability, and biocompatibility. Although TiO NPs are₂ generally considered biologically inert under dark conditions, upon photoactivation by UV light, they generate reactive oxygen species (ROS), can cause DNA damage, or lead to pulmonary toxicity after inhalation (Mohamed, 2015; Ziental et al., 2020). Another factor influencing the toxicity of TiO NPs is the crystal structure - anatase vs. rutile. The₂ observations indicate that the anatase form is more toxic due to higher photocatalytic activity leading to increased generation of ROS (Baranowska-Wójcik et al., 2020; Nosaka and Nosaka, 2017; Thakur et al., 2024). Zinc oxide nanoparticles (ZnO NPs) are also mass-scale produced and have already been introduced in many applications utilizing their antimicrobial and UV protective functions and relatively high biocompatibility. Interestingly, enhancement of drug delivery was also reported. (Gudkov et al., 2021; Islam et al., 2022; Silva et al., 2019). On the other hand, ZnO NPs can cause oxidative stress, lipid peroxidation, DNA damage, and ultimately, cell death, which are ascribed to the formation of reactive oxygen species (ROS) or due to the release of zinc ions (Zn²⁺) (Canta and Cauda, 2020; Chang et al., 2012; Chen et al., 2019; Fukui et al., 2012; Song et al., 2010). Another industrial large-scale produced semiconductor nanomaterial is cerium dioxide (CeO2). CeO2 NPs can act both as antioxidants and prooxidants, due to their ability to switch between cerium (III) and cerium (IV) oxidation states, often used as catalyst. The nature of CeO2 NPs allows to scavenge ROS and protect cells from oxidative damage continuously (Singh et al., 2020), and for that reason, among many others, their applications can be found in sunscreens or cosmetics. (Pansambal et al., 2023). Nevertheless, at higher concentrations, CeO2 NPs can induce oxidative stress, cause DNA damage, or trigger immune responses and inflammation (Forest et al., 2017; Gao et al., 2014). Silica, i.e. silicon dioxide (SiO ) nanoparticles, is the second most widely₂ produced oxide nanomaterial after alumina. SiO NPs are widely used in₂ food, cosmetics, and pharmaceuticals because of their biocompatible and chemically stable properties (Slowing et al., 2008). While generally recognized as considerably less toxic than other nanoparticles, SiO NPs₂ can induce toxicity associated with surface reactivity, which can lead to the generation of ROS, inflammatory responses, membrane damage, and DNA fragmentation (El-Sayed et al., 2021; ESKİN, 2022; Liman et al., 2020; Rafieepour et al., 2021; Yang et al., 2019). Due to their exceptional antimicrobial and biocompatible properties, metal silver nanoparticles (Ag NPs) are used in many fields such as agriculture, electronics, textile, hygiene or medicine and medical devices. When it comes to the mechanism of toxicity, primarily, it is the release of silver ions (Ag⁺) that induces oxidative stress. Direct nanoparticle interaction with the cell membrane was also described. Next, even in the absence of ROS, Ag NPs can cause damage to the mitochondrial function, leading to cell death (Bruna et al., 2021; Luceri et al., 2023; Nie et al., 2023; Xu et al., 2020). Another widespread material is copper (Cu) and copper-based NPs, often in the form of metal oxide (CuO). Among others, metal Cu NPs are used as materials for superconductors in electronics or as antimicrobial agents for coatings and food preservatives (Chakraborty et al., 2022; Naika et al., 2015; Yip and Sauls, 1992). However, when they interact with biological systems, copper ions (Cu2⁺) may be released, which may induce oxidative stress, or cause mitochondrial malfunction and DNA damage (Ahamed et al., 2010; Akintelu et al., 2020; Fahmy and Cormier, 2009; Naz et al., 2020). The logical counterpart in the evaluation of a toxic agent is its target. The target is determined by the method used – in vitro (outside of a living organism) or in vivo (within a living organism) (Savage et al., 2019). The scope of this paper focuses on in vitro methods, specifically measurements conducted on cell lines. Each cell line used for testing represents specific tissue characteristics, metabolic capacities, and stress response mechanisms, resulting in their different sensitivity to various compounds. Therefore, selecting appropriate cell lines is critical for understanding toxicity mechanisms, predicting in vivo responses, and evaluating potential risks (Kumar et al., 2017). Among the cell lines examined in the studies collected in the NanoCommons database, five cell lines are relevant in this study, namely HepG2, HepaRG, A549, RAW264.7, and HCT116. This selection covers four important areas of human organism toxicity, i.e., toxicity to the liver, gastrointestinal tract, pulmonary system, and inflammation responses. The widely used model for hepatic metabolism and cytotoxicity studies are HepG2 cells which are derived from human liver carcinoma (Ahamed et al., 2020; Elje et al., 2020; Yuan et al., 2011). In particular, HepG2 cells are relevant for studying effects of nanoparticles that release metal ions when dissolving, such as ZnO, TiO2, Ag or Cu, due to their role in metabolizing reactive compounds and detoxifying oxidative stress (Brandão et al., 2020; Brkić Ahmed et al., 2017; Liu et al., 2021; Reihani et al., 2024). Compared to HepG2, HepaRG (Human Hepatocellular cells) are considered a superior model for studying hepatotoxicity (Guo et al., 2020). It is due to their ability to replicate the metabolic and inflammatory responses of liver cells in the human body towards various compounds, including NPs. Furthermore, their sensitivity to oxidative stress and metal ions makes them an ideal model used for distinguishing between direct cytotoxic effects and metabolism-dependent toxicity (Helvenstein et al., 2015; Kuhn et al., 2024; Voss et al., 2020). The response of gastrointestinal tract is modelled by HCT116 cell line which is derived from human colorectal carcinoma. HCT116 cells are often used to study the gastrointestinal toxicity of nanoparticles, particularly those administered orally or exposed via dietary sources. These cells are highly proliferative and exhibit strong DNA repair mechanisms, which make them a robust model for studying genotoxicity and oxidative stress (Ahmed et al., 2015; Gurunathan et al., 2018; Jia et al., 2020; Vecchiotti et al., 2021). A commonly used model for respiratory toxicity are A549 cells which are derived from human alveolar basal epithelial cells (Ahamed et al., 2019; Ahmad et al., 2017). A unique feature of A549 cells is their suitability for testing the particle size-dependent uptake and cytotoxicity, which is due to the larger nanoparticles remaining in the extracellular space, while smaller particles are translocated deeper and absorbed by cells (Chairuangkitti et al., 2013; Kim et al., 2006; Rafieepour et al., 2022). The model frequently used to investigate how the immune system reacts to nanoparticles are RAW264.7 cells, derived from murine macrophages. Besides their origin, it is due to their ability to produce high level of ROS when exposed to foreign substances (Bo et al., 2023). For