Antiangiogenesis therapy has become a hot field in cancer research. Given that tumor blood vessels often express specific markers related to angiogenesis, the study of these heterogeneous molecules in different tumor vessels holds promise for advancing anti-angiogenic therapy. Previously using phage display technology, we identified a targeting peptide named GX1 homing to gastric cancer vessels for the first time. However, GX1 also showed some non-specific binding with normal gastric vessels, which can lead to toxic side effects on normal endothelial cells. Therefore, we urgently need to adopt new screening strategies to avoid non-specific binding to normal vessels and obtain gastric cancer vascular targeting peptides with higher specificity.
In this study, we designed a new strategy which combined “positive screening” in vivo and “negative screening” in vitro for the first time. An in vivo positive screening was conducted using tumor bearing nude mice to identify peptides that were specifically enriched within the vasculature of gastric cancer. Concurrently, an in vitro negative screening process was conducted on normal vasculature endothelial cells, including human umbilical vein endothelial cells (HUVECs) and human microvascular endothelial cells (HMVECs), to eliminate peptides binding to normal vasculature. After four rounds of iterative screening, a targeting peptide specifically targeting gastric cancer vasculature was obtained. In addition, an in vitro co-culture model by culturing HUVEC in tumor conditioned medium (Co-HUVEC) was established to investigate the affinity of these peptides. The targeting peptide was labeled with fluorescein isothiocyanate (FITC) for competitive and inhibitory assays.
Blood vessel density analysis confirmed redundant capillary vessels in the xenografts, indicating that the mouse model was suitable for positive screening. Following four rounds of panning, a significant enrichment for phages specifically binding to gastric cancer vasculature was observed, with minimal binding to normal endothelial cells. The peptide CNTGSPYEC exhibited the highest reproducibility. In vitro immunofluorescence staining confirmed that the peptide CNTGSPYEC could specifically enrich in Co-HUVECs while showing no binding to normal vascular endothelial cells. In vivo immunofluorescence staining revealed that the peptide CNTGSPYEC could only bind to vascular endothelial cells specifically in gastric cancer but show no non-specific binding with normal tissue. Competitive and inhibitory assay also verified the targeting characteristics of the peptide with the fluorescence intensity of 17.13. As the concentration increases, the competitive inhibition rate can be incrementally raised to 93% (_p_< 0.05). Endothelial tube formation assay indicated that the peptide could suppress neovascularization, with the microvessel count reducing by 40% (_p_< 0.05). Furthermore, Cell Counting Kit-8 assay (CCK8) showed that the targeting peptide could partly inhibit cell proliferation of Co-HUVEC (61.7%).
Our novel strategy of the combined in vitro and in vivo screening outperforms previous methods that relied solely on negative/positive screening. In vivo and in vitro test confirmed the high targeting characteristic of the new peptide. Therefore, the peptide CNTGSPYEC may be a potential candidate in diagnosis and anti-angiogenesis therapy of gastric cancer. Our further exploration employs it as a vehicle for mediating drug accumulation in gastric cancer tissue.
Angiogenesis and vasculature are critical for tumor growth, progression and metastasis. The key to vascular-targeted therapy is to identify specific molecules within tumor vascular endothelial cells and use them as targeting agents to combat cancer, while sparing the normal vasculature from potential toxic side effects.
Although numerous studies have delved into the unique characteristics of tumor vessels compared to normal vessels, the identification of specific molecules within gastric cancer vascular endothelial cells remains a challenge. This is primarily due to the difficulties in isolating and culturing gastric cancer endothelial cells, as well as the low-level expression of tumor vascular heterogeneity molecules, which makes them challenging to detect, isolate, and purify using standard techniques. Recently, several technologies are available for the screening of vascular target molecules, among which the random phage display peptide library technology stands out. This technology integrates the binding activity of target molecules with the ease of phage amplification, resulting in an efficient, cost-effective, and highly feasible screening method.
