Discovery of a Novel Aminopeptidase N Inhibitor Dino-Leucine Borate

Aminopeptidase N(APN/CD13), a Zn 2+ -dependent ectopeptidase localized on the cell surface, is widely considered to influence the invasion of tumor cells. We found that boroleucine and dino-leucine borate exhibited a strong inhibitory effect on the enzyme activity of aminopeptidase N. The tested assay indicated that both compounds had an anti-proliferative effect on triple-negative breast cancer cells. Wound healing assay, migration test and matrigel-coated transwell assay showed that both boroleucine and dino-leucine borate inhibited the migration and invasion of breast cancer cells. Immunoblot analysis showed that both compounds down-regulated the expression of matrix me-talloproteinase-2/9. In the capillary tube formation assay of human umbilical vein endothelial cells (HUVECs), dino-leucine borate showed better antian-giogenic activity than ubenimex even at a low concentration (10 μM). More-over, compared with ubenimex, the anti-metastatic activity of dino-leucine borate in vivo was similar to or even better than that of ubenimex in the H22 pulmonary metastasis mouse model. In this paper, we found the novel APN inhibitors to markedly suppress the enzyme activity of APN and inhibit the migration and invasion of tumor cells in vitro and in vivo.

ectopeptidase that is widely expressed in mammalian cells, has been associated with different aspects of normal (e.g. myeloid progenitor cells) as well as malignant development [1]. In addition, being a member of the M1 zinc metallopeptidase family, APN plays a crucial role in a variety of functions such as migration, invasion, angiogenesis, and metastasis of tumor cells [2]. Moreover, CD13 has been described as a functional biomarker of angiogenesis and surface marker of semi-dormant liver cancer stem cells (CSCs) in human liver cancer cells [3] [4] [5]. Scientists from Osaka University [6] found that CD13 is also a potential therapeutic target for liver cancer. Compared with the CD13 low group, the CD13 high group has earlier recurrence and shorter survival time, and CD13 enrichment is associated with early recurrence and poor prognosis in hepatocellular carcinoma (HCC) patients.
Aminopeptidase N is often overexpressed in tumor cells and is associated with angiogenesis and cancer progression. Guzmrojas et al. [5] compared the effects of APN deficiency in allografted malignant and non-malignant tumor cells on tumor growth and metastasis in mice, and found that APN activity in tumor and host cells jointly promoted tumor angiogenesis and growth in two independent tumor transplantation models. Therefore, it is important to find new CD13 inhibitors. At present, the only medicine as an APN inhibitor on the market is ubenimex, N-[(2S, 3R)-3-amino-2-hydroxy-4-phenylbutanoyl]-L-leucine, also named Bestatin. Ubenimex was first marketed in Japan in 1987, and is often used as an adjuvant for immune enhancers in the treatment of leukemia or other anti-cancer drugs [7]. In recent years, studies have found that ubenimex has a good application prospect in the treatment of solid tumors, which can inhibit gastric cancer, non-small cell lung cancer, and colon cancer to a certain extent and prolong the survival of patients [8] [9] [10] [11]. Boroleucine((S)-1-amino-3-methyl-butylborate) and dino-leucine borate ((S)-1amino-3-methyl-butylborate pinnae diol ester) are α-aminoborate derivatives, which are effective aminopeptidase inhibitors to suppress the enzyme activity of enkephalinase (EC 3.4.24.11) [12] and are often used as drug adjuvants. The inhibitory effect of ubenimex on leucine enkephalin metabolism was much weaker than that of boroleucine [13] [14]. According to the chemical structure of boroleucine and dino-leucine borate, we think they also can suppress the enzyme activity of APN (EC 3.4.11.2)? Therefore, these compounds have attracted our attention as APN inhibitors. Breast cancer is the most common cancer and the leading cause of cancer death among females worldwide [15] [16]. Breast cancer treatment involves a multidisciplinary approach, such as traditional drug therapy, surgery, and radiation therapy. Effective breast cancer treatment requires maximum therapeutic effect with minimal adverse effects to ensure a good quality of life for patients [17] [18]. The common treatment is to give chemical medicine to patients, which can kill tumor cells and significantly improve the survival rate of patients. Therefore, it is important to find new effective APN inhibitors with small side effects. In this study, we found a new lead compound as an APN inhibitor that has good anti-tumor activity. The results showed that boroleucine and dino-leucine borate markedly suppressed the enzyme activity of APN and migration, invasion, and metastasis of tumor cells in vitro and in vivo. Therefore, the compound can be used as a lead compound for new drug development.

