Immunosafety
As mentioned above, despite attempts to design nanomedicines capable of escaping immune recognition and entrapment, the majority of injected nanodrugs does still go to the RES. Even nanomedicines that do not carry cytotoxic drugs and are not directly toxic for the immune cells can have an effect in terms of alteration of the normal development of the defense reactions. While in the absence of challenges this may not be seen, what has to be considered is if whether immune cells loaded with nanomedicines will be as efficient as normal cells in reacting to a danger (e.g., an infection) and adequately protect our body.
If immune cells engulfed with NPs are impaired in their defensive and surveillance capacity, this may leave the organism unguarded with the risk of developing severe infections and tumors (immunosuppression). On the other hand, there is also the possibility that immune cells exposed to NPs may react exaggeratedly to harmless stimuli or wrongly recognize self-antigens, with the risk of developing chronic inflammatory and autoimmune diseases. All these diseases are multifactorial, with the possible contribution of genetic, nutritional and environmental factors to their pathogenesis. In this context, anomalous innate immunity/inflammatory reactions are considered a major initiating event (see for example [45,46]). Thus, testing the effects of nanomedicines on the normal development of the innate immune and inflammatory responses is of key importance when assessing the immunosafety profile of a nanodrug.
To this end, specific and representative assays need to be designed. The use of animal models in immunological research and testing is wide and very important and, in several cases, difficult to replace. How much animal models are representative of the human situation is an issue that needs to be accurately assessed. An example may be that of the most used allergy model in the mouse (ovalbumin as allergen in presensitized BALB/c mice), a model in which neither the allergen nor the biological system (BALB/c mice are genetically prone to mount Th2 allergic responses) reflect the development of an allergic reaction in humans. Indeed, in the human population a very low number of individuals respond with an allergic reaction to an antigen that is otherwise harmless for the vast majority of the population. In several cases no strict molecular correlation has been found between immune reactions in mice versus humans, and there is evidence of alternative molecular pathway usage.[47] With the clinical use of nanomedicines, the possibility of monitoring some immunological parameters in patients and in healthy volunteers would be of great help, in particular for assessing the kinetical evolution of responses. For screening purposes, in vitro assays that are currently used have the advantage of rapidity and robustness, by using murine or human continuous cell lines with stable characteristics. Since these cell lines are mostly tumor derived or transformed, the possibility that they may not reflect what happens in normal primary cells must be evaluated. For example, cytotoxicity of NPs on a highly proliferating macrophage cell line can be quite different from its toxic effect on primary resident macrophages that are noncycling cells. On the other hand, testing for instance, the ability of NPs to induce TNF-α in continuous versus primary macrophages may yield very similar results. Thus, the use of cell lines is very important for rapid and reproducible testing, but only for validated biomarkers that reflect the response of normal human cells. Therefore, as far as possible, the immunoassays for predicting the nanorisk for human health should be based on human primary cells, and cell lines can be used afterwards for validated markers only. In the case of the mouse allergy models mentioned above, it is clear that in vitro models based on human cells are needed. Primary human cells from allergic patients exposed to NPs plus patient-specific allergen may yield information on the interaction between specific NPs and allergens. This can be highly relevant since allergens may be part of the NP bioshell and thus arrive within a specific context. It is possible that the association of allergens could lead to an ordered structure of allergens, which may even result in repetitive patterns, a type of structure that is particularly well recognized by the immune system. Of note, allergens are generally recognized in specific contexts, often associated with particles, such as pollen fragments. For the study of allergic asthma, appropriate mouse models exist and should be utilized. NPs may not only show more pronounced effects in diseased versus healthy lungs, but due to the breach of barriers that occurs in chronically inflamed tissues, they may also have better access to other parts of the body, resulting in systemic differences in the response of healthy and sick organisms.
Another major point that should be considered when testing immune responses is that these are highly dynamic reactions with a clear evolution. Thus, the immune response is activated by 'danger' signals (including, for instance, foreign particles entering the body) and, once the danger is eliminated, the response is readily deactivated. Pathology occurs when the defense response is anomalous in extent or duration, but never or very rarely is there a qualitative difference. This implies that the molecular end points one can select are not unique to pathological risk, and what distinguishes physiological from pathological response is its duration, extent and distribution. For example, IL-1β is a key initiator of the inflammatory response and of the defense reaction, which must be produced in the tissue in order to initiate the defensive reaction. However, IL-1β can cause severe tissue damage when present in the wrong place in the wrong amount and at the wrong time. For instance, the presence of IL-1β in the circulation causes diffuse intravascular coagulation and hypotension (as in acute myelogenous leukemia and in septic shock; reviewed in [48,49]), while its prolonged presence in an inflammatory site can cause destructive tissue damage (e.g., in rheumatoid arthritis or inflammatory bowel disease).[50] Thus, production of IL-1β can be an excellent biomarker of inflammation, but it does not discriminate pathological inflammation from physiological reaction unless it is tested in a more complete fashion, for example in a kinetic study.
