Modulation and generation of plasma membrane cation channels by silica nanoparticles

Silica nanoparticles (SiNPs) have numerous beneficial properties and are extensively used in cosmetics and food industries as anti-caking, densifying and hydrophobic agents. However, the increasing exposure levels experienced by the general population and the ability of SiNPs to penetrate cells and tissues have raised concerns about possible toxic effects of this material. The airway epithelium is one of the most important entry routes of SiNPs into the body. However, the molecular mechanisms whereby SiNPs affect the functions of this tissue remain largely unknown. Recent studies indicate that sensory TRP channels expressed in the respiratory tract are involved in specific cellular responses to particulate matter. These channels have been identified as mediators of different environmental toxic agents in the airways including apoptotic and pro-inflammatory responses.

Considering these precedents and the fact that TRP channels are highly expressed in the sensory neurons and in epithelial cells where they are involved in the mechanisms of mechano- and chemo-sensation, we hypothesized that TRP channels may be modulated by SiNPs and play a key role in cellular and systemic responses to nanoparticles. To address this hypothesis we initially outlined the following aims: 1) to evaluate the effect of amorphous SiNPs on different TRP channels in heterologous and native expression systems. 2) to determine the consequences of these effects on cellular processes depending on the activity of the TRP channels. 3) to characterize the biophysical and pharmacological properties of the effect triggered by SiNPs combined with other TRP channel agonists.

In preliminary experiments we found TRPV4 to be the most sensitive sensory TRP channel to SiNPs, displaying a strong reduction in its response to the synthetic agonist GSK1016790A. We therefore focused on the study of this channel. Using fluorometric measurements of intracellular Ca2+ concentration ([Ca2+]i) we found that SiNPs inhibit activation of TRPV4 by the synthetic agonist GSK1016790A in cultured human airway epithelial cells 16HBE and in primary cultured mouse tracheobronchial epithelial cells. Inhibition of TRPV4 by SiNPs was confirmed in intracellular Ca2+ imaging and whole-cell patch-clamp experiments performed in HEK293T cells over-expressing this channel. In addition to these effects, SiNPs were found to induce a significant increase in basal [Ca2+]i, but in a TRPV4-independent manner. SiNPs enhanced the activation of the capsaicin receptor TRPV1, demonstrating that these particles have a specific inhibitory action on TRPV4 activation. Finally, we found that SiNPs abrogate the increase in ciliary beat frequency induced by TRPV4 activation in mouse airway epithelial cells. Our results show that SiNPs inhibit TRPV4 activation, and that this effect may impair the positive modulatory action of the stimulation of this channel on the ciliary function in airway epithelial cells. These findings unveil the cation channel TRPV4 as a primary molecular target of SiNPs.

The fact that SiNPs inhibit the activation of TRPV4 by GSK1016790A, prompted us to determine their possible effects on natural TRPV4 agonists such as LPS, AA and heat. Using Ca2+ imaging experiments we demonstrated that SiNPs in combination with LPS, AA or heat induce an increase in intracellular [Ca2+]. This effect was found in cells expressing TRPV4 (TRPV4 transfected HEK293T cells with mouse TRPV4, native HEK293T, 16HBE and mTEC cells), but also in non-expressing TRPV4 (CHO cells), indicating that the effect was independent of this channel. We confirmed the lack of TRPV4 role by application of the TRPV4 inhibitor HC067047, which failed to inhibit the Ca2+ signal. The intracellular Ca2+ response induced by the combination of SiNPs and LPS, AA or heat is due to Ca2+ entry pathway through the plasma membrane. This was demonstrated by a lack of intracellular [Ca2+] increase when Ca2+ was absent in the extracellular solution and by the intracellular Ca2+ signal blockage by RR.

Finally, we aimed at characterizing the pharmacological and biophysical properties of the Ca2+ entry pathway induced by SiNPs and AA. Using whole-cell patch-clamp we measured robust outward currents induced by the SiNPs-AA combination. The amplitude of the SiNPs-AA-induced currents increased with the SiNPs concentration, with an effective concentration  of 1.1 ± 0.5 µg/ml SiNPs for currents measured at -75 mV. These values are extremely low in comparison to those typically used in cell toxicity experiments, suggesting for a relevance of our findings in the understanding of the cellular actions of SiNPs in the context of airway inflammation featuring increased levels of AA.