Cell Culture
Human K562 myeloid lymphoblastoma cell, human breast cancer cells MDA-MB-231 and MDA-MB-468 were obtained from the Cell Bank of Shanghai (Shanghai, China) and maintained in Roswell Park Memorial Institute (RPMI)-1640 supplemented with 10% fetal calf serum. The cells were incubated at 37˚C in a humidified atmosphere containing 5% CO 2 .

Molecular Dynamic Simulation
The molecular docking process was performed by Glide (schrodinger Inc, supported by Shanghai Institute of Materia Medica Chinese Academy of Sciences). The crystal structure of APN (PDB Entry: 2DQM) was chosen as the receptor in the docking study. Structural modifications were performed to make the protein suitable for docking. Water molecules and ligands crystallized in the protein structure were removed. OPLS 2005 force field was assigned to the refined structure. The structure of molecule 1 and 2 were sketched by maestro and prepared by LigPrep. The active site was defined as a cubic box containing residues at a distance of 20 Å aroundthe ligand ubenimex.

Enzyme Activity Screening Assay
The activity of APN enzyme was detected by spectrophotometry using L-leucinep-nitroanilide as an APN substrate. Monoclonal K562-APN cells (K562 cells with APN overexpression) were disrupted with ultrasonication in phosphate buf-fer saline (PBS). After centrifugation (1000 g, 10 min), cell precipitate was discarded, and the supernatant was seeded into 96-well plateas enzyme source. Then different compounds were added to wells for 10 min. L-leucine-p-nitroanilide substrate was added into each well (1.2 mM). After incubation for 30 min, the enzyme activity was estimated by measuring the absorbance at 405 nm. The inhibition rate of compound was calculated by the following formula (OD control − OD test )/OD control × 100% [8].

K i Value Determination
30 μL phosphate buffer solution was added to the 96-well plate, followed by 10 μL compound and enzyme solution with different concentrations. After 5 min of incubation, 50 μL substrate and buffer solution was added and incubated at 37˚C. Shake the plate quickly and thoroughly. The enzymatic reaction curve was measured at 405 nm for 15 min, and the absorbance value was measured every 1min. The inhibition type was determined by Lineweaver-Burk mapping. The inhibition constant was determined by a second plot of Km/Vapp m and the inhibitor concentration.

EdU Assay
The MDA-MB-231 and MDA-MB-468 cells were seeded at a density of 5 × 10 3 cells/well in 96-well plates. The EdU kit (cat. C0088L; Beyotime Institute of Biotechnology, Shanghai, China) was used to detect cell proliferation according to the manufacturer's instruction. The cells were treated with drugs for 48 h or 72 h, then washed with PBS, and incubated with 10 μM EdU in culture medium at room temperature for 2 h. The resulting absorbance was measured at 450 nm on a spectrophotometer. Each treatment was performed in triplicate wells per experiment.

Colony-Formation Assay
3000 cells/2ml/well MDA-MB-231 or MDA-MB-468 cells were put into six-well plate. After 12 h of seeding, cells were treated with drugs. Until each colony had more than 50 cells, then the experiment terminated. Washed twice with PBS and fixed with methanol-for 10 min. Then cells were stained with 1% crystal violet for 5 min. Finally, colonies were washed with PBS and the photographs were captured under a microscope (IX81, Olympus).

Migration Assay
Breast cancer cells, treated with different compounds (20 μM), were plated in 6-well plates and incubated at 37˚C, 5% CO 2 for 2 d. Then a line was drawn by the tip of 10 μL pipette and washed by PBS. The images of migrating cells were taken at 0 h and 24 h respectively. The experiment was repeated three times.

Invasion Assay
Breast cancer cell invasion assay was performed using transwell chamber Y. Q. Ma et al. (transwell, 8 μm pore size; Corning Inc, Corning, New York, USA). The matrigel (Cat.356234; BD Biosciences, California, USA) was diluted with serumfree RPMI-1640 at 1:10, and 50 μl solution was added to the upper membrane for 2 h. Then 500 μl medium containing 1% FBS was added to the lower chamber. The cells (1 × 10 5 cells/well), pretreated with different compounds (20 μM) for 2 d, were plated into the upper chamber and incubated at 37˚C for 24 h. Then the matrigel on the upper membrane was wiped. The invaded cells were fixed with methanol and stained with 5% crystal violet dye. The pictures were taken with a microscope (IX81; Olympus, Japan). The experiment was repeated three times.

HUVEC Tubular Structure Formation Assay
M199 medium (50 μL) and matrigel (50 μL) were added into 96-well plates for 0.5 h at 37˚C. Add 100 μL compound (10 μM) and 50 μL cell suspension (20,000 cell/50μL/well) and incubated at 37˚C for 20 h. Images of five random fields per well were analyzed by Motic Image Plus 2.0 software (Motic Instruments Inc., Canada). The numbers of branch points of the formed tubes were counted and the average numbers were calculated. The experiments were repeated three times.