Eventually studies examining the immunosafety of nanomedicines should also take into consideration the particulate nature of the drug. For instance, NPs should be tested for interference with the detection assay used to identify the selected biomarkers (e.g., a colorimetric assay), which means that nanoimmunosafety assays should be specifically designed for NP testing.
The following characteristics are reviewed for a reliable nanoimmunosafety assay:
Rapid and easy to perform (e.g., in vitro assays);
The ability to take into account the physicochemical characteristics and behavior of NPs (e.g., the optical interference with assay readouts);
Predictive of the NP effects in vivo in humans (e.g., using human primary cells in vitro and kinetic evaluation of biomarkers);
Robust and reproducible (e.g., standardized with validated cell lines and selected reporter genes).
To address these points, we have proposed a human model of innate/inflammatory defense response that recapitulates in vitro the different phases of the defense response, from recruitment and initiation, to development of inflammation, and its eventual resolution [Boraschi D et al., Manuscript in preparation]. The model is based on human primary blood monocytes exposed in culture to sequential changes of microenvironmental conditions (chemokines and cytokines, temperature, bacterial-derived molecules) for 24 h. Profiling of the innate reactivity during the development of the reaction allows the identification of dynamic signatures that characterize the 'physiological' innate defense response. As an example, the release of the inflammatory cytokine IL-1β peaks at the time of the full development of the inflammatory reaction and decreases thereafter. On the other hand, the IL-1β inhibitor IL-1Ra is still produced at its maximum amount in the phase of resolution, since inflammation resolves when IL-1β is effectively inhibited by IL-1Ra (Figure 3).
Figure 3.
A kinetical in vitro assay for profiling the physiological innate immune response of human blood monocytes. The physiological innate immune defense reaction (physiological inflammation) is reproduced in vitro with human monocytes. CD14+ monocytes isolated from peripheral blood of normal donors were exposed in vitro to a sequence of stimuli mimicking the development of the inflammatory reaction (chemokine [C-C motif] ligand 2 from 0 to 2 h at 37°C, lipopolysaccharide from 2 to 14 h at 39°C, TNF-α from 3 to 14 h at 39°C, IFN-γ from 7 to 14 h at 39°C, culture medium alone from 14 to 24 h at 39°C). The release of inflammatory IL-1β and of its specific inhibitor IL-1Ra was quantitatively measured at different time points by ELISA. The normal innate reaction is characterized by a significant release of the inflammatory factor at the peak response, followed by a decrease, while the anti-inflammatory factor is produced more abundantly in the resolution phase.
The inflammatory versus anti-inflammatory signature, once better defined and validated, can then be used for monitoring the possible interference and alterations in the physiological response caused by nanomedicines in comparative experiments. Analogous models of 'pathological' inflammation (mimicking chronic inflammatory and autoimmune diseases) are being set up and validated in order to define pathology-related molecular signatures. These, in turn, will be used as reference benchmarks for evaluating whether the NP-induced changes can be associated with a pathology-related signature, thus allowing us to predict risk.
While for inflammatory monocytes a reliable in vitro model can be set up with a single cell type, in the case of barrier defense more complex models are required. Furthermore, in such cases, it is important that human primary cells are used, and that the architecture and the general condition of the tissue are reproduced. Along this line, for studying, for instance, the interaction between immune cells of the small intestine with ingested NPs, an epithelial layer should be set up on a suitable permeable substrate in transwell, intraepithelial lymphocytes should be introduced in the layer, macrophages and dendritic cells added to the lower surface of the substrate, synthetic mucus should cover the epithelial surface and the presence of the intestinal flora mimicked with killed bacteria. Only in these kinds of conditions can the interaction of NPs with the intestinal immune cells yield meaningful results. Studies in this direction are being performed, and are paving the way for a more reliable assessment of NP effects on human health.[51]
Nanomedicine. 2012;7(1):121-131. © 2012 Future Medicine Ltd.