In our prior research, we utilized phage display technology to identify a targeting peptide, GX1, that can specifically bind to gastric cancer vessels. However, the efficacy in tumor targeting of GX1 is insufficient, and it also showed some non-specific binding to normal gastric vessels. Consequently, the current study was designed to improve the screening strategy.
Specifically, we established an in vivo model for positive screening and two in vitro cell line models for negative screening for the first time. Through an in vivo positive screening approach in tumor-bearing nude mice, we aim to obtain a specific binding peptide that accumulates within gastric cancer vasculature. Simultaneously, we will conduct an in vitro negative depletion screening on normal endothelial cells to eliminate peptides that bind to normal blood vessels. After four rounds of rigorous selection, we aim to achieve a targeting peptide that specifically targets gastric cancer vasculature. This study will provide potential drugs for gastric cancer vascular targeted therapy. So as to improve the prognosis of gastric cancer.
Human gastric carcinoma cells (MKN45),human umbilical vein endothelial cells (HUVECs) and human microvascular endothelial cells (HMVECs) were purchased from the Chinese Academy of Sciences Cell Bank (Shanghai, China) and cultured in M200 basal medium (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS), low serum growth supplement (Cascade Biologics, USA), 100 U/mL penicillin, and 10 mg/mL streptomycin. The cells were maintained in a CO2 incubator (Forma Scientific). Co-HUVECs were prepared using the method established Jaffe et al. The MKN45 and HUVEC cells were co-cultured in transwell plates with a pore diameter of 0.4 μm. Briefly, two cells were separated by semi-permeable membranes of transwell plate, where the upper and lower chambers were interlinked and thus HUVEC and MKN45 cells could interact with each other by secreted soluble factors, which simulates the tumor microenvironment.
Nude mice (4–6 weeks old, 16–20 g) were purchased from Zhuhai BesTest Bio-Tech Co., Ltd. (Guangdong, China). All animals were maintained within specific pathogen-free facilities at the Laboratory Animal Center of Nanfang Hospital. All animal protocols were approved by the Animal Care and Use Committee of Nanfang Hospital (permit and approval number: 202105A028) and performed in accordance with the institutional guidelines of Nanfang Hospital.
Following the SRC method described by Bogden et al., four-week-old female nude mice with an average weight of approximately 16 g were selected. The mice were anesthetized by intraperitoneal injection, and under aseptic conditions, the kidney was isolated, the renal capsule was separated. Fresh gastric adenocarcinoma specimens resected surgically were washed several times with serum-free RPMI1640 (Gibco) culture medium to remove connective tissue and necrotic tissue. The tissue was then cut into 1 mm 3 pieces and implanted into the subcapsular space of the mouse kidney. Seven days later, the transplanted tumors from the nude mice were dissected, fixed with 10% paraformaldehyde, routinely embedded in paraffin, and then sectioned and stained with HE. The tumor tissue was observed under a microscope to assess the angiogenesis.
The Ph.D.-C7C Phage Display Peptide Library Kit (New England Biolabs, Beverly, USA) was employed to screen for specific peptides that bind to the tumor vasculature in a nude mouse subcapsular human gastric cancer xenograft model. The phage display library contains random cyclic heptapeptides constrained at the N terminus of the minor coat protein (cpIII) of M13 phage. The library has a high titer of 2 × 10 13 plaque-forming units (pfu) and a complexity of 1.2 × 10 9 individual clones. This represents the complete range of possible cyclic 9-mer peptide sequences, with a structural constraint imposed by a disulfide bond between two cysteine residues flanking the variable region. This constraint allows for a wide range of sequences with no obvious positional biases, as revealed by extensive sequencing of the naive library. Escherichia coli host strain ER2738 (A robust F+ strain with a rapid growth rate, New England Biolabs) was used for M13 phage propagation.
In vivo phage selection was performed as described with modifications.
