Palladium Nanoparticles: Toxicological Effects and Potential Implications for Occupational Risk Assessment

The increasing technological applications of palladium nanoparticles (Pd-NPs) and their consequent enhancing release into the community and occupational environments, have raised public health concerns regarding possible adverse effects for exposed subjects, and particularly for workers chronically and highly exposed to these materials, whose toxico-kinetic and dynamic behavior remains to be fully understood. Therefore, this review aimed to critically analyze literature data to achieve a more comprehensive knowledge on the toxicological profile of Pd-NPs. Results from available studies demonstrated the potential for these chemicals to affect the ecosystem function, to exert cytotoxic and pro-inflammatory effects in vitro as well as to induce early alterations in different target organs in in vivo models. However, our revision pointed out the need for future studies aimed to clarify the role of the NP physico-chemical properties in determining their toxicological behavior, as well as the importance to carry out investigations focused on environmental and biological monitoring to verify and validate experimental biomarkers of exposure and early effect in real exposure contexts. Overall, this may be helpful to support the definition of suitable strategies for the assessment, communication and management of Pd-NP occupational risks to protect the health and safety of workers.


Introduction
Modern science has experienced one of its major breakthroughs with nanotechnology in the last decades. This enabled nano-dimensional materials (in the 1-100 nm size domain) to be produced and applied in a number of technological and consumer fields due to their peculiar physico-chemical properties [1].
Metallic based nanomaterials have attracted great scientific attention in the nanotechnology field. Particularly, given that noble metals are characterized by remarkable catalytic, electronic, magnetic, optical, and mechanical properties per se [2][3][4], the study of the characteristics and possible uses of nano-sized noble metal materials has emerged as hugely important and valuable for the benefits potentially offered in a number of different applications.
In this regard, palladium (Pd) is a rare and precious metal that belongs to the Platinum group elements. It is largely employed as an active catalyst material in automotive catalytic converters, but finds also application in the electronic, engineering, biomedical, and jewelry sectors [5][6][7][8]. Palladium-NPs offer the opportunity for being even more effective catalyst materials due to their high surface area to volume ratio and high surface energy [9].  Battke et al. [28] Pd-NPs entrapped in an aluminum hydroxide matrix Lettuce seeds Seeds were planted immediately (day 0) or 15 day after adding NPs to the soil (0.013 and 0.066% w/w)

Seed germination and growth
Shoot/root ratio: no significant influence. on the difference in the ratio when the seeds were planted on day 0. palladium-NPs at low concentration significantly increased the ratio (2.5 cm) compared to controls (1.41 cm) in soils incubated with NPs for 15 d.
Shah and Bazelerova, [26] Size: 5-10 nm Kiwifruit pollen from plants of the male genotype of Actinidia deliciosa var. deliciosa Pollen was exposed to 0-7 mg/L NPs for up to 90 min Pollen performance and lethality Tube emergence and growth: significant inhibition began at 0.1 mg/L, complete growth cessation at 0.4 mg/L. Lethality: as concentration increased, viability exponentially decreased (LC50 at 90 min: 1.0 ± 0.3 mg/L).
Speranza et al. [27] Halloysite supported Pd-NPs. Radish seeds were exposed to 10 mL of nanomaterials at concentrations ranging 0-1500 mg/L for up to 72 h Germination and seedling development Exposure to Halloysite -PdNPs had no significant influence on germination, seedling development, xylem differentiation, or mitotic index in both lots.
Bellani et al. [39] Animal cells RTgill-W1 Cells were exposed to 5-25 mg/L NPs for 1 h and 3 days Cell viability Metabolic activity and membrane integrity showed a significant decrease after 1 h exposure due to cellular stress. Cell viability was full restored after 3 days.
Hildebrand et al. [40] Size: 10 ± 6 nm Rat-1 Cells were exposed to 1 and 2 µg/mL NPs for 2-120 h Cell viability; Oxidative stress reaction; cell cycle distribution; DNA damage Cell viability was significantly reduced by both concentrations at 120 h. Cell cycle distribution: time-dependent increase in G0/G1 phase (from 45% in controls to 70% following 2 µg/mL for 120 h). Decrease of cell percentage in the S phase (from 30% in controls to 15% following 2 µg/mL for 120 h). DNA damage: a significant effect was evident only following treatment with 2 µg/mL for 120 h. ROS production: significant increase compared to controls only after 2 µg/mL NPs for 2 and 4 h.

