Use of Labelled tLyP-1 as a Novel Ligand Targeting the NRP Receptor to Image Glioma

Background Neuropilin (NRP) receptors are overexpressed in glioma tumor tissue, and therefore may be a potential target for imaging markers. We investigated whether labelled tLyP-1, an NRP targeting peptide, could be used as the targeting ligand for developing reagents for imaging glioma tumors. Methods The tLyP-1 peptide (CGNKRTR) was labeled with 5-carboxyfluorescein (FAM) or 18F-fluoride. A control peptide (MAQKTSH) was also labeled with FAM. The in vitro binding between FAM-tLyP-1 and U87MG cells and in vivo biodistribution of FAM-tLyP-1 in a U87MG glioblastoma xenograft model (nude mouse) were determined. The in vivo biodistribution of 18F-tLyP-1 was also determined by microPET/CT. Results In vitro, FAM-tLyP-1 was strongly taken up by U87MG cells at very low concentrations (1μM). In vivo, FAM-tLyP-1 accumulated in glioma (U87MG) tumors, but uptake was minimal in the normal brain tissue 1 h after administration. The distribution of FAM-tLyP-1 in the tumor tissue was consistent with expression of NRP1. The tumor/brain fluorescence intensity ratio in mice treated with FAM-tLyP-1 was significantly higher than the control FAM-labeled peptide 1 h after administration (3.44 ± 0.83 vs. 1.32 ± 0.15; t = 5.547, P = 0.001). Uptake of FAM-tLyP-1 in glioma tumors could be blocked by administering an excess of non-conjugated tLyP-1 peptide. [Lys4] tLyP-1 was labeled with 18F to synthesis a PET (18F-tLyP-1). MicroPET/CT imaging showed the tumor was visualized clearly with a high tumor/brain radiolabel ratio at 60 min (2.69 ± 0.52) and 120 min (3.11±0.25). Conclusion Taken together, our results suggest that tLyP-1 could be developed as a novel fluorescent or radio labelled tracer for imaging glioma.


Results
In vitro, FAM-tLyP-1 was strongly taken up by U87MG cells at very low concentrations (1μM). In vivo, FAM-tLyP-1 accumulated in glioma (U87MG) tumors, but uptake was minimal in the normal brain tissue 1 h after administration. The distribution of FAM-tLyP-1 in the tumor tissue was consistent with expression of NRP1. The tumor/brain fluorescence intensity ratio in mice treated with FAM-tLyP-1 was significantly higher than the control FAMlabeled peptide 1 h after administration (3.44 ± 0.83 vs. 1.32 ± 0.15; t = 5.547, P = 0.001). Uptake of FAM-tLyP-1 in glioma tumors could be blocked by administering an excess of non-conjugated tLyP-1 peptide. [Lys4] tLyP-1 was labeled with 18 F to synthesis a PET ( 18 F-tLyP-1). MicroPET/CT imaging showed the tumor was visualized clearly with a high tumor/brain radiolabel ratio at 60 min (2.69 ± 0.52) and 120 min (3.11±0.25).

Conclusion
Taken together, our results suggest that tLyP-1 could be developed as a novel fluorescent or radio labelled tracer for imaging glioma.