that reason, they are recognized as a valuable model for researching inflammation and oxidative stress brought on by nanoparticles (Dussert et al., 2020; Giovanni et al., 2015; Hsieh et al., 2015). Data and Methods This work aimed to subject the publicly accessible dataset from the NanoCommons database (available under Creative Commons Attribution 4.0 International (CC BY 4.0)) to statistical analysis to verify how different parameters, specifically nanoparticle diameter, nanoparticle composition, and testing method influenced toxicity scores. Initially, the whole database was examined, counting in a total of 1,631 different measurements over 10 different research studies, 12 different examined compounds, and 6 different cell lines. After a more detailed examination, it became clear that measurements were observing different parameters. For instance, there were only 19 results for mortality or malformation and only 46 results for cell death, or cell death using the s.e.m method. For that reason, it was necessary to subject the whole dataset to a more detailed review to identify which parameter had the biggest representation and was able to interpret the toxicity of the studied compounds. The list of all available parameters in the NanoCommons database is displayed in Table 1. Table 1: List of parameters available in NanoCommons database Parameter Values present for given parameter Toxicity measurement 1631 _ToxExp.Sample 1631 _ToxExp.Dataset 1631 _ToxExp.Measurement Name 1631 _ToxExp.Time 1631 _ToxExp.Dosis [µg/ml] 1631 _ToxExp.Cell 1599 _ToxExp.Particle 1575 _data.Viable Cell Count.Value 1541 _data.Cell membrane damage.Value 1273 _data.Mitochondrial membrane potential.Value 1273 _data.Nuclear Intensity.Value 1273 _data.Nuclear Size.Value 1273 _data.Lysotracker.Value 273 _data.Cytoplasmic caspase 3.Value 261 _data.Fraction of caspase negative cells.Value 261 _data.Lipid droplet size.Value 261 _data.Nuclear caspase.Value 261 _data.Total caspase.Value 261 _data.Cell Death.Value 131 _data.Cell Death Hoechst PI [%].Value 131 _data.Cell Death [% of pos control].Value 46 _data.Cell Death s.e.m.Value 46 _data.IL-8 release [pg/ml].Value 46 _data.IL-8 release s.e.m..Value 46 _data.Viability [% of negative control].Value 46 _data.Viability s.e.m..Value 46 _data.Hatching inhibition [%].Value 33 _data.Malformation [%].Value 19 _data.Mortality [%].Value 19 _data.Nuclear caspase., 1 _data.Nuclear Intensity., 1 _data.Nuclear Size., 1 It was identified that the parameter with the biggest representation was data.Viable Cell Count.Value, which was present in 1,540 measurements out of 1,631, representing almost 95% coverage of the whole dataset. This parameter was selected as the main observed value, as it was also a commonly used indication of the toxicity of nanoparticles. The next step was to verify other parameters, namely: compound, exposure doses, exposed cell line, diameter of NPs, and exposure time, and check their presence in the dataset. After closer examination, it was found that 542 measurements had no information about nanoparticle diameter, so these measurements were removed from the dataset. The next refinement was focused on compounds, specifically on different doses, sufficient count of measurements, or different particle diameters. It was found that 48 measurements represented positive control, and 50 measurements of Au NPs had only one diameter – 12 nm on one cell line – HepaRG. Similarly, 10 measurements of Ag S (36 nm, HepaRG) and 10 measurements of AgO₂ (50 nm, HepaRG). These measurements were excluded from statistical analysis as they did not meet the required variation of parameters and values. During the final observation of the dataset, 880 measurements were preserved and subjected to statistical analysis, which accounted for roughly 54% of the original dataset. Statistical analysis The statistical analysis was performed in R-Studio. Plots were generated with the ggplot2 package, and an interactive chart was created using the Plotly package. Multiple Ordinary Least Squares (OLS) regression was employed to analyze the data. This statistical technique estimates the coefficients of the linear relationship between the dependent variable and several independent variables by minimizing the sum of the squared differences between the observed and predicted values. In this paper, the relationship between cell viability and multiple regressors, such as dose, particle diameter, compound, and cell line was modelled. For this analysis, two different model specifications were explored. The first was a Multiple OLS Model without interactions (Model 1), aimed at exploring the significance of certain variables and their effect on cell viability without considering any interactions. The second was a Multiple OLS Model with interactions (Model 2), which aimed to explore the significance of certain variables on cell viability while also considering interactions between some of the regressors. The structural form of Model 1 was defined by Formula 1: Cell Viability=β0+ β1 (log (Dose ) )+β2 (Particle Diameter )+β3 (Compound )+β4 (Cellline )+ϵ(1 Explanation of the Terms applicable for Formula 1:  β0: The intercept, representing cell viability when all independent variables are at their baseline values.  β1: The coefficient for the logarithmic exposure dose, showing how relative changes in the dose affect cell viability.  β2: The coefficient for particle diameter, indicating the effect of nanoparticle size on cell viability.  β3: The coefficient for the compound variable. Since compound is a categorical variable, β3 denotes a vector of estimates (for particular compounds).  β4: The coefficient for the exposed cell line variable. Similarly to β3, it is a vector of coefficients.  : The error term, capturing variability not explained by the model.ϵ The structural form of Model 2 was as follows: Cell Viability=β0+ β1 (log (Dose ) )+β2 (Particle Diameter )+β3 (Compound )+β4 (Cellline )+β5 (Particle Diameter )× (Compound )+ϵ (2 ) The difference between Model 1 (without interactions) and Model 2 (with interactions) was the addition of the interaction term between particle diameter and compounds. This enabled examination whether the effect of particle diameter was dependent on the compound or not. Besides that, the sensitivity analyses were conducted, by sub-setting the final sample. First, the relationship between the independent variables and cell viability by compounds was explored. The goal was to investigate whether the significance (and the direction of the effect) of the dose and particle diameter (when applicable) stayed the same as in the full model. The structural equation for the Multiple OLS model by compounds was as follows: Cell Viabilityi=β0+β1 (log (Dosei ))+β2 (Particle Diameter i)+ β3 (Cell linei )+ϵ i (3) where i=1 ,…,N . Second, the sample was divided based on the cell lines. The structural equation for Multiple OLS model by cell lines was as follows: Cell Viabilityi=β0+β1 (log (Dosei ))+β2 (Particle Diameter i)+ β3 (Compounds i )+ϵ i (4) where i=1 ,…,N . Explanation of the Terms applicable for Formula 3 & Formula 4:  β0: The intercept, representing cell viability when all independent variables are at their baseline values.  