On February 13, 2023, Professor Huang Xinglu of Nankai University, Yan Xiyun’s team from the Institute of Biophysics of the Chinese Academy of Sciences, and the team of Tian Jie from the Institute of Automation of the Chinese Academy of Sciences published an article entitled “Machine-learning-assisted single-vessel analysis of nanoparticle permeability in tumour” in the journal Nature Nanotechnology vasculatures”.
This study proposes a single-vessel quantitative analysis method based on protein nanoprobes and machine learning to quantitatively study the process by which nanoparticles cross blood vessels to reach tumors. Studies have found great heterogeneity in vascular permeability of different tumors, and successfully developed protein nanoparticles with enhanced transendothelial transport ability, which points the direction for the rational design of the next generation of anti-cancer nanodrugs.
The corresponding authors of the paper are Huang Xinglu, Yan Xiyun, Tian Jie; The first authors are Zhu Mingsheng and Zhuang Jie.
In recent years, nanodrug targeted delivery for the treatment of tumors has become a research hotspot, and its mechanism mainly relies on the enhanced permeability and retention (EPR) effect, that is, the high permeability and long-stay effect of nanoparticles in tumor tissue. However, with the failure of multiple nanomedicine clinical studies, research on this effect has gradually been questioned. Recent studies have shown that up to 97% of nanoparticles cross blood vessels into tumors through active transendothelial transport mechanisms, rather than passive extravasation mechanisms. Nevertheless, more researchers believe that the penetration ability of nanodrugs in different tumor blood vessels is heterogeneous, and the mechanism of nanoparticle penetration in different types of tumors is quite different. However, the current endorsement of this conclusion is more of a qualitative or empirical judgment, and effective quantitative analysis methods are lacking to support this argument.
Therefore, Professor Huang Xinglu of Nankai University and his collaborators proposed a single vessel quantitative analysis method (Nano-ISML) based on machine learning to quantitatively study the process of nanoparticles passing through blood vessels to reach tumors. After quantitative analysis of > 67,000 blood vessels in 32 different tumor types, the researchers showed in the form of data the heterogeneity of the vascular penetration ability of nanoparticles in different tumor types, and the osmotic mechanism of nanoparticles in different tumor types is also heterogeneous. Based on this discovery, the researchers successfully developed protein nanoparticles with enhanced transendothelial transport capabilities, which provides direction for the rational design of the next generation of anti-cancer nanodrugs.
Figure 1: Single-vessel quantification method based on protein nanoprobes and machine learning.
Figure 2: Quantitative analysis of vascular permeability for different tumor types using the Nano-ISML method.
Using the uniform size of ferritin nanocages (FTn), the researchers developed a tumor visualization technology based on protein nanoprobes, and developed a machine learning-based single-vessel analysis method (Nano-ISML) based on the large number of confocal images obtained by this technology. Using Nano-ISML, the researchers quantitatively analyzed > 67,000 blood vessels from 32 different tumor types, found that there was a more than 100-fold difference in the permeability of different blood vessels, and artificially divided different tumor blood vessels into three osmotic types, namely hypotonic, mesolatic and hypertonic blood vessels, based on quantitative data, and this difference depends on the type of blood vessel and tumor type.
Figure 3: Heterogeneity of vascular permeability mechanisms.
Next, the researchers generated magnetic Fe3O4 particles inside ferritin and observed the mechanism by which they passed through different tumor blood vessels by electron microscopy. The study found that due to the large number of vascular openings, the penetration of high-permeable blood vessels by nanoparticles mainly relies on passive extravasation through the endothelial space and VVO (vesiculo-vacuolar organelle), while low-permeability blood vessels mainly rely on endothelial cell capture and transendothelial transport. Vascular osmotic heterogeneity for different tumors depends on the ratio of highly permeable vessels and hypopermeable vessels.
Figure 4: Characterization of ferritin and its modified nanoparticles.
Based on the above findings, the researchers developed protein nanoparticles with enhanced transendothelial transport capabilities to enhance the vascular penetration ability of nanoparticles in hypopermeability tumors. Human serum albumin (HSA) is bound to the ferritin surface and lysosomal escape functional peptide H2E is modified to enhance the interaction of nanoparticles with vascular endothelial cells, while reducing lysosomal degradation in endothelial cells and increasing transendothelial transport of nanoparticles. The synergistic effect of these two strategies has been confirmed to effectively enhance ferritin cross-cell transport and improve its vascular penetration ability in hypopermeability tumors, and has shown stronger tumor enrichment and therapeutic potential than unmodified ferritin. Studies have also shown that this modified ferritin nanoparticle does not significantly enhance the vascular permeability and therapeutic effect of hypertonic tumors.
Figure 5: In vivo anti-tumor capacity of different ferritin preparations.
This study systematically elucidates the high heterogeneity of tumor vascular penetration through high-throughput quantitative analysis of tumor single blood vessels, and reveals the biological mechanism of the formation of vascular endothelial cell osmotic heterogeneity, and proposes a new strategy for differentiated design of nanomedicines for this mechanism, which will provide theoretical basis and design principles for the development of next-generation personalized nanomedicines. The deep integration of nanotechnology, artificial intelligence and synthetic biology also provides a new paradigm for solving key scientific problems at the intersection of disciplines. This work was supported by the National Natural Science Foundation of China (91959129, 32271448, 82072054, 31870999), the National Key Research and Development Program of China (2022YFA1105100), and the Tianjin Synthetic Biology Special Project (TSBICIP-KJGG-014-03). (Source: Science Network)
Related paper information:https://doi.org/10.1038/s41565-023-01323-4