Introduction
Gliomas originate in the brain or spinal cord and account for approximately 30% of all brain and central nervous system tumors and 80% of all malignant brain tumors. High-grade gliomas are highly vascular tumors and have a tendency to spread [1][2][3]. They are not easily excised and often recur at the surgical margin. It is crucial that the glioma tumor is completely removed, while protecting surrounding normal brain tissue as much as possible, to improving patients' survival and their postoperative quality of life. However, removing the tumor completely can be challenging because it is not easy to differentiate tumor from surrounding normal brain tissue before and during surgery [4][5][6].
Positron emission tomography (PET) and fluorescent imaging are highly sensitive imaging techniques that could be used to identify microinvasions from a glioma tumor into surrounding tissue to guide surgical removal of the tumor. There are several imaging reagents that are currently available including: 18 F-fluorodeoxyglucose ( 18 F-FDG), 11 C-methionine, 11 C-choline, and 18 F-tyrosine for PET and 5-aminolevulinic acid for fluorescent imaging [7][8][9][10][11][12][13][14][15]. However, the available reagents are not tumor specific imaging agents and they are insufficiently sensitive and specific [7,[14][15]. Chlorotoxin (CTX), which can preferentially bind to matrix metalloproteinase-2 (MMP-2) in glioma cells, has been developed as an optical imaging contrast agent (CTX: Cy5.5) by conjugation to the fluorescent marker Cy5.5. CTX: Cy5.5 can potentially "paint" a glioma tumor and improve intraoperative detection and resection of malignant glioma [16][17][18]. However, CTX is a 36 amino acid peptide with four disulfide bridges bonds, and although it can effectively show glioma tumors, it would be difficult to commercialize due to its complex molecular structure [19][20]. Thus, there is an unmet need to develop a glioma specific, sensitive imaging reagent that can be easily commercialized.
The tLyP-1 peptide (CGNKRTR) is a truncated form of the cyclic tumor-homing peptide LyP-1 (CGNKRTRGC). tLyP-1 contains a CendR element and can penetrate tissue via the NRP1 and NRP2 dependent CendR internalization pathway. The studies by Roth L have shown that the tLyP-1 phage binds to immobilized NRP1 (120-fold) and NRP2 (8-fold) with markedly greater affinity than a control phage [35]. In addition, tLyP-1-conjugated nanoparticles selectively home to tumor tissue, penetrate the blood vessels, and enter the tumor parenchyma with greater potency than the parent peptide LyP-1 [35][36][37][38]. Given that NRPs are overexpressed in glioma tumor tissue and that tLyP-1 can enter tumors, we hypothesized that tLyP-1 might be a potential imaging ligand for glioma. To address this question, we conjugated tLyP-1 to the fluorescein of FAM and positron emitter, 18 F-fluoride, respectively for use as a molecular imaging probe.

Tumor cell lines
A human glioma cell line U87MG that overexpresses NRP [31] was purchased from the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Hyclone, America) supplemented with 10% fetal calf serum (Hyclone, America) at 37°C in a humidified 5% carbon dioxide-containing atmosphere.

Animal model
Animal experiments were conducted under a protocol approved by the Nanfang hospital animal ethics committee at the Southern Medical University (Application No.: NFYY-2012-115).
Male and female BALB/C athymic nude mice (nude mice) 4-6 wk of age were obtained from the Laboratory Animal Center at the Southern Medical University. Glioma xenografts (U87MG cells) were inoculated into the mice by injecting 1 × 10 6 cells intramuscularly into the left flank. Tumor xenografts were monitored until the largest tumor diameter was approximately 0.5-1 cm, which took 4-5 weeks.
In vitro binding and blocking assay using FAM-tLyP-1 in U87MG cells The ability of FAM-tLyP-1 to bind U87MG cells in vitro was assessed via fluorescence microscopy. U87MG cells were seeded on cover slips in 6-well plates and incubated in DMEM/F-12 (0.5 mL/well) overnight. The following day, the U87MG cells were incubated at 37°C for 1 h with different concentrations of FAM-tLyP-1 in PBS + 1% BSA (0, 1, 5, 10, 20, and 40μM). To assess whether FAM-tLyP-1 binding could be blocked by unlabeled tLyP-1 we incubated U87MG cells with FAM-tLyP-1 (1 or 4μM) in the presence or absence of tLyP-1 (20μM) for 37°C for 1 h. The cells were then washed three times with PBS. The slides were imaged by a blinded observer under blue light using a fluorescent inversion Olympus IX71 microscope (Olympus, America).
When imaging using the fluorescent microscope, we first used white light to confirm that tumor cells were in the field of view. Then, in the same field of view, we used the blue light to visualize the green light emitted from the cells and take the photos. The comparisons of FAM-tLyP-1 uptake by tumor cells in the in vitro binding and blocking experiments were strictly qualitative.
The distribution of FAM-tLyP-1 in tumor tissue U87MG tumor-bearing nude mice (U87MG tumor model) were intravenously injected with 150 μL of FAM-tLyP-1 (1mM). The mice were sacrificed after 1 h by cervical dislocation under ketamine/xylazine anesthesia. The tumors were removed, embedded in freezing liquid (McCormick, France), and sectioned with a CM1850 UVLeica freezing microtome (Leica, Germany) into 7 μm thick sections. The sections were mounted in DAPI-containing mounting media. NRP1 was stained using an anti-NRP1 antibody (1:1000) and a Dylight 647-conjugated secondary antibody. The tissue distribution of NRP1 and FAM-tLyP-1 peptides in the same tumor section was determined by a blinded observer using an Olympus DP71 fluorescent microscope (Olympus, America). Red fluorescence indicated NRP positivity and green fluorescence indicated positive staining with FAM-tLyP-1.