The amplitude of SiNPs-AA-induced currents displayed a voltage dependency similar to that of TRPM7. However, we demonstrated the lack of role of this channel by showing that the use of high concentration of Mg2+ in the pipette solution to block TRPM7, failed to inhibit the currents triggered by SiNPs-AA combination.

We found that SiNPs-AA-induced currents are mainly carried by cations, and estimated a relative permeability of Na+ over Cl- of ~50. The cationic nature of these currents was supported by experiments showing a voltage-dependent inhibitory effect of the unspecific cationic channel blocker RR. We also found that the SiNPs-AA-induced ionic pathway poorly discriminates between monovalent cations, and found an Eisenman I selectivity sequence (Cs+ > Rb+ > K+ > Na+ > Li+), which is indicative of a permeation pathway containing a weak-field binding site for cations. Subsequent ionic substitution experiments allowed us to confirm that the SiNPs-AA-induced ionic pathway sustains Ca2+ permeation and that it displays little selectivity between divalent cations, with a sequence of relative permeability with respect to Na+ of Ca2+ > Sr2+ ≥ Ba2+ > Mg2+. We found that heavy metals reversibly blocked SiNPs-AA-induced currents with a potency sequence:  La3+ > Zn2+ > Ni2+ > Cd2+ > Co2+, and that La3+ inhibits the currents measured at -75 mV with an effective inhibitory concentration of 0.033 µM.

Interestingly, SiNPs-AA-induced currents displayed multiple typical features of currents carried by voltage-gated channels, including activation and deactivation kinetics and a voltage dependence of current amplitudes that could be well-described by the multiplication of the whole-cell conductance, the driving force and a Boltzmann curve featuring a voltage for half-maximal activation (Vact) of ~40 mV. The associated slope factor (sact) was relatively large (~25 mV), which is much larger than those of “classical” voltage-gated channels and closer to those of voltage-gated TRP channels.

At this point it is impossible to put forward a molecular mechanism explaining the SiNPs-AA-induced currents, but taken together, our results suggest that these currents are due to the formation of cationic pores in the plasma membrane. We tentatively “baptise” these as nanoparticle-induced pores, or NPIP. Interestingly, the fact that NPIP were observed when combining SiNPs and AA, LPS or heat strongly suggest the highly feasible hypothesis that the nanoparticles are more effective in conditions in which the membrane fluidity is increased. We propose that this Ca2+ entry pathway may be relevant for the toxicological properties of SiNPs in cells.

As future perspectives we propose using as working hypothesis that NPIP properties can be studied using the biophysical methods applied for the study of classical voltage-gated ion channels. For this we foresee the evaluation of the existence of “single-pores” (in analogy to single-channels), by measuring NPIP currents in cell-attached recordings. In addition, it would be interesting to estimate the pore size by monitoring the relative permeability of cations of increasing diameter.

We also hypothesize that NPIP may display a set of pharmacological properties that distinguish them from other cationic pores. To test this hypothesis we may further characterize the effects of inorganic cation channel blockers (Ni2+, Zn2+, Cd2+, Co2+, La3+) on NPIP, in terms of concentration, voltage dependence, and effects on current kinetics. In addition we may determine the effects of organic voltage-gated Ca2+ channel blockers (dihydropyridines, phenylalkylamines and benzothiazepines) and of TRP channel inhibitors on NPIP and perform a medium throughput screening of NPIP modulators out of a library of 2400 compounds using medium throughput measurements of [Ca2+]i in a Flexstation 3 Reader and an automated patch-clamp setup (Patchliner, Nanion).

Finally, since we found NPIP in airway epithelial cells (see above), we hypothesize that these pores can be formed also in other native cells known to be in contact with polluting nanoparticles. To test this hypothesis we may determine whether similar pores can be formed in mouse airway macrophages and dentritic cells and skin and oesophageal keratinocytes. This screening may be performed using intracellular Ca2+ imaging, using the properties of NPIP found in the above-described experiments in HEK293T cells as features for comparison.

These studies may shed light on the thus far elusive molecular mechanisms underlying nanoparticle-induced toxicity and may set the bases for a subsequent project aimed at determining to what extent Ca2+ entry via NPIP results in altered cellular functions and tissue damage. In the future we may also incorporate in vivo experiments in mice, providing for a translational approach through the comparison of results obtained in mouse and human cells.