β1: The coefficient for the logarithmic exposure dose, showing how relative changes in the dose affect cell viability.  β2: The coefficient for particle diameter, indicating the effect of nanoparticle size on cell viability.  β3: The coefficient for the cell line (formula 3) or compound variable (formula 4).  ϵ i: The error term, capturing variability not explained by the model. Results The refined database can be found in the supplementary documents of this paper, and the 880 measurements from seven datasets contained five cell lines and six compounds, and their combinations is displayed in Table 2. Table 2: Compounds vs. Exposed Cell Line counts. ExposedCells/ Compounds A549 HCT116 HepaRG HEPG2 RAW264.7 Ag 16 16 50 16 16 CeO2 16 16 20 16 16 Cu 0 0 20 0 0 SiO2 24 24 170 24 24 TiO2 48 48 100 48 48 ZnO 16 16 40 16 16 Subsequently, the particle diameters ranged from 8 to 150 nm, and exposure doses ranged from 0.3 to 250 [µg/ml], as displayed in Table 3. Interestingly, ViableCellCount variable contained also positive values, which can be attributed to less toxic results compared to the control sample. Additionally, the exposure time was constant for all measurements – 24h. Table 3: Summary statistics of continuous variables used in the model. Statistic Min Mean Median Max Std Dev ExposureDosis 0.30 33.85 7.80 250.00 58.53 ParticleDiameter 8.00 39.86 20.00 150.00 43.57 ViableCellCount -26.98 -2.02 -0.64 2.43 4.38 ExposureTime 24.00 24.00 24.00 24.00 0.00 Multiple OLS Model (with and without interactions) The aim was to explore the significance of variables on cell viability variable (CVV), displayed in Table 4. Table 4 : Results of model with and without interactions Dependent variable: Viable Cell Count Model 1 Model 2 (1) (2) log(ExposureDosis) -1.002*** -1.002*** (0.063) (0.063) ParticleDiameter 0.011* 0.003 (0.006) (0.036) CompoundsCeO2 3.094*** 4.610*** (0.482) (1.740) CompoundsCu -5.821*** -14.424*** (0.823) (2.953) CompoundsSiO2 3.215*** 2.867*** (0.382) (0.915) CompoundsTiO2 2.494*** 2.417*** (0.369) (0.893) CompoundsZnO -2.369*** 3.408 (0.854) (9.477) ExposedCellHCT116 -0.823* -0.823* (0.430) (0.428) ExposedCellHepaRG -1.330*** -1.371*** (0.365) (0.367) ExposedCellHEPG2 0.980** 0.980** (0.436) (0.435) ExposedCellRAW264.7 0.608 0.608 (0.436) (0.435) ParticleDiameter:CompoundsCeO2 -0.056 (0.066) ParticleDiameter:CompoundsCu 0.776*** (0.250) ParticleDiameter:CompoundsSiO2 0.013 (0.036) ParticleDiameter:CompoundsTiO2 0.003 (0.037) ParticleDiameter:CompoundsZnO -0.033 (0.074) Constant -1.556*** -1.362 (0.444) (0.891) Observations 880 880 R2 0.429 0.437 Adjusted R2 0.422 0.426 Residual Std. Error 3.331 (df = 868) 3.317 (df = 863) F Statistic 59.266 *** (df = 11; 868) 41.842*** (df = 16; 863)   Note: *p<0.1; **p<0.05; ***p<0.01 The first variable, log(ExposureDosis), was negative and statistically significant for both models, having the same value (−1.002). A 100% increase in the exposure dose was associated with approximately a 1 percentage point decrease in CVV. Results for compounds highlight the varying effect of each compound in comparison to the base compound, which was Ag. Compared to Ag, CeO ,₂ SiO , and TiO had a positive and significant effect. On the other hand, Cu₂ ₂ and ZnO resulted in a strong negative effect, indicating a substantial reduction in cell viability compared to Ag. Interestingly, results for ZnO nanoparticles in Model 1 are significant (compared to Ag, ZnO is associated with a 2.369 percentage point decrease in CVV), whereas in Model 2, which contains the interaction with particle diameter, the estimates are not significant. For cell lines, the results were stable across both models in terms of significance and values. HCT116 and HepaRG cells displayed a negative effect on CVV and were statistically significant, while HEPG2 and RAW264.7 displayed a positive effect on CVV but lacked statistical significance, raising questions about the resilience (sensitivity) of specific cell lines to different compounds. The variable ParticleDiameter had a statistically significant positive effect (p < 0.05) in the model without interactions. However, in the model with interactions, the effect of the diameter and compounds became more complex, as the direct effect was not immediately apparent. We can derive the effect of diameter on CVV by taking the first derivative of CVV with respect to diameter (from Equation 2). The derivative is equal to + β₂ β5*Compound, hence, the effect of diameter is now dependent on the compound. For Ag, the effect of diameter is exactly , but thisβ₂ estimate is not significant. In contrast, diameter appears to play a role for Cu compounds. Increasing the diameter by 1 (or 10) nm is linked to a 0.776 (7.76) percentage point increase in CVV for Cu compared to Ag. Analogically, we can analyze the difference between Ag and the other compounds. By taking the first derivative of CVV with respect to the compound, we get + β₃ β5*Diameter. If the diameter was 0, there is no interaction effect and the effect of a compound on the outcome is exactly the respective estimate of . However, the minimum value of diameter isβ₃ 8 nm, meaning that the interaction term is taken into account. Particularly, the effect of Cu on CVV depends on the particle diameter as the interaction was significant. In other words, the effect of Cu on CVV can be computed as -14.424 + 0.776*diameter. As the particle diameter increases, the effect of Cu on CVV (compared to Ag) becomes less negative. If the diameter is 1 nm, the effect of Cu on CVV (compared to Ag and holding other variables constant) is -13.359. If the diameter is 10 nm, Cu would be associated with a 6.664 percentage point decrease in CVV compared to Ag. Multiple OLS Model (by compounds) The aim was to explore how the specific factors influence CVV separately for each compound (decreasing the sample size compared to the whole- sample regression) and whether these effects are consistent with the whole-sample results. The results by compound are displayed in Table 5 below. Table 5: Results of Multiple OLS model by compounds Dependent variable: Viable Cell Count TiO2 Ag SiO2 CeO2 ZnO Cu   (1) (2) (3) (4) (5) (6) log(ExposureDosi s) -0.759 *** -1.697*** -0.289*** -0.122*** -3.067*** -3.136*** (0.086) (0.162) (0.048) (0.040) (0.273) (0.547) ParticleDiameter 0.005 -0.006 0.010*** -0.053*** -0.030 0.779** (0.007) (0.034) (0.004) (0.010) (0.096) (0.362) ExposedCellHCT1 16 -0.525 2.079* -0.475 -0.202 -5.764*** (0.517) (1.082) (0.422) (0.221) (1.723) ExposedCellHepa RG -2.418 *** 2.519*** -1.853*** -0.736*** -1.972 (0.458) (0.916) (0.335) (0.215) (1.477) ExposedCellHEPG 2 -0.227 5.956*** -1.047** -0.506** 4.246** (0.527) (1.099) (0.425) (0.226) (1.753) ExposedCellRAW2 64.7 0.032 4.392 *** -1.339*** -0.860*** 3.033* (0.527) (1.099) (0.425) (0.226) (1.753) Constant 1.194*** -3.059*** 0.794** 1.816*** 6.386 -12.017** (0.411) (1.046) (0.313) (0.323) (13.918) (4.329) Observations 292 114 266 84 104 20 R2 0.359 0.551 0.272 0.463 0.625 0.688 Adjusted R2 0.345 0.526 0.255 0.421 0.602 0.651 Residual Std. Error 2.534 (df = 285) 3.061 (df = 