Distribution of FAM-tLyP-1 by fluorescence techniques in the U87MG tumor models
Mice with U87MG tumor xenografts (n = 10) were randomly divided into an experimental group (n = 5) and control group (n = 5). The mice in the experimental group were intravenously injected with 150 μL of FAM-tLyP-1 (1mM), while the control group received 150 μL of FAM-labeled control peptide (1mM). The mice were sacrificed after 1 h by cervical dislocation under ketamine/xylazine anesthesia. The tumor and normal organs were then removed, washed with PBS at least three times, and then collected for imaging ex vivo. FAM-tLyP-1 uptake in the tumor and normal organs was described by a blinded observer under blue light with an exposure time of 60 s, using the Kodak in-Vivo Imaging System F (Kodak, American), and processed for fluorescence intensity analysis. The fluorescence intensity was analyzed visually based on the brightness of the green light indicating uptake of FAM-tLyP-1.
For whole animal imaging, U87MG tumor bearing mice were administered 150 μL of FAM-tLyP-1 (1mM) and then sacrificed and frozen at -80°C after 1.0 h. Each mouse was then cut into coronal sections. The uptake of FAM-tLyP-1 in the tumor and normal organs was observed under blue light using the Kodak in-Vivo Imaging System F (Kodak, American). To determine whether unlabeled tLyP-1 could block FAM-tLyP-1 labeling in the tumor, the mice were injected intravenously with 10 times the quantity of non-conjugated tLyP-1 peptide 30 min prior to FAM-tLyP-1 administration.
The intensity of FAM-tLyP-1 uptake in the tumor and normal organs was analyzed blindly using the Kodak MI analysis software (Kodak, American). The fluorescence intensity was analyzed visually from high to low levels based on a color scale of white, red, yellow, green, blue, and black. Regions of interest (ROIs) were drawn around the border of the tumor and normal organs on the fluorescence images and the average fluorescence intensity was measured. The tumor/non-tumor ratios (T/NT ratios) were calculated by dividing the fluorescence intensity in the tumor by that of the normal organs.
Synthesis of PET molecular probe, 18 F-tLyP-1 As previously described [32][33][34], [Lys4] tLyP-1 was labeled with 18 F by coupling the Lys 4 amino group with-succinimidyl-4-18 F-fluorobenzoate ( 18 F-SFB) under slightly basic condition (pH 8.5). 18 F-SFB was synthesized, as previously reported and purified by HPLC [39]. The purified 18 F-SFB was trapped on the cartridge and eluted with diethyl ether (5 mL) into a 10 mL V-vial in a customized module. The diethyl ether (Et 2 O) was removed under helium stream at ambient temperature and the dried labeling agent was reconstituted with the peptide in anhydrous DMSO. The conjugation between 18 F-SFB and peptide was executed using 250 μg tLyP-1 peptide with DIPEA (40 μL) at 40°C for 20 min and was quenched by adding 5% acetic acid. The crude product was purified using semi-preparative HPLC. The 18 F-tLyP-1 fraction was collected. The solvent was removed by rotary evaporation and the residue was resolubilized in saline. The formulated saline mixture was sterile-filtered into a sterile product vial.

MicroPET/CT imaging analysis
MicroPET/CT scan was performed on a SIEMENS Inveon scanner (Siemens, Germany). U87MG tumor bearing mice (n = 5) were intravenously injected with 3.7 MBq (200 μCi) of 18 F-tLyP-1. MicroPET/CT images were acquired as 10-min static images at 30, 60, and 120 min after the injection with the mice under isoflurane anesthesia. The images were reconstructed by a 3-dimensional ordered subsets expectation maximum (OSEM) algorithm and CT correction was applied for attenuation correction.
In the PET images, ROIs were measured with the Inveon Research Workplace (IRW) 3.0 software (Siemens, Germany). The ROI was determined by manually superimposing the ellipsoid volume of interest (VOI) to the target tissue. The activity concentrations were determined by the mean pixel intensity within each VOI, and converted to μCi/mL using a calibration constant. Assuming the tissue density of 1 g/mL, the ROI activity was converted to μCi/g and normalized as percent injected dose per gram (%ID/g). The tumor/normal brain ratios were calculated by dividing the ROI activity in tumor by that in the normal brain [40].