Plant Models
The emerging application of Pd-NPs in many fields inevitably led these chemicals to contaminate environmental matrices, potentially resulting in bioaccumulation in living organisms. This raised interest in assessing the effects of such metal-NPs on plant models. In this regard, Battke et al. [28] provided the first report regarding the plant uptake of Pd-NPs and their phyto-toxic effects. As a sign of abiotic stress, leaf length of barley plants decreased or became extremely variable, and leaves developed as rigid and slightly convoluted in response to Pd-NP treatment.
Palladium-NPs were also reported to significantly affect the growth of lettuce seeds, as measured by shoot to root ratios, when these were planted in soil samples previously incubated with NPs for 15 days [26]. Conversely, Pd-NPs supported on halloysite nanotubes showed no influence on radish germination and development of high and low vigor seeds of Raphanus sativus, mitotic index and chromosomal figures, therefore supporting the biological safety of such nanomaterials [39]. Speranza et al. [27] assessed the effects of Pd-NPs on the performance of kiwifruit pollen cultures, as an experimental model largely employed to assess the chemical impact on plant cell metabolism. Palladium-NPs were able to induce morphological alterations of pollen grains, damage to cell membranes, and depletion of endogenous calcium, therefore resulting in impaired pollen tube emergence and growth. When NP induced effects were compared to those exerted by PdCl 2 on the same endpoints, higher cyto-toxicity was reported for the nano-sized form, maybe in relation to a greater metal adsorption derived from NP exposure.

Animal Cells
Cell viability tests with RTgill-W1 (Oncorhynchus mykiss) fish cell lines showed no obvious acute toxicity for Pd/magnetite NPs. Metabolic activity and membrane integrity, as cell viability parameters, showed a slight decreasing trend after 1 h of particle exposure. However, this was attributed to a cellular stress response induced by NP addition, more than to a specific cytotoxic effect of NPs. After three days of exposure, cell viability was fully recovered. The third endpoint under study, lysosomal integrity, did not show any alteration at both periods investigated [40].
Conversely, in the normal diploid Rat-1 fibroblast cell line acutely and sub-acutely treated with Pd-NPs, a significant dose and time-dependent cell growth inhibition was observed [30]. A cell cycle arrest with accumulation of cells in G0/G1 phases and a significant DNA damage was evident, although only a slight increase in the intracellular reactive oxygen species (ROS) levels could be detected.

Cytotoxic Effects
The exposure to Pd-NPs was reported to induce different toxic effects on various cellular models. The assessment of cell viability in human colon adenocarcinoma Caco-2 cells and HaCaT keratinocytes, treated with Pd-NPs, revealed only minor effects, maybe related to the cellular stress caused by the NP application [40]. On the other hand, Wilkinson et al. [29] showed that the exposure of primary bronchial epithelial cells (PBECs), and lung carcinoma epithelial cells (A549) used to simulate the upper and lower respiratory tract, respectively, to Pd-NPs resulted in a concentration-dependent cytotoxicity. Of note, PBECs were markedly more affected by Pd-NPs than A549 cells. In line with these findings, apoptosis was induced in a dose-dependent manner by Pd-NPs in PBECs, but not in A549 cells [29].
Nanoparticle chemical composition appeared also relevant to determine cellular viability effects, since individual Pd-NPs resulted in less cytotoxicity compared to bimetallic Pt-Pd-NPs on human epithelial cervical cancer HeLa cells [32]. Such enhanced effect could be attributed to a synergistic action of both components. Also, the type of NP surface coating was reported as a possible influencing factor for the Pd-NP toxicological profile. Positively charged cysteamine functionalized Pd nanosheets, in fact, showed a much higher cytotoxicity on human liver HepG2 cancer cells, compared to negatively charged 3-mercaptopropionic acid functionalized NPs [35]. This was supposed to be related to the stronger contact that two dimensional, cysteamine functionalized nanosheets may have with cell surface compared to the mercaptopropionic-acid nanosheets, which aggregated in three dimensional particles, therefore changing their contact level with the cellular membrane. However, a dose-dependent cytotoxic effect on HepG2 cells was also reported by Rajakumar et al. [33] with Pd-NPs synthesized using the aqueous leaf extracts of Eclipsa prostrata. Increasing concentrations were able to induce cytopathic effects leading to cell damage and necrosis.