107) 1.461 (df = 259) 0.625 (df = 77) 4.874 (df = 97) 4.852 (df = 17) F Statistic 26.592***  (df = 6; 285) 21.909***  (df = 6; 107) 16.139***  (df = 6; 259) 11.062***  (df = 6; 77) 26.928***  (df = 6; 97) 18.741***  (df = 2; 17) Note: *p<0.1; **p<0.05; ***p<0.01 The effect of log(ExposureDosis) was significant and negative (decreasing CVV) for all measurements, which was also by the previous results of models with/out interactions. The strongest effect was observed for ZnO (- 3.067) and Cu (-3.136), while the weakest effect was noted for CeO2 (- 0.122) and SiO2 (-0.289), which was primarily due to elemental characteristics. Regarding ParticleDiameter, there were contradictive results that varied across compounds. Only for three compounds, particle diameter was statistically significant; SiO2 and Cu showed a positive relationship with particle size, while CeO2 exhibited a negative effect. Positive effect indicated – the larger particles, the higher cell viability, the lower toxicity. The results from these regressions indicate that HepaRG cells were particularly sensitive within most compounds (showing significant negative effects on viability) compared to the base A549 cells. Specifically, HepaRG cell lines were associated with a 2.418, 1.853, and 0.736 percentage point decrease in CVV for TiO2, SiO2, and CeO2 (compared to A549 cell lines). This cell line was not significantly different from the base for ZnO, however, the HCT116 cell line was found to have a significantly negative effect on CVV (-5.764) in ZnO regression (compared to A549). Interestingly, ZnO NPs showed a positive and significant effect on HEPG2 and RAW267.7 cells, and Ag NPs were positive and significant for all cell lines, with the highest positive effect of 5.796 for HEPG2. These results indicate that these cell lines might have had higher resistance to these NPs, or in the case of lower doses, ZnO NPs or Ag NPs could have acted as a nutrient, promoting cell growth or metabolic activity, which resulted in positive outcomes. Multiple OLS Model (by cell lines) The aim was to explore how the specific factors influence CVV separately for each cell line and whether these effects are consistent with the whole- sample results. The regression results are displayed in Table 6 below. Table 6: Results of Multiple OLS model by cell lines   Dependent variable: Viable Cell Count A549 HCT116 HepaRG HEPG2 RAW264.7 (1) (2) (3) (4) (5) log(ExposureDosis ) -0.615*** -0.917*** -1.203*** -0.712*** -0.722*** (0.118) (0.148) (0.105) (0.087) (0.102) ParticleDiameter -0.004 -0.001 0.014 -0.009 -0.016 (0.013) (0.017) (0.009) (0.010) (0.011) CompoundsCeO2 5.556*** 3.252*** 4.136 *** 0.979* 2.231*** (0.738) (0.922) (1.102) (0.534) (0.626) CompoundsCu N/A N/A -6.033*** N/A N/A N/A N/A (1.110) N/A N/A CompoundsSiO2 5.905*** 3.351*** 3.207*** 0.559 1.831*** (0.669) (0.836) (0.691) (0.484) (0.567) CompoundsTiO2 5.632*** 3.022*** 1.915** 0.561 2.394*** (0.598) (0.748) (0.726) (0.433) (0.508) CompoundsZnO 4.091** -4.183* -4.162*** 1.375 2.540 (1.820) (2.276) (1.432) (1.317) (1.545) Constant -4.420*** -2.039*** -2.216*** 0.591 -0.817 (0.601) (0.751) (0.683) (0.470) (0.551) Observations 120 120 400 120 120 R2 0.544 0.584 0.427 0.390 0.421 Adjusted R2 0.519 0.562 0.417 0.357 0.390 Residual Std. Error 2.072 (df = 113) 2.590 (df = 113) 4.164 (df = 392) 1.499 (df = 113) 1.758 (df = 113) F Statistic 22.424*** ( df = 6; 113) 26.488*** ( df = 6; 113) 41.701 *** (df = 7; 392) 12.022*** ( df = 6; 113) 13.673*** ( df = 6; 113)   Note: *p<0.1; **p<0.05; ***p<0.01 Based on the results above, it is possible to see that the effect of log(ExposureDosis) is statistically significant and negative for all observations, indicates universal trend of lower CVV while increasing exposure dose. The strongest negative effect was observed for HepaRG cells (-1.203), while the weakest negative effect was displayed for A549 (- 0.615), emphasizing unique sensitivities of cell lines for specific exposure doses. Particle diameter was not found to be significantly associated to CVV in any of the cell line regressions. The significant positive coefficients for CeO2, SiO2, and TiO2 indicate lower toxicity in comparison to Ag across various cell lines (A549, HCT116, HepaRG and RAW264.7). The effect of ZnO on CVV, compared to Ag, was significantly different in A549, HCT116 and HEPARG regressions. For A549 cell line, the effect of ZnO NPs was positive (4.091), and for HCT116 (-4.183) and HepaRG (- 4.162), the effect was negative. One of possible explanation could be that when interacting with A549, ZnO NPs could promote cellular mechanisms to maintain viability, or for HCT116 and HepaRG could induce cytotoxic effects more strongly, leading to lower CVV. The last negative instance was observed for Cu NPs that were measured on HepaRG cell line, negatively affecting CVV by 6.033 percentage points. These results can be linked to either elemental characteristics or HepaRG liver cells being particularly vulnerable due to their high metabolic activity and susceptibility to oxidative damage and metal ion toxicity. Discussion The results above represent an exemplary attempt at a comprehensive statistical analysis of the data extracted from the NanoCommons database. Among a large number of available datasets, the attention is focused on six widely studied compounds in the form of NPs and their concentration effects on different cell lines while their size is also considered as the factor. The compounds selected from the database were zinc oxide (ZnO), copper (Cu), silver (Ag), titanium dioxide (TiO ), silicon₂ dioxide (SiO ), and cerium dioxide (CeO ), where each compound exhibited₂ ₂ distinct toxicological behaviours associated with its properties, such as ion release, oxidative stress induction, or photocatalytic activity. The logical counterpart in consideration of any toxic agent effects is the living target, represented in vitro by the various cell lines derived from different human tissues having thus different susceptibility and response to the questioned toxic agent. So, to evaluate the effect of the choice of the cell lines, the CVV data subsets for HepG2, HepaRG, A549, RAW264.7, and HCT116 were investigated. Prior to any other consideration, it should also be acknowledged that this study has several limitations. First, the analysis was conducted on the limited dataset, although being extracted from the dataset is as representative as possible. Moreover, factors such as shape, surface charge or modifications of NPs were not included in this study as they were not included in the original database. Next, the analysis is done in full awareness that the statistic evaluation enables us to find correlations without establishing causality relations. The results of statistical analysis of different variables affecting CVV displayed wide range of findings in terms of interactions of various compounds, cell lines and NPs size. Their graphical visualisation in the form of the interactive 3D chart is available in supplementary data of the paper. In all models, the exposure dose was consistently displaying significant negative effect on CVV, which is supporting the universal