Statistical analysis
Descriptive data were expressed as the mean ± standard deviation. Statistical Package for the Social Sciences, version 13.0 (SPSS Inc.), was used for the statistical analysis. The nonparametric one-sample kolmogorov-smirnov test was applied to assess for normality. A P value greater than 0.05 indicated the data was normally distributed. The T/NT ratios of FAM-tLyP-1 and FAM-control peptide were compared using the independent samples t test. One-way ANOVA was used to copare the uptake of 18 F-tLyP-1 in the tumor and normal organs at 60 and 120 min post injection. A P value less than 0.05 was considered statistically significant.

FMA-tLyP-1 is strongly taken up by U87MG cells
We first tested the ability of FAM-tLyP-1 to bind to U87MG cells in vitro. FAM-tLyP-1 was strongly taken up by the U87MG cells at concentrations as low as 1μM. The uptake of FAM-tLyP-1 in U87MG cells increased slightly in a dose dependent manner (Fig 1). There was no apparent green fluorescence in cells treated with PBS. We then assessed whether unlabeled tLyP-1 could prevent the binding of FAM-tLyP-1. FAM-tLyP-1 uptake in U87MG cells was markedly reduced when the U87MG cells were first incubated with a 5-or 20-fold excess of non-conjugated tLyP-1 peptide. This suggested that tLyP-1 was likely competitively inhibiting FAM-tLyP-1 binding (Fig 2).

FAM-tLyP-1 accumulates in tumor tissues that express NRP1
We next assessed whether FAM-tLy-P-1 accumulated in tumors in vivo and whether its distribution was consistent with NRP binding in vivo. NRP1 expression was confirmed on the U87MG tumor tissue used in the present study (Fig 3C). In addition, FAM-tLyP-1 was shown to accumulate in the tumor tissue 1.0 h after administration by fluorescence microscopy ( Fig  3B). The distribution of FAM-tLyP-1 was consistent with NRP1 expression in the same tumor section (Fig 3).
FAM-tLyP-1 targets to the glioma tumor in vivo with a high tumor/brain ratio To assess whether FAM-tLyP-1 had a sufficient signal to background ratio in vivo, we assessed the intensity of the green fluorescence in tumor tissue and normal tissue from individual organs in nude mice with U87MG tumors. One (1) hour after administration of the FAM-tLyP-1 the fluorescence intensity of FAM-tLyP-1 in the tumor was very high, compared to minimal fluorescence in the normal brain tissue (Fig 4A). The fluorescence intensity in the tumor was significantly greater than the normal brain with a tumor/brain ratio of 3.44 ± 0.83 ( Table 1). The T/NT fluorescence ratio of the FAM-control peptide was 1.32 ± 0.15 in the brain after 1 h (Fig 4B), and was significantly lower than FAM-tLyP-1 (t = -5.547, P = 0.001; Table 1).
The fluorescence intensity in other organs, such as the heart, lungs, spleen, and muscle was also minimal and the mean T/NT ratios in these organs also exceeded 2.0 ( Table 1). The T/NT ratio was below 1.0 in two tissues of the intestine (0.88±0.74) and kidney (0.84±0.35; Fig 4A; Table 1).
We also conducted whole animal imaging using coronally sectioned U87MG tumor bearing mice. These images also indicated high fluorescence intensity in the tumor, and minimal fluorescence in the brain, heart, lungs, liver, muscle, and bone. As with the isolated organs, intense fluorescence distribution was also noted in the intestine (Fig 5A). FAM-tLyP-1 uptake by tumors in vivo could be blocked by administering an excess of non-conjugated tLyP-1 peptide (Fig 5B). Overall, the results of the isolated organs and sectioned whole animal imaging were consistent, and indicated high levels of uptake by tumor, but not normal tissue. Having established that the fluorescent marker FAM-tLyP-1 was an effective imaging agent in vivo, we synthesized 18 F-tLyP-1 for use in PET/CT imaging and assessed its in vivo imaging potential. The amount of time required for 18 F-tLyP-1 synthesis was approximately 250 min, including the final HPLC purification. The overall radiochemical yield (without decay correction) was approximately 8-12%. The radiochemical purity of the labeled peptides was greater than 98% measured by analytical HPLC. The 18 F-tLyP-1 molecular marker was then tested in the U87MG tumor model. The maximum-intensity images (MIP) of the in vivo microPET/CT imaging from 30, 60, and 120 min after 18 F-tLyP-1 injection are presented in Fig 6. There was high radioactivity in the tumor at 30, 60, and 120 min post administration. The radioactivities of the glioma tumors were 2.97 ± 0.40% ID/g at 60 min and 2.22 ± 0.27% ID/g at 120 min (Fig 7). Minimal radioactivity was found in the brain during this time period (Fig 6). The radioactivity in the brain was 1.16 ± 0.20% ID/g at 60 min and 0.72 ± 0.10% ID/g at 120 min (Fig 7). The tumor/brain ratios of 18 F-tLyP-1 reached 2.69 ± 0.52 and 3.11±0.25, respectively at 1 h and 2 h post-injection and the tumor was visualized clearly (Fig 6B and 6C).  The radioactivity in the blood pool, heart, and liver was high at 30 min, but had decreased to very low levels by 60 min (Fig 6). Intense radioactivity was found in the gallbladder, intestine, kidney, and bladder from 30-120 min (Fig 6). The radioactivity levels observed in tumor were significantly higher than in the brain, lung, heart, liver, spleen, and muscle at 60 and 120 min (Fig 7; all P<0.001). However, the radioactivity in tumor was significantly lower than observed in the intestine at 60 and 120 min (Fig 7; all P<0.001). In the kidney, the level of radioactivity compared to the tumor was similar at 60 min (P = 0.691) and lower at 120 min (Fig 7; P<0.001). The biodistribution of 18 F-tLyP-1 was similar to FAM-tLyP-1.