When Pd-NP acute and sub-acute cytotoxic effects were investigated on peripheral blood mononuclear cells (PBMCs) leaving quiescence through a phytohemagglutinin-L stimulation, a timeand dose-dependent reduction in cell viability could also be detected [31]. Interestingly, non-cytotoxic Pd-NP doses caused a significant increase of cells within the G0/G1-phase, as demonstrated by an increased percentage of mitogen activated PBMCs with a diploid DNA content, and a significant reduction of cells in GS-and G2/M-phases. In line with these results [31], a time and dose-dependent inhibition of cell growth has been recently demonstrated by Iavicoli et al. [30] in exponentially growing cultures of A549 cells treated with Pd-NPs for sub-acute periods of time (48-120 h). However, such cell viability inhibition was less evident compared to that determined in the same study on Rat-1 fibroblasts, suggesting a species-specific susceptibility to the NP toxicity. Moreover, Pd-NP-induced inhibition of cell growth was not associated with the induction of apoptosis. As reported by Petrarca et al. [31], Pd-NPs induced a progressive cell cycle arrest with accumulation of cells in the G0/G1 phase of the cell cycle, which could suggest a potential toxic effect of Pd-NPs on DNA. According to this statement, the authors found that Pd-NPs were able to induce a significant increase in DNA breaks already after 4 h of exposure and such damages become even more evident after longer periods of treatment even if only a slight increase in the intracellular ROS levels could be detected. Alarifi et al. [36] demonstrated that Pd-NPs not only inhibited human skin malignant melanoma A375 cell proliferation in a dose-and time-dependent manner, but also induced apoptosis, DNA damage and a cell cycle arrest. Inhibition of cell growth was specifically related with a reduction of the percentage of cells in the G0/G1 phase and an accumulation of those in the S and G2/M phases of the cell cycle.
A recent study investigated the possible effects that Pd-NPs exerted on human eosinophil cells known to have a key role in lung diseases and allergies [43]. In contrast to previous findings, Pd-NPs did not induce necrotic or apoptotic cellular viability alterations in both human 3D10 eosinophils and primary eosinophil cells in vitro. However, these NPs were able to induce actin cytoskeleton rearrangement in exposed cells, therefore determining their increased adhesion to endothelial cells. This was confirmed by the pre-treatment of cells with cytochalasin D, a potent inhibitor of actin polymerization, which in turn determined levels of adhesion comparable to those found in untreated controls.
Concerning possible modes of action, oxidative stress is considered as one of the most important mechanisms for cytotoxic effects of NPs [44]. However, regarding Pd-NPs, an oxidative stress induction was detected only in A375 human skin malignant melanoma cells [36] and in A2780 human ovarian cancer cells [4] based on ROS production following an acute or sub-acute treatment, which could be responsible for the induction of DNA damage and apoptotic-mediated cell death. Gurunathan et al. [4] also hypothesized that excessive production of ROS could trigger an autophagic phenomenon in ovarian cancer cells, eventually leading to cytotoxicity. Interestingly, the surface-specific ROS production of Pd-NPs was reported as being significantly related to the primary particle size, with a maximum around 12 nm, for both acellular and human leukaemia cellular environments [41]. On the other side, no significant ROS production was observed for Pd-NPs in Caco-2 and Hacat [40], A549 cells [29,30], PBECs [29], PBMCs [31], and human eosinophils [43].
