What makes nanoparticles different from the same material in bulk form? Two fundamental physics phenomena: an enormously increased surface-to-volume ratio (meaning a far higher proportion of atoms sit exposed on the surface, dramatically altering chemical reactivity), and the dominance of quantum effects at the nanoscale, which can change properties like conductivity, optical transparency, and magnetism. A material that is chemically inert at the gram scale may behave very differently at the nanometer scale. That is precisely what makes nanomaterials useful — and precisely what makes their toxicology unpredictable.

The research selected six particles with strong industrial relevance, all produced and characterized by the Fraunhofer Institute for Ceramic Technologies and Systems (IKTS) in Dresden:

Cell types chosen to model human exposure routes included skin keratinocytes (HaCaT), lung epithelial cells (A549), and colon cells (CaCo-2) — representing the three main routes by which industrial workers encounter nanoparticles: dermal contact, inhalation, and accidental ingestion. For environmental relevance, a rainbow trout gill cell line (RTgill-W1) was also included.

The first and perhaps most startling result of the study: every type of nanoparticle tested entered every type of cell tested. This sounds almost self-evident, but it challenges a long-held assumption in cell biology.

Conventional wisdom held that only specialized phagocytic cells — macrophages, monocytes — were capable of engulfing foreign particles. The immune system's "cleaners." Non-phagocytic cells (skin cells, lung lining cells, gut epithelial cells) were thought to be largely impermeable to particles of any size. This study contradicts that view decisively.

Using a combination of light microscopy, transmission electron microscopy coupled with energy-dispersive X-ray spectroscopy (EDX), and flow cytometry, the researchers confirmed particle uptake into all seven cell types studied. EDX was critical here — it allowed chemical element identification, confirming that the objects seen inside cells were indeed the test particles and not preparation artefacts. Live-cell imaging further revealed the time-dependent process of uptake in real time.

Particles were consistently found in the cytoplasm — the cell's main interior compartment — and co-localised with lysosomal structures. Lysosomes are the cell's degradation machinery: membrane-bound organelles filled with enzymes that break down waste and foreign material. Finding nanoparticles accumulating there raises an important question: if the lysosomes cannot degrade the particles (and many engineered nanoparticles are specifically designed for their durability), they may simply accumulate. Long-term consequences are unknown.

Crucially, particles were never found in cell nuclei. This is significant because the nucleus houses DNA, and nuclear entry would dramatically increase genotoxic risk. The nuclear pore complex — the gated channel through the nuclear membrane — appears to act as a reliable size barrier. However, the researchers note that this does not eliminate the possibility of indirect DNA damage through other pathways.

Flow cytometry provided a semi-quantitative measure: particles inside cells increase cellular "granularity" — a light-scatter signal that changes measurably. Changes in granularity were generally more dependent on particle type than cell type, suggesting that particle properties (size, surface chemistry, charge) matter more than the biological identity of the receiving cell. Only primary particle size showed a weak correlation with uptake rates after short exposure times.

The researchers also tested whether blocking actin filaments (using Cytochalasin D, an inhibitor of phagocytosis) would stop particle entry. It didn't — at least not fully. In most cell types, the inhibitor had no significant effect on uptake, reinforcing the conclusion that nanoparticles enter cells through multiple pathways, not exclusively the actin-dependent phagocytic route typically reserved for immune cells.

Having established that particles enter cells, the next question is: what do they do there? For WC and WC-Co nanoparticles, the findings were striking.

Cells were exposed to three conditions: WC nanoparticles alone, WC-Co nanoparticles, and dissolved cobalt chloride (CoCl₂) at concentrations equivalent to the cobalt content of the WC-Co particles. Cell viability (metabolic activity via AlamarBlue) and membrane integrity (via CFDA assay) were the key readouts.

This pattern echoes earlier work with micro-sized WC-Co particles, where Dominique Lison's group in Belgium demonstrated similar enhanced toxicity. The leading hypothesis involves the particle surface: the interface between WC and Co may catalyze the generation of reactive oxygen species (ROS) — chemically reactive molecules that damage cell membranes, proteins, and DNA. Additionally, the "Trojan Horse" hypothesis (first proposed by Limbach et al., 2007) suggests that nanoparticles may carry cobalt ions directly into cells, bypassing the charge-based barriers that normally limit ion entry into the cytoplasm, dramatically increasing intracellular cobalt concentrations.

