Nanotechnology industry is expanding at a rapid rate but in-depth exploration of the health and environmental effects of these materials is still warranted 1 . There is increasing evidence linking the NPs with human health problems. It has already been shown that inhaled carbonaceous NPs possess the potential to aggravate existing respiratory disorders, such as asthma or bronchitis 2 3 . Translocation of NPs from the lungs towards other organs has been demonstrated and possible consequences include inflammation, heart rate and function anomalies, homeostatic disturbances and oxidative stress 4 5 . More recently it has also been shown that pre-injected titanium dioxide nanoparticles can transform benign cells into aggressive metastatic tumor cells 6 . On the basis of current knowledge, there is increasing need for the risk assessment of both CB and TiO₂ due to increased environmental and occupational exposures. CB and TiO₂ are among the most abundantly produced and widely utilized NPs. Major sources of CB NPs include combustion (considered as combustion derived ultrafine particles) and industry. These particles also represent the core of atmospheric pollution particles. TiO₂ NPs are used in the preparation of sunscreens, cosmetics and tooth pastes 7 8 . Some recent estimates of annual global nano TiO₂ production range between 5000-6400 metric tones 9 10 . These enormous amounts of nanomaterial produced raise the possibilities of occupational and environmental exposures.

NP exposures can lead to disturbances in the cellular homeostatic mechanisms resulting either in adaptive cellular responses or cell death. NP-induced perturbations of cellular mechanisms might act as basis of different pathophysiological processes depending upon the concentration and duration of exposure 11 . Cell death could occur either by an abrupt process named necrosis or by a tightly regulated or programmed process (apoptosis and autophagy). Necrotic cell death occurs in different human pathologies like cerebral ischemia, myocardial infarction and acute organ failures. Apoptosis is a key event in many physiological, biochemical as well as pathological phenomenon. Either an excess or a reduced apoptotic process is involved in many pathological conditions such as autoimmune diseases, neurodegenerative disorders and carcinogenesis 12 . Apoptosis plays an important role in the pathogenesis of different respiratory disorders such as asthma, emphysema, and acute respiratory distress syndrome 13 14 .

Reactive oxygen species (ROS) play a dual role in the fate of cell i.e. causing cell death as well as acting as second messengers to induce an adaptive cell response 15 . Indeed, oxidative stress has been shown to induce cell death by a variety of mechanisms 16 17 18 . A hierarchical model for NP toxicity also describes the possibility of higher oxidative stress levels leading to cell death induction 11 . Different types of NPs have been shown to induce oxidative stress 19 20 21 but the role of oxidative stress in NP induced cell death has not yet been completely elaborated.

This work was carried out on human bronchial epithelial cells (16HBE14o- cell line and normal human bronchial epithelial cells) which represent one of the target cells at the portal of entry of NPs in human body. The present study was designed to investigate the molecular mechanism/pathways implicated in the cell death induced by two chemically distinct but nearly same sized NPs (CB and TiO₂). Furthermore, the involvement of oxidative stress in cell death induction by NPs was also investigated. To the best of our knowledge, this is the first report elaborating the molecular pathways involved in apoptosis induction by CB and TiO₂ NPs showing that distinct cellular mechanisms are involved in apoptosis induction by these two NPs.

NPs pose an emergent challenge to toxicologists as they are gaining diversified utilizations in many fields of life. A detailed description of molecular mechanisms induced by these NPs and their risk assessment is clearly warranted. This experimentation aimed at deciphering the mechanisms underlying the cytotoxic effects of CB and TiO₂ NPs. We have previously shown that these NPs are capable of inducing cytotoxicity in bronchial epithelial cells 21 . There have been a few studies which attributed apoptosis as the basic phenomenon for cytotoxic effects of different types of NPs but these observations were mostly made on the basis of caspase activation or annexin V staining for apoptosis detection and did not address the molecular pathways 7 22 23 . A recent article by Pan and colleagues describes that 1.4 nm gold NPs induce necrosis by oxidative stress and mitochondrial damage in Hela cells 24 . A deeper insight into the signaling pathways triggered by NP exposures in their main target cells will help in identifying the potential diseases/disorders which may occur after NP exposures and might also help in understanding the contribution of NPs in the pathogenesis of these disorders. In vitro studies are preferred for the comprehension of molecular mechanisms and signaling pathways and it has been shown that in vitro doses of 20 μg.cm^-2 represent exposures of high risk individuals (asthmatics etc) to ambient levels of atmospheric particles 25 26 .

