Advanced histology methods for biodistribution and safety screening of next generation nanobiomaterials

Abstract

Over the past few decades nanobiomaterials (NBMs) have shown a promise in improving many areas of medicine such as diagnosis and treatment of diseases, with number of formulations currently approved for use by the regulators. However, there are still concerns regarding biodistribution, accumulation, and toxicity of NBMs in vivo as well as their impact on the immune system. To gain a better understanding on the complex interplay of NBMs with biological systems and structures, assays used in preclinical assessment of the next generation NBMs need to be expanded to address those specific concerns. This PhD project aimed to contribute to the scientific knowledge on the biodistribution, accumulation, and toxicity of representative NBMs in vivo by improving their detection in tissue sections by means of accepted histological markers and methodologies. The second aim was to provide an efficient and objective method for semi-quantitative analysis of histological data by utilising machine learning. To achieve those goals, the research undertaken in this PhD project develops in four main case studies, where histological techniques were specifically tailored to the type of NBMs under investigation: namely liposome and iron oxide nanoparticles. The first two case studies presented in chapter 3 were undertaken to address the lack for consistent microscopic detection of liposomes in histological tissue sections. For this purpose, two type of liposomes were selected as NBM models: a liposome containing fluorescent dye and a PEGylated liposome. By using histological stain for lipid detection (Oil red O), liposomes were successfully detected with fluorescent microscopy in different organs (liver, spleen, kidney, and small intestine). Moreover, by imaging fluorescent stain incorporated in the first liposome, or immunofluorescence labelled PEG of the second liposome, these NBMs were effectively discriminated from naturally occurring dietary lipids or fats. The following two case studies presented in Chapter 4 and 5 addressed the need for a subjective and time efficient method for quantitative histology studies. This was achieved by utilising machine learning to assess biodistribution and morphology of aggregates formed by magnetic iron oxide nanoparticles, as NBMs in in vivo animal murine model. Perls? Prussian blue stained was used to stain iron in tissue sections, with quantitative analysis carried out in traditional 2D histology slides and 3D histology volumes reconstructed from a series of consecutive tissue sections. Data extrapolated from histology sides using this method provided information on patterns of accumulation of magnetic iron oxide NBMs in terms of number and size of aggregates formed in lungs, liver, kidney, and spleen following intravenous injection. Additionally, in the last case study, immune reaction in response to NBMs and their degradation in vivo was evaluated using specialised histology stains. Results of this analysis suggested that dextran used to coat the magnetic iron oxide nanoparticles induced an acute immune response in lungs which was not seen in the NBM when this coating was degraded. Overall, the work presented in this thesis highlights the extensive potential that tailored histology could have in evaluating the safety of NBMs in vivo. Various specialised stains allowed to detect NBMs in tissue sections on a microscopic level, to evaluate their degradation in vivo and type of immune reaction activated in response to the presence of these novel type of materials. Furthermore, traditional histological methods were brought into the twenty-first century by using ML to extrapolate quantitative data from representative images showing iron aggregates in 2D and 3D histology volumes. Future work in this area should evaluate if this type of analysis can be extended to assess the safety of other types of NBMs.