In Vivo Mouse H22 Pulmonary Metastasis Model
Four-week-old male Kunming mice were purchased from Hunan SJA Laboratory Animal Company (Hunan, China) and fed under specific pathogen-free con-

Statistical Analyses
All statistical analyses were performed using SPSS 17.0 statistical software package. Quantitative data was presented as mean ± SEM. Statistical analysis was performed by one-way analysis of variance. For data of migration and invasion assay, significant difference between groups was performed by Student's t-test. P < 0.05 was considered statistically significant in all experiments.

Boroleucine and Dino-Leucine Borate Effectively Suppress APN Enzyme Activity
According to the chemical structure of boroleucine and dino-leucine borate, two compounds may suppress the enzyme activity of APN. Therefore, we employed a method to detect the effect of two compounds on suppressing homosapien APN enzyme activity, as we previously reported [19]. The enzyme suppression IC 50 of ubenimex, boroleucine and dino-leucine borate was 20.17 μM, 0.15 μM, 0.007 μM, respectively (Figure 1(A), Figure 1(B)). This data suggested that boroleucine and dino-leucine borate have better suppression effect on APN enzyme activity than ubenimex. The IC 50 value may vary between different laboratorys, K i is a parameter to a compound. Then we detected K i of three compounds. Using Lineweaver-Burk double reciprocal plots, a series of straight lines intersected in the second quadrant indicated that three compounds belong to competitive inhibition type. K i value of ubenimex, boroleucine, and dino-leucine borate were 18.78 μM, 1.73 μM, 0.77 μM, respectively (Figure 1(C)). This result also indicated that boroleucine and dino-leucine borate strongly combine with APN enzyme.
According to the chemical structure of dino-leucine borate, it is unstable and may degrade to boroleucine. In order to verify this hypothesis, we placed boroleucine and dino-leucine for several days to measure the inhibition rate of APN enzyme activity. The inhibition rate of boroleucine was only 4.3% on the first day and decreased slowly over time. However, the inhibition rate of dino-leucine borate was 68% on the first day and decreased rapidly over time to 0.27%, which was as same as that of boroleucine (Figure 1(D)). Therefore, we think that the degradation product of dino-leucine borate is boroleucine, and dino-leucine borate can convert into boroleucine over time.

Boroleucine and Dino-Leucine Borate Inhibit the Proliferation of MDA-MB-231 and MDA-MB-468 Cells
To test the effect of compound on cell proliferation, we performed EdU assay. The data suggested that ubenimex, boroleucine, and dino-leucine borate had little effect on cell proliferation in MDA-MB-231 and MDA-MB-468 cells when concentration was 1 mM. (Figures 2(A)-(D)) According to other data (Data not shown), we chose 20 μM to do the following test. At this concentration, three compounds rarely inhibited the proliferation of tumor cells. Therefore, the concentration of 20 μM was selected to treat cells in wound healing, migration and trans-well invasion assay. Colony-formation assay is an effective method to determine the proliferation ability of single cell in vitro. Therefore, we evaluated the effect of boroleucine, dino-leucine borate, and ubenimex on the colony formation ability in triple negative breast cancer cells. We found that all compounds inhibited colony formation, and boroleucine had the same inhibitory effect as ubenimex. However, dino-leucine borate had a stronger inhibitory effect than ubenimex on colony formation (Figure 2(E), Figure 2(F)).

Boroleucine and Dino-Leucine Borate Inhibit the Migration and Invasion of Breast Cancer Cells and Downregulate the Expression of Metastasis-Related Proteins
The effects of boroleucine and dino-leucine borate on migration and invasion of   Figure 3(B)). We also employed transwell migration assay without matrigel to test cell migration. The results showed that the number of trans-membrane cells in the three compound groups was significantly reduced compared with Ctrl group. While the inhibitory effect of boroleucine and dino-leucine borate was better than that of ubenimex. The inhibitory effect of boroleucine and dino-leucine borate on cell migration was also verified in MDA-MB-468 cells at the same condition (Figure 3(C), Figure 3(D)). Then, human umbilical vein endothelial cells (HUVECs) tubular structure formation assay was used to evaluate the anti-angiogenesis activity. The results showed that dino-leucine borate significantly reduced the capillary tube formation compared with ubenimex (Figure 3(E)).  Figure 4(B)), while the effect of three compounds on the expression of MMP-2/9 were detected. WB result showed that all three compounds down-regulated the expression of MMP-2/9 at 20 μM. Moreover, the downregulation effect of dino-leucine borate was more potent than that of boroleucine and ubenimex (Figure 4(C), Figure 4(D)). These results together suggested boroleucine and dino-leucine borate inhibit migration and invasion of tumor cells.