Therefore, other possible molecular mechanisms should be elucidated for their role in Pd-NP cytotoxicity. Dahal et al. [34] confirming these results, demonstrated that Pd-nanocubes induced apoptosis in transformed human epithelial HeLa cells only in serum-free medium. The effect of serum on compromising the efficacy of Pd-nanocubes in inducing apoptosis can be related to the healing properties of proteins, antibodies and antigens present in serum on cytotoxic effects, as well as on the possibility that these components may form different bio-molecular coronas around nanocubes changing their biological reactivity. Apoptosis was the primary route of Pd-nanocube-induced cell death, although it was associated with a hyperpolarization of mitochondria, contrary to common depolarization initiated by ROS. This may suggest that the mechanism of Pd-nanocube toxicity was different from cytotoxic effects mediated by oxidative stress reactions.
Regarding the role of solubility in NP-induced effects, Petrarca et al. [31] suggested that Pd ions, per se or released from NPs could be the inducers of Pd toxicity as assessed comparing effects induced by Pd(IV) ions and Pd-NPs on PBMCs. Pd ions resulted more toxic, compared to Pd-NPs, causing almost complete loss of cell viability earlier, and at a lower dose, and a significant ROS increase. Following Pd (IV) ion exposure, more marked subcellular alterations, i.e., the presence of numerous auto-phagosomal vacuoles containing damaged mitochondria, and/or undigested cytoplasmic material, and a significant amplification of cell cycle alterations described for Pd-NPs, could be demonstrated. This may suggest that Pd ions released from Pd-NPs may have a role in their toxicological behavior. Metal ions may accumulate in mitochondria damaging their function which may in turn arrest the cell cycle and cause cell death. However, deeper research is necessary to verify and confirm such possible mechanism of action.

Immunological and Inflammatory Responses
The immune potential of Pd-NPs was evaluated in a couple of studies through the analysis of cytokine release and expression in PBMCs of non-atopic [37] and Pd-sensitized women [38]. In the first study [37], a significant reduction in IL-17 and TNF-α concentrations was induced by NP exposure, with or without lipopolysaccharide (LPS) stimulation, while the INF-γ release was significantly enhanced by the highest concentration of NPs in LPS stimulated cells. These results were partially comparable to those obtained with a Pd salt in the same conditions of exposure, with the exception of the Pd-NP stimulatory action on the INF-γ release. Overall, this may suggest that Pd-NPs may exert an important immune modulatory effect in vitro which was also confirmed when such cytokine changes were investigated in PBMCs obtained from Pd-sensitized non-atopic volunteers [38]. The addition of 10 −5 M Pd-NPs did not significantly modify the release of cytokines in LPS un-stimulated cultures, but induced a significant inhibition of TNF-α release in LPS stimulated PBMCs. These findings suggested a possible immuno-modulatory effect of Pd-NPs on cytokine release. However, deeper investigation is necessary to define underlying modes of action in PBMCs from Pd-sensitized or not-sensitized subjects and possible long-term health implications of such alterations.
Interleukin-8 secretion was also investigated by Wilkinson et al. [29] as a possible biomarker of the inflammogenic potential of Pd-NPs on PBECs and A459 cells. Additionally, the levels of prostaglandin E 2 (PGE 2 ), potentially involved in a number of biological functions, including vasodilatation and both anti-and pro-inflammatory action, were also evaluated. A non-linear trend could be detected for the secretion of IL-8 in response to Pd-NP treatment by A549 lung epithelial cells. A dose-dependent decrease was evident in the lower concentration range (0.01-1 µg/mL), while a significant increase compared to controls was achieved at the highest concentration (10 µg/mL). On the other hand, a dose-dependent decrease could be detected for the PGE 2 secretion in both cell types. Notably, these effects are seen at a non-cytotoxic concentration of Pd-NPs, indicating that these biomarkers serve as sensitive indicators of the biological effects of Pd-NPs. Recently, a significant, dose-dependent increase in IL-8 secretion was also reported in a three-dimensional cellular model obtained co-culturing PBECs and human fetal lung fibroblasts MRC-5 at an air liquid interface, able to mimic an in vivo healthy and chronic bronchitis-like mucosa acutely exposed to Pd-NPs [42]. This demonstrated that Pd-NPs could induce an inflammatory response in both mucosal models, although IL-8 concentrations were significantly higher in the inflamed model compared to the normal one.