trend of increased toxicity at higher doses, as can be seen in Figure 1. Figure 1: Effect of ExposureDosis on CVV from perspective of Exposed Cells (left) and Compounds (right) Another principal view on the data shown in Figure 2 demonstrates the analysis of trends in CVV dependence on particle diameter. Figure 2: Effect of ParticleDiameter on CVV from perspective of Exposed Cells (left) and Compounds (right) The particle size, although displaying negligible overall effect on CVV, showed compounds-specific impact in the model with interactions, particularly for Cu NPs where the increase in particle size was associated with positive effect on CVV (0.776), thus decreasing the total negative effect of Cu NPs on CVV. The results from model without interactions displayed that Cu and ZnO were associated with negative impact on CVV as expected (Ahamed et al., 2010; Fahmy and Cormier, 2009; Fukui et al., 2012; Naz et al., 2020; Song et al., 2010) while CeO2, SiO2 and TiO2 were displaying positive effect on CVV, due to each compound characteristics – CeO2 possibly behaving as a prooxidant (Pansambal et al., 2023; Singh et al., 2020), and TiO2 possibly behaving relatively inert unless photo- activated (Arun et al., 2023; Baranowska-Wójcik et al., 2020; Ziental et al., 2020). In the case of SiO2 there is no clear explanation why SiO2 nanoparticles resulted to have positive effect on CVV. There are numerous studies displaying contradictory conclusions of SiO2 toxicity, primarily focusing on the factor of surface modification (Marzaioli et al., 2014; Nayl et al., 2022; Yang et al., 2018), which in this study, had not been examined. The multiple OLS model by compounds displayed that particle diameter was significant in three cases – for SiO2, CeO2 and Cu. The estimates were small for the first two, but for Cu NPs, the effect of particle size on CVV was significantly bigger compared to other compounds. For SiO2 and CeO2, the effect on all cell lines was negative, underlining the increased impact or higher sensitivity of selected cell lines within these two compounds. The multiple OLS model by cell lines showed that compared to Ag, there was a positive relationship between the remaining compounds and CVV for A549 and RAW264.7 regressions. It can be due to Ag characteristics (Chairuangkitti et al., 2013; Giovanni et al., 2015) where specifically in these measurements decreasing CVV more than other nanoparticles or indicating a possible resistance of A549 and RAW264.7. For the latter, there is no literature or research supporting this assumption. Next, only in three cases, the results surprisingly suggest a negative effect of a specific compound compared to Ag on CVV. In HEPARG regression, ZnO and Cu were found to decrease the CVV by 4.162 and 6.033 percentage points, respectively, compared to Ag. The next negative effect was also observed for ZnO NPs when exposed on HCT116 cells, decreasing CVV by 4.183. Since both ZnO and Cu can induce oxidative stress (Fukui et al., 2012; Song et al., 2010), the increased sensitivity of HCT116 and HepaRG cells could be attributed to their biological characteristics, with HepaRG liver cells being particularly vulnerable due to their high metabolic activity and susceptibility to oxidative damage and metal ion toxicity (Helvenstein et al., 2015; Kuhn et al., 2024; Voss et al., 2020). In summary, the results have indicated a hierarchy within the four investigated factors influencing the toxicity scores of NPs in CVV as illustrated in Figure 3. First, it is the primary pair of factors, i.e. the composition and the choice of the counterpart cell line, followed by the concentration (or dose) which governs the overall manifestation of the NPs’ toxicity in a significant decrease of CVV. The NPS size is of lesser importance, which means it governs the toxicity effect within the subset defined by the choice of the primary factor pair. This is contra-intuitive assuming the key role of size-related effects in nanomaterials, including the nanotoxicity as revealed by many experimental mechanistic studies (see (Macko et al., 2023) and references therein). It should be kept in mind that these results do not suggest that the size of NPs does not play any role in their toxicity mechanism (e.g. for Ag, TiO , and ZnO in this study),₂ but the effect of various sizes has not been significantly manifested at testing conditions used in the original studies collected and integrated into the NanoCommons knowledge base. Figure 3: Factors importance in nanoparticles toxicity Future research When it comes to the future research of nanoparticles toxicity, and at the same time, enrichment of NanoCommons database, it would be beneficial to have more data on toxicity of Alumina NPs (Al2O3), which can release Al³⁺ ions, which are known to be neurotoxic and can interfere with cellular functions (Gao et al., 2022; Park et al., 2016, 2015; Strigul et al., 2009). To complete the picture of the most produced nanoparticles, Carbonaceous NPs should be subject to continuous research, as these also poses toxicological properties (Awasthi et al., 2024; Uo et al., 2011). Another addition to the database could be done on ultrafine dust particles (UFPs) collected at different locations. UFPs consist of particles below 100 nanometres, and pose various risks (cancer, neurological or respiratory health issues), at the same time, it is the most common nanoparticle in the environment (Hammond et al., 2022; Heinzerling et al., 2015; Kolling et al., 2011; Ophir et al., 2019). For the reasons above, addition and toxicological research on the abovementioned elements could enhance the relevance of the database and its metanalysis. Conclusion This study provides insights into the relationships between nanoparticle characteristics and cell viability. The effects of variables – NPs composition, cell line, dose and NPs size were explored via two main models – the first model was a Multiple OLS Model without interactions (Model 1), aimed at exploring the significance of certain variables and their effect on cell viability without considering any interactions. The second was a Multiple OLS Model with interactions (Model 2), which aimed to explore the significance of certain variables on cell viability while also considering interactions between some of the regressors. Subsequently, the sensitivity analyses were conducted by sub-setting the final sample. First, the relationship between the independent variables and cell viability by compounds was explored. Second, the sample was divided based on cell lines. The results displayed significant dose dependency on CVV across all models and confirmed the universal trend of increased toxicity at higher doses. The particle size had a negligible effect on CVV, but compound-specific impacts were observed. Overall, the results did not provide a clear trend on how other variables, such as cell lines or particle size, influence CVV but can serve as another supporting material in understanding nanoparticle toxicity. The factors can be ordered within a hierarchy of influence. The primary pair of factors is the composition and the choice of the counterpart cell