GFNPs show strongest signal in orthotopic pancreatic cancer tumor derived from SW1990 cells, followed by liver and kidney tissues. Such phenomenon in pancreatic tumor tissue consistent with our previous results in primary liver cancer based on tumor cell membrane permeability target mechanism (CMPT)[26]. On the one hand, GFNPs were primarily endocytosed to Kupffer cells within the liver. On the other hand, small water-soluble nanoparticles with the proper surface charge such GFNPs may get targeted to reticuloendothelial cell cells in the liver, spleen and other organs, finally excreted by the kidney [29]. All primary tumors originate from tumor cells acting on surrounding normal cells, which tumor cells can alter the gene expression of normal cells in the same environment, mainly in the form of activation of proto-oncogenes and inactivation of anti-oncogene, and the gradual transformation of normal cells into cancer cells through immune evasion [30]. Fluorescence signal of GNPs inside the tumor tissue changed weaker from one point to the surrounding area, similar to the invasion pattern of tumor cells. Besides, GFNPs fluorescence signal intensity was proportional to the tumor proliferation cycle (Fig. S2). GFNPs showed decreased distribution in the liver, lung and kidney tissue, obviously increased distribution in the orthotopic pancreatic tumor tissue compared with normal control mice. Herein, we hypothesized that such changes of GFNPs fluorescence signal were related to certain cells in the TME of pancreatic cancer, of which affected the tumor proliferation. Single-cell suspension of PDAC tumors from untreated patients were processed for scRNA-seq to identify the cell populations present in the tissue, which including cancer cells, fibroblasts, ductal cells, etc [31]. It has been reported to increase in malignancy through interactions with the abundant stroma. CAFs commonly identified by expression of αSMA, are a major source of Col1 fibers in tumors and promote tumor progression [32]. Which are major contributors to PDAC progression through pro-tumor signal and the generation of fibrosis, account for a large proportion of PDAC stroma [33], and the amount of αSMA-positive area decreased significantly as the tumor grade increased from Grade1 to Grade3 [34]. Although the concept of differentiation of mouse orthotopic pancreatic cancer has not been proposed, we found that the spatial relationship between αSMA-positive CAFs in the tumor stroma and tumor cells was related to the cycle of tumor proliferation. In the 2-weeks tumor, CAFs were widely distributed in the tumor stroma, strongly expressing αSMA near the tumor cells. There are no obvious pattern in the spatial relationship between tumor cells and CAFs. IF quantification revealed that the proportion of αSMA-positive CAFs was cut the value in half among 2-week to 4-week tumor. Studies of the 4T1 tumor showed that more than 85% of NPs were delivered to the α-SMA + stroma[35]. GFNPs localize in the nuclei of both αSMA-expressing fibroblasts and adjacent tumor cells, demonstrated outstanding dual-targeting ability which can simultaneously target cancer cells and CAFs without complex targeting ligand modifications [36]. Given that therapeutic bottleneck of PDAC arises from the dense stroma. Such smaller NPs are better for future therapeutics due to their deeper tumor penetration [37, 38]. NPs could be efficiently and rapidly internalized by macrophages and vascular smooth muscle cells via i.v. injection [39]. PDAC included abundant ducts and tumor neovascularization, which GFNPs possess this size advantage and ability to specifically mark the medial side of the catheter lumen, showing an in-adjacent position to Vimentin and an out-adjacent position to αSMA. In comparison with other NPs showed only a few localized in the vicinity of tumor vessels, GFNPs are not only abundantly deposited nearby tumor vessels but also penetrate tumor tissues to reach those regions far away from tumor vessels.
Current studies on the localization and quantification of individual tumor cells focus on circulating tumor cells, and researchers are proposing tumor treatments based on the capture of tumor cells [40, 41]. Tumor cells are treated like labelled in advance with bioluminescence and fluorescence when studying the interaction between tumor cells and drugs, however, such exit many problems. Wang report the design and synthesis of a trifunctional probe for seeing and counting cancer cells using both fluorescence imaging and inductively coupled plasma mass spectrometry (ICP-MS) but the components are also very complex while the function is achieved [42]. Based on simple synthetic GFNPs and excellent performance in actively targeting tumor cells, we found the size and weight of tumors formed by different numbers of tumor cells at the same time do not differ significantly, so in this case we cannot directly determine the malignancy of the tumor, but the fluorescent signal of the GFNPs show significant decrease in orthotopic pancreatic model. However, we did not reach the same conclusion in the subcutaneous pancreatic tumor model. IHC studies using an antibody against proliferating cell nuclear antigen revealed that the small subcutaneous tumors contained a larger fraction of proliferating cells than the large tumors [43]. Tumors grown in the liver had significantly lower vessel density, especially in the center coincident with central necrosis, than the subcutaneous tumors. However, macromolecular vascular permeability in the orthotopic liver tumor was significantly higher than in the subcutaneous tumor [44]. On the other hand, orthotopic tumors metastasized more readily than subcutaneous tumors due to the differences of types and ratios of various cell types between orthotopic pancreatic model and subcutaneous tumors, particularly CAFs. IHC of αSMA in tumor sections and immunoblot analysis of αSMA in tumor extracts in the orthotopic and subcutaneous tumors reveal tumors identified a higher number of αSMA-CAFs in orthotopic tumors [32]. The fluorescence intensity of GFNPs in subcutaneous tumors formed by tD-Tamato-BXPC-3 cell does not show a single linear variation, but rather the strongest fluorescence signal is found in subcutaneous tumors formed by specific cell count. GFNPs co-located with the nuclei of tumor cells in areas where tD-Tamato signals are present.