Ex Vivo Studies
Skin contact with Pd can result in sensitization and allergic contact dermatitis [45][46][47]. The Pd-NP higher surface area to mass ratio may be responsible for greater biological activities compared to the bulk form of the metal, as well as an easier release of reactive metal ions [48,49], with a greater potential for skin permeation. To verify this latter aspect, Larese et al. [50] investigated the penetration of 0.60 mg/cm 2 of 10.7 ± 2.8 nm sized Pd-NPs applied on intact and needle-damaged human skin pieces obtained as a surgical waste from four different donors in an in vitro diffusion cell system. Human skin absorption of Pd-NPs occurred in both intact and damaged skin, although lesions increased the metal absorption significantly. The Pd content in intact skin decreased significantly from the epidermis to the dermis. Inside the skin, these NPs can be a long-term source of metal that could be involved in the sensitization process or can be available for systemic diffusion.

In Vivo Studies
Several biological models, from bacterial strains [51] to mammalian organisms were reported to be affected by Pd-NP treatment. The immune [52,53], renal [54] and endocrine systems [55] were demonstrated as critical targets for the toxic action of Pd-NPs ( Table 2). The elemental Pd content in urine, and changes in cytokine serum levels, urinary protein content and hormonal serum concentrations, emerged as possible biomarkers of Pd-NP exposure and early effect, although further studies are necessary to confirm these preliminary results and validate their possible employment in real exposure settings. Table 2. Studies investigating the effects of Pd-NPs on in vivo models.

Bacterial models
Pd-NPs entrapped in an aluminum hydroxide matrix Microcosm soil NPs were added to the soil at a final concentration of 0.013% or 0.066% (w/w) Bacterial growth No significant effects on the number of colony forming units or on the total soil community metabolic fingerprint. Shah and Belozerova, [27] Bimetallic Pd-iron NPs Sphingomonas sp. PH-07 Bacterial strain was exposed to 0-0.5 g/L Bacterial growth No differences compared to controls were evident up to 0.1 g/L, whereas bacterial growth was significantly inhibited with greater NP concentrations.
Gram negative E. coli, Gram positive S. aureus bacterial strains Bacteria were exposed to 2. V. fischeri were exposed to 0-9 µg/mL for 5-30 min. Marine cultures were exposed to 5 and 50 mg/kgdw for upt to 18 weeks Bacterial growth of V. fischeri; Respiratory metabolisms and structure of the marine microbial community Acute toxicity: no significant effects. Respiratory metabolisms: Pd-NPs did not impact the dechlorination activities; while dose-dependently inhibited the sulfate reduction and methanogenesis. Pd-NPs increased the richness of the microbial community.
Iavicoli et al. [52] Female Wistar rats (n. 5 per group) Animals were exposed to repeated (on day 1, 30 and 60) intravenous injections of Pd-NPs at 0-12 µg/kg dose. Serum cytokine concentrations were assessed at the 90th day post-exposure Effects on the immune system The highest dose of Pd-NPs (12 µg/kg) induced a significant reduction of IL-1α, IL-4, IL-6, IL-10, IL-12, and GM-CSF compared to controls. The 1.2 µg/kg dose induced also a significant reduction of IL-1α. Interleukin-10 was significantly reduced at all the investigated concentrations.

Bacteria and Microbial Communities
To understand the interactions between nanomaterials and micro-organisms can be helpful to define NP toxicological profiles on simple biological entities and extrapolate information to be verified on more advanced organisms within more complex environmental conditions of exposure. In this regard, an anti-microbial effect of Pd-NPs was reported by Adams et al. [51] toward Gram negative Escherichia coli and Gram-positive Staphylococcus aureus bacterial cultures, although with a greater inhibition in this latter strain, maybe in relation to the stronger resistance of the cell wall of negative bacteria compared to positive ones. Interestingly, the anti-bacterial activity resulted dependent on the NP size, and time-periods investigated. Antibacterial properties of Pd-NPs were also subsequently confirmed on E. coli and S. aureus cultures [58] and on the novel multidrug resistant Cronobacter sakazakii strain in which NPs were able to eradicate also biofilm bacterial growth, responsible for their multidrug resistance [60]. A comparable antimicrobial impact was also evident with bimetallic-Pd-iron-NPs, which demonstrated to significantly inhibit the growth rate of the dyphenyl ether degrading bacterial strain Sphingomonas sp. PH-07 in a dose-dependent manner, at concentrations greater than 0.1 g/L [56].