line for testing, followed by the secondary factor of concentration (or dose), which governs the overall manifestation of the NPs’ toxicity in a significant decrease of CVV. The tertiary factor of NPs size is manifested of lesser importance, which might be rather due to the testing conditions than due to the irrelevance of the size at all. This study also demonstrated the usefulness of data collections present in the NanoCommons database and indicates its future potential for data extraction in the field of nanotoxicity. Moreover, the amount of gathered FAIR (Findable, Accessible, Interoperate, and Reusable) data surpasses the magnitude of any individual experimental study. The observed correlations may be a stepping stone for targeting new studies and developing more sensitive nanotoxicity testing methods to overbridge the gap between mechanistic and statistical screening studies. Acknowledgments Tomas Bata University in Zlin is acknowledged for all support provided in kind. Fundings The Ministry of Education, Youth and Sports of the Czech Republic supported this study financially within the framework of DKRVO (project number RP/CPS/2024-28/007). References Abbasi, R., Shineh, G., Mobaraki, M., Doughty, S., Tayebi, L., 2023. Structural parameters of nanoparticles affecting their toxicity for biomedical applications: a review. Journal of Nanoparticle Research 25, 43. https://doi.org/10.1007/s11051-023-05690-w Ahamed, M., Akhtar, M.J., Alhadlaq, H.A., 2020. Influence of silica nanoparticles on cadmium- induced cytotoxicity, oxidative stress, and apoptosis in human liver HepG2 cells. Environ Toxicol 35, 599–608. https://doi.org/10.1002/tox.22895 Ahamed, M., Akhtar, M.J., Alhadlaq, H.A., 2019. Preventive effect of TiO2 nanoparticles on heavy metal Pb-induced toxicity in human lung epithelial (A549) cells. Toxicology in Vitro 57, 18–27. https://doi.org/https://doi.org/10.1016/j.tiv.2019.02.004 Ahamed, M., Siddiqui, M.A., Akhtar, M.J., Ahmad, I., Pant, A.B., Alhadlaq, H.A., 2010. Genotoxic potential of copper oxide nanoparticles in human lung epithelial cells. Biochem Biophys Res Commun 396, 578–583. https://doi.org/https://doi.org/10.1016/j.bbrc.2010.04.156 Ahmad, J., Siddiqui, M.A., Akhtar, M.J., Alhadlaq, H.A., Alshamsan, A., Khan, S.T., Wahab, R., Al-Khedhairy, A.A., Al-Salim, A., Musarrat, J., Saquib, Q., Fareed, M., Ahamed, M., 2017. Copper doping enhanced the oxidative stress–mediated cytotoxicity of TiO2 nanoparticles in A549 cells. Hum Exp Toxicol 37, 496–507. https://doi.org/10.1177/0960327117714040 Ahmed, F.R., Shoaib, M.H., Azhar, M., Um, S.H., Yousuf, R.I., Hashmi, S., Dar, A., 2015. In- vitro assessment of cytotoxicity of halloysite nanotubes against HepG2, HCT116 and human peripheral blood lymphocytes. Colloids Surf B Biointerfaces 135, 50–55. https://doi.org/https://doi.org/10.1016/j.colsurfb.2015.07.021 Akintelu, S.A., Folorunso, A.S., Folorunso, F.A., Oyebamiji, A.K., 2020. Green synthesis of copper oxide nanoparticles for biomedical application and environmental remediation. Heliyon 6. https://doi.org/10.1016/j.heliyon.2020.e04508 Arun, J., Nachiappan, S., Rangarajan, G., Alagappan, R.P., Gopinath, K.P., Lichtfouse, E., 2023. Synthesis and application of titanium dioxide  photocatalysis for energy, decontamination and viral disinfection: a review. Environ Chem Lett 21, 339–362. https://doi.org/10.1007/s10311- 022-01503-z Awasthi, S., Srivastava, A., Kumar, D., Pandey, S.K., Mubarak, N.M., Dehghani, M.H., Ansari, K., 2024. An insight into the toxicological impacts of carbon nanotubes (CNTs) on human health: A review. Environmental Advances 18, 100601. https://doi.org/https://doi.org/10.1016/j.envadv.2024.100601 Baranowska-Wójcik, E., Szwajgier, D., Oleszczuk, P., Winiarska-Mieczan, A., 2020. Effects of Titanium Dioxide Nanoparticles Exposure on Human Health—a Review. Biol Trace Elem Res 193, 118–129. https://doi.org/10.1007/s12011-019-01706-6 Bezza, F.A., Tichapondwa, S.M., Chirwa, E.M.N., 2020. Fabrication of monodispersed copper oxide nanoparticles with potential application as antimicrobial agents. Sci Rep 10, 16680. https://doi.org/10.1038/s41598-020-73497-z BioXMTM Knowledge Portal." NanoCommons [WWW Document], 2024. . https://ssl.biomax.de/nanocommons/cgi/login_bioxm_portal.cgi. Bo, R., Wang, J., Rui, L., Liu, X., Li, Jiawen, Tao, Y., Hong, H., Xu, S., Huang, M., Liu, M., Pang, H., Li, Jingui, 2023. Immunoregulatory effects on RAW264.7 cells and subacute oral toxicity of ultra-large pore mesoporous silica nanoparticles loading Lycium barbarum polysaccharides. J Drug Deliv Sci Technol 82, 104192. https://doi.org/https://doi.org/10.1016/j.jddst.2023.104192 Brandão, F., Fernández-Bertólez, N., Rosário, F., Bessa, M.J., Fraga, S., Pásaro, E., Teixeira, J.P., Laffon, B., Valdiglesias, V., Costa, C., 2020. Genotoxicity of TiO2 Nanoparticles in Four Different Human Cell Lines (A549, HEPG2, A172 and SH-SY5Y). Nanomaterials 10. https://doi.org/10.3390/nano10030412 Brkić Ahmed, L., Milić, M., Pongrac, I.M., Marjanović, A.M., Mlinarić, H., Pavičić, I., Gajović, S., Vinković Vrček, I., 2017. Impact of surface functionalization on the uptake mechanism and toxicity effects of silver nanoparticles in HepG2 cells. Food and Chemical Toxicology 107, 349–361. https://doi.org/https://doi.org/10.1016/j.fct.2017.07.016 Bruna, T., Maldonado-Bravo, F., Jara, P., Caro, N., 2021. Silver nanoparticles and their antibacterial applications. Int J Mol Sci. https://doi.org/10.3390/ijms22137202 Canta, M., Cauda, V., 2020. The investigation of the parameters affecting the ZnO nanoparticle cytotoxicity behaviour: a tutorial review. Biomater Sci 8, 6157–6174. https://doi.org/10.1039/D0BM01086C Chairuangkitti, P., Lawanprasert, S., Roytrakul, S., Aueviriyavit, S., Phummiratch, D., Kulthong, K., Chanvorachote, P., Maniratanachote, R., 2013. Silver nanoparticles induce toxicity in A549 cells via ROS-dependent and ROS-independent pathways. Toxicology in Vitro 27, 330– 338. https://doi.org/https://doi.org/10.1016/j.tiv.2012.08.021 Chakraborty, N., Banerjee, J., Chakraborty, P., Banerjee, A., Chanda, S., Ray, K., Acharya, K., Sarkar, J., 2022. Green synthesis of copper/copper oxide nanoparticles and their applications: a review. Green Chem Lett Rev 15, 187–215. https://doi.org/10.1080/17518253.2022.2025916 Chang, Y.N., Zhang, M., Xia, L., Zhang, J., Xing, G., 2012. The toxic effects and mechanisms of CuO and ZnO nanoparticles. Materials. https://doi.org/10.3390/ma5122850 Chen, P., Wang, H., He, M., Chen, B., Yang, B., Hu, B., 2019. Size-dependent cytotoxicity study of ZnO nanoparticles in HepG2 cells. Ecotoxicol Environ Saf 171, 337–346. https://doi.org/https://doi.org/10.1016/j.ecoenv.2018.12.096 Dussert, F., Arthaud, P.-A., Arnal, M.