Discussion
Here we show that the tLyP-1 peptide can be successfully conjugated to fluorescent and radio label markers to selectively target glioma tumors both in vitro and in vivo. The FAM-tLyP-1 reagent binds to U87MG cells at low concentrations (1μM) in vitro. In vivo, FAM-tLyP-1 accumulates in NRP-positive U87MG tumors and mirrors the distribution of NRP1. These findings suggest that FAM-tLyP-1 might be an alternative to CTX:Cy5.5. Compared with CTX, tLyP-1 is a much smaller (only 7 amino acid residues) and has a less complicated molecule structure, which will likely make it is easier to synthesize, purify, and translate to the clinic [35].   We also demonstrated that the radiolabeled 18 F-tLyP-1 is a viable candidate for an imaging reagent. 18 F-FDG is a commonly used PET radiopharmaceutical to detect malignancies [41][42][43][44]. However, it plays limited role in detection of malignancies in the brain because of high uptake of 18 F-FDG in normal brain tissue [45][46][47]. Here we show that uptake of FAM-tLyP-1 and 18 F-tLyP-1 in the brain was minimal. This resulted in high T/brain ratio in the brain and clear tumor visualization. Similar to 18 F-tLyP-1, the tumor/brain ratios of FAM-tLyP-1 was also high, indicating both molecular reagents could likely by used to identify glioma microinvasions before the surgery and intraoperation to guide the resection of malignant glioma.
Given that NRPs are overexpressed in angiogenic vessels in 98-100% of carcinomas and in most malignant tumor cells [22][23][24][25][26][27][28][29], FAM and 18 F-floride labeled tLyP-1 might have potential as broad spectrum tracers to detect other tumors. We observed relatively low levels of FAM-tLyP-1 and 18 F-tLyP-1 uptake in the head and neck, lungs, heart, liver, spleen, bone, and muscle suggesting these reagents may be useful for detecting tumors in these regions. Unfortunately, there was high accumulation of FAM-tLyP-1 and 18 F-tLyP-1 in the gallbladder, intestine, kidneys, and urinary bladder that would make it difficult to visualize tumors in the abdomen. This also suggested that FAM-tLyP-1 was excreted via the urinary system and hepato-biliary tract. Additional modifications of tLyP-1 are warranted to reduce the biliary excretion and improve in vivo pharmacokinetics.

Conclusion
This study demonstrated that FAM and 18 F-labeled tLyP-1 peptide can selectively target gliomas and provide a high tumor-to-background ratio 1 h post administration. Both FAM-tLyP-1 and 18 F-tLyP-1 are potential molecular tracers to characterize NRP receptor expression, help visualize tumors, and improve the diagnosis of glioma.