In order to address whether the known antibacterial properties of the conducting polymer polyaniline [61], could be enhanced via the incorporation with metal NPs, Pt-Pd colloidal solutions as well as Pt-Pd-polyaniline nanocomposites were explored for their action against selective Gram positive (Streptococcus sp. and Staphylococcus sp.) and Gram-negative bacteria (E. coli and Klebsiella sp.) [57]. Nanocomposite materials exhibited improved antibacterial activity compared to pristine polyaniline. Interestingly, Pt-Pd individual nano-metal colloids did not exhibit any antibacterial activity, suggesting that the enhanced biological impact could be dependent on the interaction between metal NPs and a pristine polyaniline matrix as well as on the greater reactivity of smaller NPs in the nano-composites. The possibility to employ Pd-NPs in the remediation of groundwater, wastewater and sediments, raised concerns among possible effects these NPs may exert on environmental microbial communities once released into the environment [59]. Therefore, the impact of Pd-NPs on the marine bacteria was assessed on a single eco-toxicological model of marine bacterium, V. fischeri, as well as on active bio-remediating microbial communities incubated under laboratory biogeochemical conditions mimicking those occurring in situ. Pd-NPs showed nor toxic effects on V. fischeri, nor alterations in the total bacterial community structure, as its biodiversity was increased compared to the not exposed community. Additionally, concerning respiratory metabolisms, the negligible effect on organohalide respiration activities, as well as the dose-dependent inhibition of sulfate reduction and methanogenesis, further support the lack of ecotoxic effects of Pd-NPs on marine microbiome. Comparably, no significant effects on the number of colony forming units or on the total soil community metabolic fingerprint could be detected in soil microbial communities exposed to Pd-NPs in experimental microcosm conditions of exposure serving as a direct analog of ecological system [26]. No differences could be also detected between the variable NP concentrations employed, maybe in relation to the immobilization of NPs on soil organic compounds, such as humic acids.
Additionally, the antiplasmodial activity of Pd-NPs was demonstrated by Rajakumar et al. [33] in mice intraperitoneally inoculated with Plasmodium berghei-infected red blood cells. The untreated controls showed a progressively increasing parasitemia, while animals treated via the oral route with Pd-NPs showed a 78.13% decrease in parasitemia with a reported inhibitory concentration 50 value of 16.44 mg/kg/body weight.

Animal Models
Concerning in vivo results, investigations conducted on female Wistar rats, proved the ability of Pd-NPs to significantly affect the immune system of treated animals inducing alterations in the serum levels of several cytokines secreted by different T helper (Th) lymphocyte subsets [52,53]. A subacute (14 days) exposure to increasing doses of Pd-NPs was able to induce a general, although not significant, upward trend, in all cytokine serum concentrations investigated, while a significant increase was only determined by the highest dose of Pd-NPs (12 µg/kg) [52]. These findings may suggest that Pd-NPs could exert a significant and generalized stimulatory impact on the immune system of exposed rats. However, when investigated in a sub-chronic time frame, such cytokine increasing trend was not more evident [53]. Repeated injections of Pd-NPs, in fact, induced a decreasing trend in serum levels in most of the cytokines investigated, with the highest concentration (12 µg/kg) determining significant inhibitory effects compared to controls, supporting a more complex interaction between these nanomaterials and the immune system, related also to the time points investigated.