-E., Dalzon, B., Torres, A., Douki, T., Herlin, N., Rabilloud, T., Carriere, M., 2020. Toxicity to RAW264.7 Macrophages of Silica Nanoparticles and the E551 Food Additive, in Combination with Genotoxic Agents. Nanomaterials 10. https://doi.org/10.3390/nano10071418 Elje, E., Mariussen, E., Moriones, O.H., Bastús, N.G., Puntes, V., Kohl, Y., Dusinska, M., Rundén-Pran, E., 2020. Hepato(Geno)Toxicity Assessment of Nanoparticles in a HepG2 Liver Spheroid Model. Nanomaterials 10. https://doi.org/10.3390/nano10030545 El-Sayed, R.A., El-Demerdash, F.M., El-Magd, M.A., 2021. Ginseng ameliorates pulmonary toxicity induced by silicon dioxide nanoparticles in rats. Asian Pac J Trop Biomed 11. ESKİN, A., 2022. Effects of silicon dioxide nanoparticles (SiO2 NPs) on total hemocyte count and hemocyte viability of Galleria mellonella. Int J Trop Insect Sci 42, 2617–2623. https://doi.org/10.1007/s42690-022-00790-x Fahmy, B., Cormier, S.A., 2009. Copper oxide nanoparticles induce oxidative stress and cytotoxicity in airway epithelial cells. Toxicology in Vitro 23, 1365–1371. https://doi.org/https://doi.org/10.1016/j.tiv.2009.08.005 Forest, V., Leclerc, L., Hochepied, J.-F., Trouvé, A., Sarry, G., Pourchez, J., 2017. Impact of cerium oxide nanoparticles shape on their in vitro cellular toxicity. Toxicology in Vitro 38, 136–141. https://doi.org/https://doi.org/10.1016/j.tiv.2016.09.022 Fukui, H., Horie, M., Endoh, S., Kato, H., Fujita, K., Nishio, K., Komaba, L.K., Maru, J., Miyauhi, A., Nakamura, A., Kinugasa, S., Yoshida, Y., Hagihara, Y., Iwahashi, H., 2012. Association of zinc ion release and oxidative stress induced by intratracheal instillation of ZnO nanoparticles to rat lung. Chem Biol Interact 198, 29–37. https://doi.org/https://doi.org/10.1016/j.cbi.2012.04.007 Gao, Y., Chen, K., Ma, J., Gao, F., 2014. Cerium oxide nanoparticles in cancer. Onco Targets Ther 7, 835–840. https://doi.org/10.2147/OTT.S62057 Gao, Y., Wen, P., Chen, H., Wei, Y., Cui, H., Ma, J., Li, J., Qin, G., 2022. Teratogenicity of 30 nm Aluminum Oxide Nanoparticles (Al2O3NPs) in Rats by Gavage. Biol Trace Elem Res 200, 4027–4034. https://doi.org/10.1007/s12011-021-03004-6 Giovanni, M., Yue, J., Zhang, L., Xie, J., Ong, C.N., Leong, D.T., 2015. Pro-inflammatory responses of RAW264.7 macrophages when treated with ultralow concentrations of silver, titanium dioxide, and zinc oxide nanoparticles. J Hazard Mater 297, 146–152. https://doi.org/https://doi.org/10.1016/j.jhazmat.2015.04.081 Gudkov, S. V, Burmistrov, D.E., Serov, D.A., Rebezov, M.B., Semenova, A.A., Lisitsyn, A.B., 2021. A Mini Review of Antibacterial Properties of ZnO Nanoparticles. Front Phys 9. Guo, X., Seo, J.-E., Petibone, D., Tryndyak, V., Lee, U., Zhou, T., Robison, T., Mei, N., 2020. Performance of HepaRG and HepG2 cells in the high-throughput micronucleus assay for in vitro genotoxicity assessment. J Toxicol Environ Health A 83, 1–16. https://doi.org/10.1080/15287394.2020.1822972 Gupta, R., Xie, H., 2018. Nanoparticles in Daily Life: Applications, Toxicity and Regulations. JOURNAL OF ENVIRONMENTAL PATHOLOGY TOXICOLOGY AND ONCOLOGY 37, 209–230. https://doi.org/10.1615/JEnvironPatholToxicolOncol.2018026009 Gurunathan, S., Qasim, M., Park, C., Yoo, H., Kim, J.-H., Hong, K., 2018. Cytotoxic Potential and Molecular Pathway Analysis of Silver Nanoparticles in Human Colon Cancer Cells HCT116. Int J Mol Sci 19. https://doi.org/10.3390/ijms19082269 Hammond, J., Maher, B.A., Gonet, T., Bautista, F., Allsop, D., 2022. Oxidative Stress, Cytotoxic and Inflammatory Effects of Urban Ultrafine Road-Deposited Dust from the UK and Mexico in Human Epithelial Lung (Calu-3) Cells. Antioxidants 11. https://doi.org/10.3390/antiox11091814 Heinzerling, A., Hsu, J., Yip, F., 2015. Respiratory Health Effects of Ultrafine Particles in Children: a Literature Review. Water Air Soil Pollut 227, 32. https://doi.org/10.1007/s11270- 015-2726-6 Helvenstein, M., Stanicki, D., Laurent, S., Blankert, B., 2015. Interaction between Iron oxide nanoparticles and HepaRG cells: A preliminary in vitro evaluation. J Nanomater 2015. https://doi.org/10.1155/2015/463841 Hsieh, H.-C., Chen, C.-M., Hsieh, W.-Y., Chen, C.-Y., Liu, C.-C., Lin, F.-H., 2015. ROS-induced toxicity: exposure of 3T3, RAW264.7, and MCF7 cells to superparamagnetic iron oxide nanoparticles results in cell death by mitochondria-dependent apoptosis. Journal of Nanoparticle Research 17, 71. https://doi.org/10.1007/s11051-015-2886-8 Islam, F., Shohag, S., Uddin, M.J., Islam, M.R., Nafady, M.H., Akter, A., Mitra, S., Roy, A., Bin Emran, T., Cavalu, S., 2022. Exploring the Journey of Zinc Oxide Nanoparticles (ZnO-NPs) toward Biomedical Applications. Materials. https://doi.org/10.3390/ma15062160 Jia, M., Zhang, W., He, T., Shu, M., Deng, J., Wang, J., Li, W., Bai, J., Lin, Q., Luo, F., Zhou, W., Zeng, X., 2020. Evaluation of the Genotoxic and Oxidative Damage Potential of Silver Nanoparticles in Human NCM460 and HCT116 Cells. Int J Mol Sci 21. https://doi.org/10.3390/ijms21051618 Kim, J.-S., Yoon, T.-J., Yu, K.-N., Noh, M.S., Woo, M., Kim, B.-G., Lee, K.-H., Sohn, B.-H., Park, S.-B., Lee, J.-K., Cho, M.-H., 2006. Cellular uptake of magnetic nanoparticle is mediated through energy-dependent endocytosis in A549 cells. J Vet Sci 7, 321–326. Kolling, A., Heinrich, E., Susanne, R., and Heinrich, U., 2011. Relationship of pulmonary toxicity and carcinogenicity of fine and ultrafine granular dusts in a rat bioassay. Inhal Toxicol 23, 544–554. https://doi.org/10.3109/08958378.2011.594458 Kuhn, J., McDonald, A., Mongoin, C., Anderson, G., Lafeuillade, G., Mitchell, S., Elfick, A.P.D., Bagnaninchi, P.O., Yiu, H.H.P., Nelson, L.J., 2024. Non-invasive methods of monitoring Fe3O4 magnetic nanoparticle toxicity in human liver HepaRG cells using impedance biosensing and Coherent anti-Stokes Raman spectroscopic (CARS) microscopy. Toxicol Lett 394, 92–101. https://doi.org/https://doi.org/10.1016/j.toxlet.2024.02.010 Kumar, V., Sharma, N., Maitra, S.S., 2017. In vitro and in vivo toxicity assessment of nanoparticles. Int Nano Lett 7, 243–256. https://doi.org/10.1007/s40089-017-0221-3 Liman, R., Acikbas, Y., Ciğerci, İ.H., Ali, M.M., Kars, M.D., 2020. Cytotoxic and Genotoxic Assessment of Silicon Dioxide Nanoparticles by Allium and Comet Tests. Bull Environ Contam Toxicol 104, 215–221. https://doi.org/10.1007/s00128-020-02783-3 Liu, Y., Zeng, Z., Jiang, O., Li, Y.X., Xu, Q., Jiang, L.J., Yu, J., Xu, D., 2021. Green synthesis of CuO NPs, characterization and their toxicity potential against HepG2 cells. Mater Res Express 8. https://doi.org/10.1088/2053-1591/abd666 Luceri, A., Francese, R., Lembo, D., Ferraris, M., Balagna, C., 2023. Silver Nanoparticles: Review of Antiviral Properties, Mechanism of Action and Applications. Microorganisms 11. https://doi.org/10.3390/microorganisms11030629 Macko, M., Antoš, J., Božek, F., Konečný, J., Huzlík, J., Hegrová, J., Kuřitka, I., 2023. Development of New Health Risk Assessment of Nanoparticles: EPA Health Risk Assessment Revised. Nanomaterials 13. https://doi.org/10.3390/nano13010020 Marzaioli, V., Aguilar-Pimentel, J.A., Weichenmeier, I., Luxenhofer, G., Wiemann, M., Landsiedel, R., Wohlleben, W., Eiden, S., Mempel, M., Behrendt, H., Schmidt-Weber, C., Gutermuth, J., Alessandrini, F., 2014. Surface modifications of silica nanoparticles are crucial for their inert versus proinflammatory and immunomodulatory properties. Int J Nanomedicine 9, 2815–2827. https://doi.org/10.2147/IJN.S57396 Mohamed, H.R.H., 2015. Estimation of TiO2 nanoparticle-induced genotoxicity persistence and possible chronic gastritis-induction in mice. Food and Chemical