A notable methodological observation: serum albumin in culture media bound to particle surfaces and prevented agglomeration — a finding with important implications for both laboratory testing and in-vivo relevance, since blood and interstitial fluids also contain albumin.

A common assumption in nanotoxicology is that nanoparticles, when they clump together into larger aggregates (agglomerates), effectively lose their nano-properties and become less biologically active. This study tests that assumption directly.

When WC and WC-Co particles were suspended in simple aqueous media without proteins, they agglomerated significantly — forming clusters with hydrodynamic diameters far larger than the individual primary particles. When the same particles were suspended in media containing serum proteins (as in a biological environment), agglomeration was suppressed, and particles remained more dispersed.

The critical finding, tested in rainbow trout gill cells (RTgill-W1), was that agglomerated particles were still taken up by cells. The agglomeration state of the particles in the exposure medium had no protective effect on particle internalization. Cells encountered and incorporated particles regardless of whether those particles were individually dispersed or clumped together.

The toxicological effects varied with media composition — but this variation appeared to be driven by how media components (especially proteins) interact with cobalt ions, rather than by the physical agglomeration of the particles per se.

This is an important finding for environmental risk assessment: it suggests that modeling nanoparticle behavior in natural waters purely based on agglomeration state may significantly underestimate their bioavailability and toxicological impact on aquatic organisms.

The most molecularly detailed chapter of this thesis used whole-genome microarrays to map changes in gene expression across 47,000+ gene probes in human skin cells (HaCaT) after exposure to WC, WC-Co, or dissolved CoCl₂ at two time points: 3 hours and 3 days. Five independent biological replicates were used per condition. The analysis employed statistical methods including SAM (Significance Analysis of Microarrays) and Gene Set Enrichment Analysis (GSEA).

The numbers tell a striking story of differential response:

WC alone caused virtually no transcriptional response — a relatively clean particle. WC-Co caused a substantial response, particularly at 3 days. CoCl₂ caused the largest response by far, and critically, 184 of the genes altered by WC-Co at 3 days were also altered by CoCl₂ — suggesting that the cobalt ions leaching from WC-Co particles are driving the majority of the transcriptional response.

Pathway analysis of the WC-Co gene expression signature revealed something unexpected and mechanistically important: the genes showing the greatest enrichment were those associated with hypoxic response — the cellular alarm system that activates when oxygen levels drop dangerously low.

Under normal oxygen conditions, the transcription factor HIF-1α (Hypoxia-Inducible Factor 1 alpha) is rapidly hydroxylated and then degraded by the von Hippel-Lindau (VHL) ubiquitin ligase pathway. When oxygen is scarce, this degradation is blocked, HIF-1α accumulates, and it switches on a programme of gene expression that reshapes cellular metabolism — promoting anaerobic glycolysis, angiogenesis, and cell survival strategies adapted to low-oxygen environments.

Cobalt ions are known to inhibit the hydroxylases responsible for marking HIF-1α for destruction — even in the presence of normal oxygen levels. The result is a "pseudohypoxic" state: the cell behaves as though it is starved of oxygen when it isn't. Key HIF-1α target genes identified in this study included:

Several of these genes (BNIP3, CA9, PFKFB4) are well-established HIF-1α targets. Their upregulation by cobalt ions — through a mechanism distinct from oxygen deprivation — is a known pharmacological effect (cobalt was historically used to treat anaemia for this reason). Seeing this signature in the gene expression pattern of nanoparticle-exposed cells provides direct mechanistic evidence linking WC-Co toxicity to the intracellular release of cobalt and subsequent activation of the HIF-1α pathway.

The research team was careful to situate their in vitro findings within the appropriate context — cell culture experiments cannot directly predict in vivo effects, and translating laboratory concentrations to real-world exposure scenarios requires significant additional work. Nevertheless, the implications of the findings are important to consider.

1. Occupational exposure is the primary near-term concern.

   Workers in hard metal manufacturing — handling WC and WC-Co powders during sintering, grinding, and machining operations — represent the highest-risk population. The finding that nanoparticles enter even non-phagocytic cells (such as lung epithelial cells) via multiple pathways suggests that inhalation of nanosized powders poses risks beyond those predicted by models based on micron-sized particle behavior.