In addition to the primary focus, the tumor cells will also metastasize to the organs along the transvascular or lymphatic route. Nearly all deaths caused by solid cancers occur as a result of metastasis - the formation of secondary tumors in distant organs such as the lungs, liver, brain and bone [45]. Liver metastases are commonly detected in a range of malignancies including colorectal cancer, pancreatic cancer, melanoma, lung cancer and breast cancer. Interactions between tumor cells and the tumor microenvironment play an important part in the engraftment, survival and progression of the metastases. Various cells are implicated in promoting and sustaining metastases in the liver [46, 47]. Tumor foci in the liver had tortuous vascular architecture, heterogeneous blood flow, significantly lower vascular density, and significantly higher vascular permeability than normal liver tissue [44]. We also observed the fluorescence of GFNPs was significantly higher in the white tumor nodules of the liver than in the non-white normal liver tissue, and the increased fluorescence in the lung and pancreatic tissues indicated that the tumor cells had also invaded the adjacent lung and pancreatic tissues. CDK2 overexpression could facilitate lymph node metastasis. The overexpression of cyclin E and CDK2 may mainly promote the progression of early cancer. Increased cyclin E protein was related to elevated CDK2, which was further linked to higher Ki67 [46]. In cancerous liver sections, the fluorescence signals of the tumor nucleus antibody protein Ki67 and the nucleus PI almost completely overlapped, the GFNPs and the tumor nucleus antibody protein Ki67 showed localized signal overlap, and αSMA-CAFs were distributed in the vicinity of the double-positive areas of GFNPs and Ki67. The same phenomenon in the invaded pancreatic tissue also has been observed. However, GFNPs did not co-localize with the metastasize cancerous pancreatic αSMA-CAFs in contrast to the highly metastasize cancerous liver [48].
GFNPs fluorescently label clinical pancreatic cancer tumors with different degrees of malignancy with varying intensity[49, 50]. PDAC is an invasive epithelioid tumor, occurring in 80–90% of pancreatic cancers, usually accompanied by a ductal lumen, in which the tumor cells are often cubic and a marked pro-fibro-mesenchymal reaction is typical [51, 52]. As a highly fibrotic pancreatic tumor, GFNPs can label CAFs cells in the PDAC mesenchyme and a proportion of tumor cells [53]. pNET are a type of pancreatic tumor with predominantly neuroendocrine differentiation. Which are less common, accounting for 1–2% of all pancreatic tumors, and are also called islet cell tumors. They are well differentiated, with well-defined single nodules showing dense fibrosis [54, 55]. GFNPs demonstrate very good fluorescent labeling of cells in the tumor microenvironment of pNET, which could nearly Identify all kinds of cells in TME and co-localize fluorescently with molecules secreted by pancreatic isletsβcells [56]. ASC is an epithelial malignancy of the pancreas, accounting for approximately 2% of all pancreatic malignancies, with ductal and squamous epithelial differentiation [57]. ASC tumors often contained necrotic islets, which constitute a known clinical feature of ASC. Such could have been fluorescently labeled specifically by GFNPs. Besides, it exits a strong fluorescent signal inside cells surrounding Ki67-positive tumor cells, but the fluorescent signal does not overlap with CAFs in the stroma. In summary, GFNPs have the strongest fluorescent signal for pNET-labeling but are less specific, with better specificity for fluorescent labeling of the other two malignancies.
Gemcitabine is the first line and the gold standard drug for pancreatic cancer. However, the anticancer efficacy is severely limited by its instability and poor cellular uptake[58]. To enhance the clinical efficacy of GEM, we constructed a novel nanodrug delivery system based on dual-targeting performance of GFNPs against tumor cells and CAFs of pancreatic cancer. This GFNPs-GEM nanoformulation possess a high drug loading. With the features of small size and stable formulation, the obtained GEM nanoformulation could effectively accumulate in tumor site and rapid uptake in cells. GFNPs-GEM nanoformulation displayed more potent anticancer activity compared to free GEM in vivo. Moreover, the nanodrug displayed greatly reduced adverse effects and satisfactory biocompatibility. Benefiting the advantageous features of the dual-targeting performance of GFNPs against tumor cells and CAFs and nanotechnology-based drug delivery, this GEM nanosystem constitutes a promising therapeutic candidate for pancreatic cancer treatment. Our study also underlines the potential use of the dual-targeting performance of GFNPs in TME for improving drug efficacy as well as reducing drug toxicity (Fig. 9).