Detailed physico-chemical characterization of both types of NPs was done before cell culture exposures. In cell culture medium, the surface charge on the NPs measured by the zeta potential is less negative than in water. This charge reduction is caused by the compensation of negative charges due to the presence of inorganic cations contained in the physiological media 27 . These cations are attracted to the negatively charged surface of the NPs. A less negative zeta potential favours particle aggregation, since the contribution of electrostatic repulsion is decreased 28 . We observed indeed that in physiological media, the hydrodynamic particle diameter is larger than in water as a result of this enhanced aggregation. Inside the human respiratory tract particles may also become aggregated as soon as they come in contact with the surfactant and mucous covering the pulmonary cells. The observation of cytotoxic effects from the aggregated NPs in the cell culture medium is not uncommon 29 . We have already shown that, even when in the form of aggregates, the cytotoxic, pro-inflammatory and oxidative effects of NPs depend upon the primary particle size/surface area 21 . We observed a few nanometer differences in the primary particle diameters we measured by TEM and data provided by the manufacturer (Table 1). This may be due to different methodologies used but these were not communicated by the manufacturer. Furthermore, we measured the diameters after suspending NPs in the cell culture medium, which might lead to coating of NPs. Nevertheless, this observation indicates the need to determine nanomaterial physicochemical characteristics under treatment conditions before experimentation.

Peripheral chromatin condensation, membrane blebbing and formation of apoptotic bodies is classical for apoptosis 30 . Recently similar ultrastructural features have been described in pulmonary epithelial cells of NP exposed factory workers 31 . TEM analysis confirmed the presence of these typical morphological features of apoptosis whereas necrosis was not observed at low doses. It is interesting to note that at least two to three times more doses of atmospheric particles are required to induce similar apoptotic effects (data not shown). Therefore we compared the cytotoxicity of the NPs used in this study with larger sized CB (21 nm and 95 nm) and TiO₂ (50 nm) and cytotoxicity increases with decreasing particle size (CB13 nm> CB 21 nm >CB 95 nm for CB and TiO₂ 15 nm>TiO₂ 50 nm TiO₂ NPs respectively) 21 . The morphological characterization of apoptotic cell death was confirmed with cytometric observations of cells with reduced size. The decrease in cell size (apoptotic volume decrease) is considered as a hallmark of apoptotic cell death and plays an important role in regulating the activity of apoptotic nucleases and activation of caspases 32 . Our observation of caspase activation by CB and TiO₂ NPs is in agreement with other studies showing recently the activation of caspase-3 in the apoptotic events of CB and TiO₂ NPs 22 33 . We have shown here that NP exposure leads to activation of both initiator (C-8) and executioner caspases (C-3/-7). Caspase-8 activation typically occurs in death receptor signaling pathway of apoptosis induction but recently it has been shown that alternative activation systems also exist 34 35 36 . Indeed, we did not observe the activation of Fas receptor after the treatment with NPs (data not shown). Furthermore, we have shown here that caspase 8 activation leads to the activation of caspase-3/-7 as its inhibition by a specific inhibitor (z-IETD-fmk) decreases caspase-3/-7 activation (Additional file 1 figure S2). Caspases have been shown to be sensitive to the redox status of the cells 16 37 38 and we have also shown a time dependent induction of oxidative stress by both types of NPs. Activation of executioner caspases (C-3/-7) leads to proteolysis and fragmentation of DNA which after fixation with permeablizing fixatives like methanol/ethanol leaks out of the cell and hypodiploid cells are detected as a subG1-peak in cell cycle analysis using PI incorporation. We have shown here that both CB and TiO₂ NPs exposure leads to fragmentation of DNA and appearance of this sub-G1 peak occurs after the activation of caspases confirming the role of caspases in the fragmentation of the DNA.

Interestingly, after treatment with NPs a different outcome was observed for CB and TiO₂ NPs in terms of loss of ΔΨ_m which is only induced by CB. This clearly points out that different mechanisms of cell death induction can be triggered by different types of NPs according to their chemical nature. Loss of ΔΨ_m occurs due to a variety of factors including mitochondrial outer membrane permeabilization (MOMP) and ROS. We have shown by the use of PEG- catalase that the CB NPs induced perturbation in mitochondrial potential, caspase activation and DNA fragmentation is dependent on the oxidative stress (H₂O₂ production) induced by these NPs whereas PEG-catalase could not protect against apoptosis induction by TiO₂ NPs. Catalase is an antioxidant enzyme which specifically scavenges H₂O₂ and PEG modification facilitates intracellular uptake of catalase, elevates intracellular activity and augments cellular resistance to oxidative stress as shown for endothelial cells 39 . Moreover, heat inactivation of the PEG-catalase leads to loss of its protective effect indicating that the protection is due to its enzymatic activity and not due to non-specific coating on the particles (Additional file 1 figures S3). MOMP is considered as a point of no return of the programmed cell death process conducting to the release of pro-apoptotic cytochrome c, apoptosome formation and caspase activation 40 41 42 . We observed activation of Bax and release of cytochrome c only in case of CB NP exposure. In the presence of all these findings, we are convinced of the involvement of the mitochondrial pathway and oxidative stress (H₂O₂) only in case of apoptotic cell death induced by CB NPs.