2. The chemical composition of particles matters enormously.

   WC alone is relatively benign at tested concentrations; WC-Co is significantly more toxic. Regulatory frameworks that group nanoparticles by size alone, without accounting for chemical composition and potential ion leaching, will systematically underestimate the risk of composite nanomaterials.

3. Agglomeration should not be assumed to reduce risk.

   Environmental and occupational risk models that assume agglomerated particles are less bioavailable than primary particles may be incorrect. Cells took up agglomerated WC-Co and the particles remained toxic regardless of agglomeration state.

4. The HIF-1α pathway is a plausible mechanism for chronic effects.

   Sustained, cobalt-driven activation of hypoxia-response signaling could, over time, contribute to the tissue remodeling, fibrosis, and potentially carcinogenic changes seen in hard metal workers. The IARC classification of WC-Co as "probably carcinogenic" is consistent with this molecular picture.

5. Environmental monitoring for nano-sized particles needs to improve.

   If nanoparticles are released from production sites, consumer products (sunscreens, coatings, toothpastes), and waste streams, and if agglomeration does not eliminate their biological activity, then current environmental monitoring approaches — largely calibrated for larger particles — may be systematically missing biologically relevant exposures.

At the time of this research, the European chemicals regulation REACh did not include nanoparticle-specific provisions, and no standardised toxicological test battery existed for nanoparticles. The thesis explicitly calls for nanoparticle-specific safety assessment frameworks and labelling requirements for nanoparticle-containing products. While regulatory attention to nanomaterials has increased in the years since this research was published, the fundamental questions it raises — about long-term effects, in vivo translocation of particles, bioaccumulation, and the behaviour of composite nanomaterials — remain areas of active research and regulatory development.

This research is notable for several methodological strengths: multiple independent analytical techniques were used to confirm particle internalization (microscopy, EDX, flow cytometry), which reduces the risk of artifacts. The comparative experimental design — testing WC, WC-Co, and CoCl₂ side by side — allows clean attribution of effects. The use of multiple cell types (phagocytic and non-phagocytic, human and fish) broadens the applicability of findings. Genome-wide transcriptomics with five biological replicates, along with RT-PCR confirmation, represents a rigorous approach.

The honest caveats are equally worth noting. In vitro systems cannot replicate the complexity of in vivo exposure routes — the layered defenses of an intact epithelium, the role of the immune system, and the effects of particle transit through blood and lymph. The concentrations used in cell culture experiments may not reflect realistic exposure levels for workers or the environment. And critically, the enhanced toxicity of WC-Co over its components — the "synergistic" effect — could not be fully explained by the gene expression data; its exact molecular mechanism remains an open question.

The authors are admirably clear about these limitations and frame their work not as a definitive risk assessment but as a first step — identifying which materials warrant concern, characterizing mechanisms of action, and providing the foundation for hypothesis-driven in vivo studies.

Wibke Busch's doctoral research represents a careful, multi-technique characterization of how a set of industrially relevant nanoparticles interacts with mammalian and fish cells. Its core messages are:

Conclusion 1

Nanoparticles enter all kinds of cells, not just immune cells

The phagocytic capacity of a cell does not determine whether nanoparticles enter it. Industrial nanoparticles can penetrate lung cells, skin cells, gut cells, and fish gill cells via multiple uptake pathways.

Conclusion 2

Chemical composition drives toxicity; size is insufficient alone

WC is relatively benign; WC-Co is toxic at similar concentrations. Any risk assessment framework must account for the full chemical composition of composite nanomaterials and the possibility of ion leaching inside cells.

Conclusion 3

WC-Co mimics hypoxia at the molecular level.

Cobalt ions released from WC-Co nanoparticles stabilize HIF-1α and trigger a hypoxia-like transcriptional program. This provides a plausible mechanistic link between WC-Co exposure and the lung diseases seen in hard metal workers.

Conclusion 4

Agglomeration ≠ inactivity

Clumped particles are still taken up by cells and remain toxic. Risk models that discount agglomerated nanoparticles may underestimate real-world biological impact.

Nanotechnology is not going away — and nor should it. The benefits in medicine, clean energy, and materials science are genuine and significant. But those benefits must be developed in parallel with robust safety science. Research like this — systematic, mechanistic, transparent about its limitations — is exactly the foundation that responsible innovation requires.

Original Research Paper

Toxicity of Engineered Nanoparticles Towards Vertebrate Cells In Vitro

Wibke Busch · Martin-Luther-Universität Halle-Wittenberg · 2010