As another possible target organ of the Pd-NP toxic action, the effects on the renal system was recently assessed in Wistar rats acutely exposed via the intravenous route to a wide range of NP concentrations. Results showed an increasing trend in urinary Pd levels demonstrating that Pd urinary concentrations were directly associated with the administered doses. Additionally, urinary concentration of low, retinol binding protein and β2-microglobulin, and high molecular weight proteins were employed to assess the nephrotoxic action of Pd-NPs [54]. Interestingly, the highest administered dose (12 µg/kg) was able to induce a significant increase in urinary retinol binding protein and β2-microglobulin levels compared to controls. This may suggest that Pd-NPs may exert a nephrotoxic action at the renal tubular level. In fact, these low molecular weight proteins are excreted in urine in small quantities when tubular function is normal, while an increase in their urinary excretion is indicative of renal tubular damage. As a confirmation of this specific damage, ultra-structural analysis of kidney specimens revealed subcellular dose-dependent morphological alterations with the main target in the proximal and distal tubular epithelium. However, these preliminary results need to be confirmed also to define possible Pd-NP molecular mechanisms of action.
More recently, a first attempt to investigate the endocrine disruptive potential of Pd-NPs was performed by our group of research in an animal model of sexually mature female Wistar rats sub-acutely exposed to NPs via a single tail vein injection [55]. An abnormal function of the hypothalamic-pituitary-gonadal axis was detected as demonstrated by the reduced serum concentrations of estradiol (E2) and testosterone (T) and the increased level of luteinizing hormone (LH) in treated animals compared to controls, which become significant at the highest exposure dose (12 µg/kg). Overall, these results demonstrated the ability of Pd-NPs to significantly affect the normal sex hormone levels, thus suggesting that they may play an important role in disrupting the physiological functions of the female reproductive system of Wistar rats.

Discussion
Over these last years, the enhancing interest in the synthesis and application of Pd-NPs, together with their consequent increasing release into the community and occupational environments, have raised concerns regarding possible adverse effects exerted on ecosystems and human health. Particular attention has been focused on the health of workers who may be repeatedly and long-term exposed to high doses of these nanomaterials, whose toxico-kinetic and dynamic behavior is still not fully understood. In this emerging scenario, this is the first attempt to comprehensively analyze the state of the art concerning Pd-NP toxicological behavior on simple and more complex biological systems in order to extrapolate information useful to guide suitable risk assessment processes and point out future research needs.
Although no definite conclusions can be drawn concerning the eco-toxicological impact of Pd-NPs, the adverse effects reported on plant/seed models in vitro [27,28,39] do not allow to exclude a potentially serious risk for plant reproductive success and growth, as well as for the ecosystem functions.
Regarding Pd-NP antimicrobial properties, a significant inhibitory effect on bacterial growth was evident in pure microbial cultures [51,60] while it was not evident in more complex soil and marine microbial communities under experimental conditions mimicking those of the natural environment [26,59]. This may be dependent on the possible interactions that NPs may have with environmental compounds that may, in turn, prevent strong influences on microbial communities. Overall, these data point out the importance of combining standard eco-toxicity tests, microbial metabolic assessments, and analyses of microbial community composition, in a multifaced approach to evaluate the impact of NPs on the environmental microbiota and underlined molecular mechanisms of action [59].
The majority of the reviewed in vitro studies reported significant cytotoxic [4,[29][30][31]34,36] and immuno-modulatory effects [29,37,38] induced by Pd-NP exposure on diverse animal and human cells. Interestingly, different responses to NP treatment were determined in cell cultures used as models for the upper and lower part of the respiratory tract, maybe in relation to the differences in particle uptake, and culture media employed, which may change the NP bio-molecular corona, therefore affecting NP toxicological profiles [29]. In this scenario, more complex tri-dimensional in vitro models, resembling healthy and chronic bronchitis-like mucosa, may be useful to achieve data to be easily extrapolated to an in vivo context and useful to define areas of the respiratory tract, as well as pathological conditions, characterized by a greater susceptibility to the NP toxicity [42].
