br Given the substantial heterogeneity among cancers it is
Given the substantial heterogeneity among cancers, it is unlikely that any single protein molecule will have adequate specificity and sensitivity needed for accurate early diagnosis. There is a growing consensus that multiple markers, used individually or as part of a panel, are required for early disease detection. To better comprehend the potential use of the in vivo protein corona for biomarker discovery, we statistically compared the level of proteins present in the coronas formed in healthy and tumor-bearing mice. Progenesis analysis re-vealed 384 and 328 diﬀerent proteins diﬀerentially expressed in tumor compared to healthy corona samples, in the melanoma (Fig. 3A) and lung-carcinoma model, respectively (Fig. 5A). It should be noted that the equivalent analysis of plasma samples in the case of melanoma model, revealed only 6 potential biomarker proteins (Table S6).
Proteins released from the tumor microenvironment, by leakage or secretion, are massively diluted in the blood circulation and current proteomic methodologies often fail to provide suﬃcient depth of ana-lysis, required for their detection. Classification of the melanoma and lung carcinoma associated corona proteins identified in this study ac-cording to their cellular localisation, demonstrated that circulating li-posomes act as platforms for interaction with intracellular, tissue leakage products. This was further exemplified by the detection of human corona proteins released by lung carcinoma NPS-2143 in the blood circulation of SCID mice (Fig. 5C).
On a broader context, our findings suggest that the adsorption of low abundant biomolecules onto the surface of NPs, once injected in the bloodstream allows thus their successful identification. The nanoscale liposomal corona is enriched by multiple plasma proteins with potential diagnostic and disease monitoring value which renders the need for much more work on the biomarker verification and validation front necessary, but beyond the scope of this study. In contrast with other NP-based assays that are design to quantify already known blood bio-markers, the technology proposed here could facilitate the initial dis-covery phase of the cancer biomarker pipeline in which animal models are mainly used, aiming to reveal previously unseen molecules diﬀer-entially abundant between healthy and diseased states. When animal models are employed for plasma biomarker discovery, the exploitation of the molecularly richer in vivo protein corona is beneficial in com-parison to its counterpart ex vivo corona, however further investigations are required to assess the potential use of the ex vivo corona directly from human plasma samples. It is also important to understand the potential limitations of characterising protein corona fingerprints at a single time point post-inoculation of tumor cells in mice and more studies are needed in order to investigate the evolution of protein corona at diﬀerent stages of tumor growth and prove its potential at the Biomaterials 188 (2019) 118–129
early asymptomatic stages of cancer.
In this proof-of-concept study, we provide previously unreported experimental evidence of the potential exploitation of the spontaneous and often undesirable protein coating of NPs, once injected in the blood circulation, to enhance plasma proteomic analysis and to sample dis-ease-specific proteins. We demonstrated the ability of intravenously administered clinically used NPs to scavenge the blood pool of mela-noma (B16-F10) and lung carcinoma-bearing (A549) mice and to sur-face-capture a complex mixture of tumor-specific proteins that could not be identified by plasma sample analysis that was performed in comparison. The successful recovery and purification of corona-coated liposomes from the blood circulation of mice allowed the elimination of the overwhelming signal of highly abundant plasma proteins and en-abled the detection of low abundant tumor-released proteins. Protein corona was found to qualitatively and quantitatively diﬀer in the ab-sence and presence of inoculated tumors enabling the identification of 384 and 328 diﬀerentially abundant proteins for melanoma and lung adenocarcinoma respectively. The success of the nanoparticle-enabled retrieval of low abundance plasma proteins was further revealed by the concurrent detection of murine host response proteins and human in-oculated tumor-released proteins in a lung adenocarcinoma xenograft model.