Ubenimex, Boroleucine, and Dino-Leucine Borate Inhibit Lung Metastasis in Kunming Mice Bearing Hepatocellular Carcinoma H22 Cells
Based on the significant role of APN in tumor metastasis, we used in vivo H22 pulmonary metastasis model to evaluate the anti-metastasis activity of three compounds. The number of metastatic pulmonary node was counted. As depicted in Figure 5, the number of node in mice treated with dino-leucine borate (20 mg/kg, ip) was lower than that of ubenimex group (20 mg/kg, ip) ( Figure 5(A)).
The pulmonary node number of dino-leucine borate and ubenimex were 16 and 28, respectively. There were no apparent loss of body weight and no evident toxic sign in liver and spleen in mice treated with boroleucine and dino-leucine borate ( Figure 5(B), Figure 5(D)). All together, dino-leucine borate manifests strong anti-metastasis activity in vivo, no apparent toxicity.

Molecular Docking
Molecular docking was performed to predict the binding of boroleucine and dino-leucine borate in the active site of APN. As revealed in Figure 6,

Discussion
Breast cancer is a common malignancy in women, and its prevalence and incidence have increased significantly in the past decades. At present, chemotherapy is still the main method of breast cancer treatment. Meanwhile, conventional chemotherapy and radiotherapy are insensitive to breast cancer [20] [21].
In this study, we found that boroleucine and dino-leucine borates are novel small molecule aminopeptidase N inhibitors. These inhibitors interfere with the  are known to be closely related to the invasion and metastasis of cancer cells [22] [23]. In our study, we found that all three compounds significantly suppressed the expression of MMP2/9, and the down-regulation effect of dino-leucine borate was more potent than that of boroleucine and ubenimex. Meanwhile, dino-leucine borate demonstrated more potential anti-angiogenesis activity than ubenimex in the tubular structure formation assay of HUVECs. In addition, dino-leucine borate exhibited a promising anti-metastasis effect in the H22 pulmonary metastasis mouse model, which deserves further research and development as an anti-metastatic lead compound.
A large amount of evidence suggests that APN can be used as a promising marker for cancer diagnosis since APN is up-regulated in tumor cells and plays a vital role in tumor invasion, metastasis, and angiogenesis [24] [25] [26] [27] [28]. Therefore, a sensitive and selective method for in situ detection of APN expression is of great significance for the diagnosis and pathophysiological elucidation of APN-related disease. The results showed that both boroleucine and dino-leucine borate significantly inhibited the activity of APN. These compounds can be used as the forerunners of APN inhibitors in the development of new drugs, and can also be used to synthesize APN targeted fluorescent probes for the early detection of tumors.
APN-targeted fluorescent probe is used for the early detection of tumors. Li H [29]'s research group developed a novel two-photon near-infrared fluorescence probe DCM-APN, which can be used to trace APN in vitro and in vivo. The probe DCM-APN is used to distinguish normal cells from cancer cells, indicating that probe DCM-APN is an effective tool for detecting APN, and is helpful for the early diagnosis of tumours in clinical medicine. Hahnenkamp A [30] synthesized a non-peptide small molecule fluorescent imaging agent with a high affinity for APN. It plays a key role in pathophysiological angiogenesis. The low K i values of boroleucine and dino-leucine borate also indicated that they combine strongly with APN and can be used to develop fluorescence probes for early detection of tumors through targeting APN.
This study has several limitations. For example, we only discussed the activity of two compounds and did not design experiments to prove their role in vivo imaging in mice or their toxic effects on normal cells. Our current research is still in the preliminary stage of searching for potent compounds. In future experiments, we will further study the anti-tumor mechanism of APN inhibitors.
Above all, the two new structures of boroleucine and dino-leucine borate are expected to be used as lead compounds for the design and synthesis of new APN Y. Q. Ma et al. inhibitor, and targeting APN for early tumor detection through developing new fluorescence probe.
Boroleucine and dino-leucine borate significantly inhibit the enzymatic activity of APN and inhibit the proliferation, migration and invasion of triple-negative breast cancer cells as well as lung metastasis in vivo, and can therefore be used as aminopeptidase N inhibitor class anti-tumour lead compounds for new drug development.

Author Contributions
Conception and design: Yuqian Ma, Minzhi Fang, Xuejian Wang; Experiments and analyzed the data: Yuqian Ma, Yutian Li, and Dongqi Zhu; Manuscript writing: Yuqian Ma, Minzhi Fang, Xuejian Wang; Final approval of manuscript: All authors; Accountable for all aspects of the work: All authors.