Interestingly, in vitro studies are generally useful to define possible NP molecular mechanisms of action. However, current evidence on Pd-NPs do not allow to define specific conclusions in this regard, as no uniform results were reported concerning the role of such chemicals as triggers of oxidative stress reactions, apoptotic pathways, DNA damage cascades as well as cell cycle disturbances. Additionally, it cannot be excluded that the type, i.e., organic ligands, surfactants, polymers and dendrimers, and variable concentrations of chemical stabilizers employed in NP preparation, may influence their toxicological profile. Therefore, further investigation seems necessary to define the role that the complex interplay between the different NP physico-chemical properties, cellular growth media and peculiar cellular-type features may have in determining different patterns of cytotoxicity.
In this regard, Pd-NP structural diversity and plasticity on supporting matrix [34,39] size, also in a very narrow distribution [51], chemical composition as well as surface functionalization [32,35], may all represent key NP properties potentially affecting their toxicological behavior. This may finally suggest the possibility to modify nanomaterial bioactivity changing NP physico-chemical characterization, therefore supporting the "safe by design" concept. This argues for nanomaterials that, while maintaining most of their innovative and revolutionary physico-chemical properties, are at the same time characterized by absent or lower toxicity. From an occupational health perspective, this may lead to a safer production and application of such materials, possibly preventing the development of adverse health effects.
Experiments carried out on animal models are extremely important to primarily define the toxico-kinetic and dynamic-behavior of NPs, as well as to identify possible biological indicators of exposure and early effect. Several organ systems have been reported to be affected by NP exposure, from the immune to the endocrine, as well as the renal systems [52][53][54][55]62]. However, it remains to be elucidated whether early alterations detected in sub-acute, sub-chronic conditions of exposure, i.e., cytokine changes, alteration in hormonal serum concentrations, as well as increase in urinary protein content, may persist in the long term, and their possible role in disease development and manifestation. Future research, in this perspective, should be focused on the health impact of low-dose, long term conditions of exposure, as those potentially experienced in real workplace settings.
Interestingly, concerning the immuno-modulatory effects of Pd-NPs on cytokine production and/or release, preliminary investigation on this topic in vivo [52,53] failed to show a significant preferential development of a specific cytokine secretion profile that may be responsible for an imbalance toward a cellular or humoral immune response. This may reflect the complexity of the Pd-NP interaction with biological systems in vivo which may be affected by modes of exposure, penetration of physiological barriers, solubility in biological media as well as the bio-molecular corona formation, which may all change the impact of the Pd-NP insult.
From an occupational health perspective, the possibility to extrapolate from animal investigations possible biomarkers of exposure and effects seems absolutely important. Considering the still existing practical difficulties in environmental monitoring of workplace NP exposure, due to the lack of standardized procedures and dosimetric parameters to be measured, biological monitoring may provide complementary information concerning exposure occurring by all routes, considering the inter-individual variabilities in the toxicokinetic of chemicals [63]. In this regard, the Pd elemental content in biological matrices should be investigated and validated as a possible indicator of internal dose in real occupational exposure settings. On the other side, alterations in nephrotoxic, immunological and hormonal parameters should be verified and confirmed as possible biomarkers of early effects in in field studies. The suitable definition of Pd-NP target organs, may be important to understand possible conditions of susceptibility due to pre-existing alterations/pathologies in such organs, which may deserve particular attention from an occupational health point of view [64].

Conclusions
This review represents the first attempt to achieve a more comprehensive knowledge of the toxicological profile of Pd-NPs. Although the potential for these nanomaterials to affect the ecosystem function, to exert cytotoxic and pro-inflammatory effects in vitro as well as to induce early alterations in different target organs in in vivo models has been reported, further investigations appear absolutely necessary to confirm such preliminary findings.
These should take advantage of a more comprehensive characterization of the physico-chemical properties of Pd-NPs in order to get a deeper understanding of the complex interplay between their intrinsic features and the detected health effects. Additionally, field studies strongly encourage defining the environmental levels of exposure to Pd-NPs in real workplace settings. Environmental assessment should be accompanied by biological monitoring investigations focused on verifying and validating experimental biomarkers of exposure and early effects in a real exposure contexts. Overall, this seems important to support the definition of adequate strategies for the evaluation of the risks derived from Pd-NP exposure, and, in a precautionary manner, to guide the adoption/implementation of comprehensive preventive and protective strategies to manage such emerging risks and protect the health and safety of workers.