Toxicology 83, 76–83. https://doi.org/https://doi.org/10.1016/j.fct.2015.05.018 Naika, H.R., Lingaraju, K., Manjunath, K., Kumar, D., Nagaraju, G., Suresh, D., Nagabhushana, H., 2015. Green synthesis of CuO nanoparticles using Gloriosa superba L. extract and their antibacterial activity. Journal of Taibah University for Science 9, 7–12. https://doi.org/10.1016/j.jtusci.2014.04.006 NanoCommons: The European Nanotechnology Knowledge Infrastructure [WWW Document], 2025. . https://www.nanocommons.eu/. Nayl, A.A., Abd-Elhamid, A.I., Aly, A.A., Bräse, S., 2022. Recent progress in the applications of silica-based nanoparticles. RSC Adv. https://doi.org/10.1039/d2ra01587k Naz, S., Gul, A., Zia, M., 2020. Toxicity of copper oxide nanoparticles: a review study. IET Nanobiotechnol 14, 1–13. https://doi.org/https://doi.org/10.1049/iet-nbt.2019.0176 Nie, P., Zhao, Y., Xu, H., 2023. Synthesis, applications, toxicity and toxicity mechanisms of silver nanoparticles: A review. Ecotoxicol Environ Saf 253, 114636. https://doi.org/https://doi.org/10.1016/j.ecoenv.2023.114636 Nosaka, Y., Nosaka, A.Y., 2017. Generation and Detection of Reactive Oxygen Species in Photocatalysis. Chem Rev 117, 11302–11336. https://doi.org/10.1021/acs.chemrev.7b00161 Ophir, N., Bar Shai, A., Korenstein, R., Kramer, M.R., Fireman, E., 2019. Functional, inflammatory and interstitial impairment due to artificial stone dust ultrafine particles exposure. Occup Environ Med 76, 875–879. https://doi.org/10.1136/oemed-2019-105711 Pansambal, S., Oza, R., Borgave, S., Chauhan, A., Bardapurkar, P., Vyas, S., Ghotekar, S., 2023. Bioengineered cerium oxide (CeO2) nanoparticles and their diverse applications: a review. Appl Nanosci 13, 6067–6092. https://doi.org/10.1007/s13204-022-02574-8 Park, E.-J., Lee, G.-H., Yoon, C., Jeong, U., Kim, Y., Cho, M.-H., Kim, D.-W., 2016. Biodistribution and toxicity of spherical aluminum oxide nanoparticles. Journal of Applied Toxicology 36, 424–433. https://doi.org/https://doi.org/10.1002/jat.3233 Park, E.-J., Sim, J., Kim, Y., Han, B.S., Yoon, C., Lee, S., Cho, M.-H., Lee, B.-S., Kim, J.-H., 2015. A 13-week repeated-dose oral toxicity and bioaccumulation of aluminum oxide nanoparticles in mice. Arch Toxicol 89, 371–379. https://doi.org/10.1007/s00204-014-1256-0 Rafieepour, A., Azari, M.R., Khodagholi, F., Jaktaji, J.P., Mehrabi, Y., Peirovi, H., 2021. Interactive toxicity effect of combined exposure to hematite and amorphous silicon dioxide nanoparticles in human A549 cell line. Toxicol Ind Health 37, 289–302. https://doi.org/10.1177/07482337211002373 Rafieepour, A., R Azari, M., Khodagholi, F., 2022. Cytotoxic effects of crystalline silica in form of micro and nanoparticles on the human lung cell line A549. Toxicol Ind Health 39, 23–35. https://doi.org/10.1177/07482337221140644 Reihani, A., Shaki, F., Azari, A., 2024. Zinc oxide nanoparticles decrease acrylamide cytotoxicity and oxidative stress in HepG2 cells. Nutr Food Sci. https://doi.org/10.1108/NFS-07-2023- 0147 Samuel, M.S., Ravikumar, M., John J., A., Selvarajan, E., Patel, H., Chander, P.S., Soundarya, J., Vuppala, S., Balaji, R., Chandrasekar, N., 2022. A Review on Green Synthesis of Nanoparticles and Their Diverse Biomedical and Environmental Applications. Catalysts 12. https://doi.org/10.3390/catal12050459 Savage, D.T., Hilt, J.Z., Dziubla, T.D., 2019. In vitro methods for assessing nanoparticle toxicity, in: Methods in Molecular Biology. Humana Press Inc., pp. 1–29. https://doi.org/10.1007/978- 1-4939-8916-4_1 Silva, B.L. da, Abuçafy, M.P., Manaia, E.B., Junior, J.A.O., Chiari-Andréo, B.G., Pietro, R.C.L.R., Chiavacci, L.A., 2019. Relationship between structure and antimicrobial activity of zinc oxide nanoparticles: An overview. Int J Nanomedicine. https://doi.org/10.2147/IJN.S216204 Singh, K.R.B., Nayak, V., Sarkar, T., Singh, R.P., 2020. Cerium oxide nanoparticles: Properties, biosynthesis and biomedical application. RSC Adv. https://doi.org/10.1039/d0ra04736h Slowing, I.I., Vivero-Escoto, J.L., Wu, C.-W., Lin, V.S.-Y., 2008. Mesoporous silica nanoparticles as controlled release drug delivery and gene transfection carriers. Adv Drug Deliv Rev 60, 1278–1288. https://doi.org/https://doi.org/10.1016/j.addr.2008.03.012 Song, W., Zhang, Jinyang, Guo, J., Zhang, Jinhua, Ding, F., Li, L., Sun, Z., 2010. Role of the dissolved zinc ion and reactive oxygen species in cytotoxicity of ZnO nanoparticles. Toxicol Lett 199, 389–397. https://doi.org/https://doi.org/10.1016/j.toxlet.2010.10.003 Strigul, N., Vaccari, L., Galdun, C., Wazne, M., Liu, X., Christodoulatos, C., Jasinkiewicz, K., 2009. Acute toxicity of boron, titanium dioxide, and aluminum nanoparticles to Daphnia magna and Vibrio fischeri. Desalination 248, 771–782. https://doi.org/https://doi.org/10.1016/j.desal.2009.01.013 Thakur, Naveen, Thakur, Nikesh, Kumar, Anil, Thakur, V.K., Kalia, S., Arya, V., Kumar, Ashwani, Kumar, S., Kyzas, G.Z., 2024. A critical review on the recent trends of photocatalytic, antibacterial, antioxidant and nanohybrid applications of anatase and rutile TiO2 nanoparticles. Science of The Total Environment 914, 169815. https://doi.org/https://doi.org/10.1016/j.scitotenv.2023.169815 Uo, M., Akasaka, T., Watari, F., Sato, Y., Tohji, K., 2011. Toxicity evaluations of various carbon nanomaterials. Dent Mater J. https://doi.org/10.4012/dmj.2010-039 Vecchiotti, G., Colafarina, S., Aloisi, M., Zarivi, O., Di Carlo, P., Poma, A., 2021. Genotoxicity and oxidative stress induction by polystyrene nanoparticles in the colorectal cancer cell line HCT116. PLoS One 16, e0255120-. Voss, L., Yilmaz, K., Burkard, L., Vidmar, J., Stock, V., Hoffmann, U., Pötz, O., Hammer, H.S., Peiser, M., Braeuning, A., Löschner, K., Böhmert, L., Sieg, H., 2020. Impact of iron oxide nanoparticles on xenobiotic metabolism in HepaRG cells. Arch Toxicol 94, 4023–4035. https://doi.org/10.1007/s00204-020-02904-1 Xu, L., Wang, Y.Y., Huang, J., Chen, C.Y., Wang, Z.X., Xie, H., 2020. Silver nanoparticles: Synthesis, medical applications and biosafety. Theranostics. https://doi.org/10.7150/thno.45413 Yang, S.A., Choi, S., Jeon, S.M., Yu, J., 2018. Silica nanoparticle stability in biological media revisited. Sci Rep 8. https://doi.org/10.1038/s41598-017-18502-8 Yang, Y., Du, X., Wang, Q., Liu, J., Zhang, E., Sai, L., Peng, C., Lavin, M.F., Yeo, A.J., Yang, X., Shao, H., Du, Z., 2019. Mechanism of cell death induced by silica nanoparticles in hepatocyte cells is by apoptosis. Int J Mol Med 44, 903–912. https://doi.org/10.3892/ijmm.2019.4265 Yip, S.K., Sauls, J.A., 1992. Nonlinear Meissner effect in CuO superconductors. Phys Rev Lett 69, 2264–2267. https://doi.org/10.1103/PhysRevLett.69.2264 Yuan, Y., Liu, C., Lu, J., Tang, W., Gan, Q., Zhou, H., Qian, J., Lu, X., 2011. In vitro cytotoxicity and induction of apoptosis by silica nanoparticles in human HepG2 hepatoma cells. Int J Nanomedicine 1889. https://doi.org/10.2147/ijn.s24005 Ziental, D., Czarczynska-Goslinska, B., Mlynarczyk, D.T., Glowacka-Sobotta, A., Stanisz, B., Goslinski, T., Sobotta, L., 2020. Titanium Dioxide Nanoparticles: Prospects and Applications in Medicine. Nanomaterials 10. https://doi.org/10.3390/nano10020387