To understand the initial phase of apoptosis induction in the case of TiO₂ NPs we evaluated the contribution of the lysosomal compartment in the cell death pathways of these NPs. Lysosomal destabilization leads to release of lysosomal proteases particularly cathepsin B which participate in the final outcome of apoptotic phenomenon by either directly causing proteolysis or by activating other proteases like caspases 43 44 45 46 as leaked cathepsins in the vicinity of lysosomes are more likely to cleave procaspase-8 45 46 . Release of lysosomal proteases has been shown to be involved in cell death induction by a variety of stimuli (ROS, drugs, TNF alpha) 44 47 48 49 including silica NPs 50 . We observed a significant lysosomal destabilization only in case of TiO₂ NPs. We have previously reported that NPs are internalized as aggregates which are essentially taken up in the phagosomes/phagolysosomes 21 . We analyzed the internalization of these NPs by flow cytometry and TEM and observed that these are internalized in a dose dependent manner in 16HBE cells. This fact points towards the possibility that TiO₂ NPs may induce lysosomal damage during the endocytic uptake process leading to release of cathepsins. ROS are one of the variety of factors which cause lysosomal destabilization 51 52 . In addition to oxidative stress, we also observed a dose dependent lipid peroxidation by TiO₂ NPs. These results are in agreement with previous studies which described oxidative DNA damage and lipid peroxidation in BEAS-2B cells without photo-activation of TiO₂ NPs 19 . It has been postulated that ROS production (specifically hydroxyl radical production) can occur at the surface of TiO₂ NPs even in the absence of UV illumination or sun light exposure 53 and it is a proven fact that hydroxyl radicals are principal agents which lead to lipid peroxidation. This may explain the different apoptotic pathways observed for TiO₂ and CB NPs which did not induce lysosomal destabilisation. It is noteworthy that lipid peroxidation has also been shown to activate caspases in a cellular redox status depending manner 54 . Taken together, these data strongly suggest the contribution of lysosomes and lipid peroxidation in the cell death induction by TiO₂ NPs. All these mechanisms were further confirmed in normal human bronchial epithelial cells (NHBE).

Thus we conclude that the CB and TiO₂ NPs induce a rapid apoptosis in bronchial epithelial cells and initial phase of cell death induction is different in both cases. In Figure 8, we have represented a schematic diagram of pathways involved in the apoptosis induction by both CB and TiO₂ NPs. In case of CB a decrease in ΔΨ_m and oxidative stress could already be observed after 30 minutes of exposure while caspase-3 activation is significantly increased from 2 hours and subG1 peak only after four hours. Apoptotic events depend on H₂O₂ formation in case of CB NPs as they could be effectively modulated by using PEG-catalase. This implies that initial production of ROS leads to loss of ΔΨ_m causing subsequent release of cytochrome c and activation of C-3, C-7 which leads to DNA fragmentation. In case of TiO₂ NPs, initial ROS production, lipid peroxidation and lysosomal membrane destabilization (all significantly induced from 30 minutes of exposure) lead to release of cathepsin B and activation of caspase-8 after 1 hour and caspase-3/7 after 2 hours resulting in DNA fragmentation which is only observed after 4 hours.

In conclusion we have elucidated different molecular pathways that mediate the apoptotic effects of NPs in bronchial epithelial cells (16HBE14o- and NHBE cells). Indeed, the initial phase of cell death induction varies according to the chemical nature of NPs. In case of CB NPs the production of ROS is implicated in the downstream mitochondrial events whereas, in case of TiO₂ NPs the initial lysosomal destabilization and lipid peroxidation is involved in the apoptosis induction. These results will help in understanding the possible contribution of NPs in the pathogenesis of those disorders in which apoptosis plays an important role and in a longer run will assist in the development of strategies to counteract the possible adverse health effects of NPs. Our results clearly point towards the need of meticulous understanding of molecular events occurring after NP exposures and not just looking at the final outcome as the pathways may differ for diverse NPs. Indeed, the presence of distinct pathways raises the possibilities of observing different effects of the NPs at non cytotoxic concentrations which may induce other pathophysiological phenomenon.

All the chemicals were purchased from Invitrogen (Cergy-Pontoise, France) unless otherwise specified.

The authors declare that they have no competing interests.

SB and SH contributed in study design, did the experimental work, analyzed data and wrote the manuscript. LT and JM did the particle characterization. CB and JF helped in TEM analysis. IF helped in experimental work and reviewed the manuscript. FM, AB, KA critically reviewed the manuscript and gave intellectual input. All the authors have read and approved final manuscript.