Design, synthesis and preliminary activity evaluation of novel 3-amino-2-hydroxyl-3-phenylpropanoic acid derivatives as aminopeptidase N/CD13 inhibitors

Abstract

Aminopeptidase N (APN/CD13) over expressed on tumour cells, plays a critical role in tumour invasion, metastasis and tumour angiogenesis. In this article, we described the design, synthesis and preliminary activity studies of novel 3-amino-2-hydroxyl-3-phenylpropanoic acid derivatives as APN inhibitors. The in vitro enzymatic inhibitions on APN from porcine kidney showed that compound 7e had the most potent inhibitory activity against APN with the IC₅₀ value to 1.26 ± 0.01 μM, which is better than that of bestatin (IC₅₀ = 2.55 ± 0.11 μM). In addition, compound 7e also showed better inhibitory activity against APN on human ovary clear cell carcinoma cell ES-2 than bestatin with the IC₅₀ value to 30.19 ± 1.02 μM versus 60.61 ± 0.1 μM. Compound 7e could be used as the lead compound in the future for anti-cancer agent research.

Keywords::

• Aminopeptidase N inhibitors
• 3-amino-2-hydroxyl-3-phenylpropanoic acids derivatives
• structure-based drug design

Introduction

Aminopeptidase N (APN; EC 3.4.11.2) is a zinc-dependent type II membrane-bond ectopeptidase and belongs to the M1 family of MA clan of peptidase^1,2. This enzyme consists of 967 amino acids with a short N-terminal cytoplasmic domain, a single transmembrane part and a large cellular ectodomain containing the active site^3,4, which can preferentially release neutral or basic amino acids from the N-terminal of peptides^5,6. APN is widely distributed in the body of mammalian. It can be expressed on the surface of different cells such as myeloid progenitors and monocytes, epithelial cells of the intestine and kidney, synaptic membranes in the central nervous system, fibroblasts, endothelial cells, epithelial cells^7–10 and so on. APN is highly expressed on tumour cells and plays a critical role in tumour invasion, metastasis and angiogenesis. Therefore, the design and synthesis of APN inhibitors may be a clinical significance for the discovery of anti-cancer agents.

Bestatin (AHPA-Leu), a well-known APN inhibitor, was first isolated from the culture filtrate of Streptomyces olivoreticuli in 1976¹¹, and now it is widely used in clinical practise for the treatment of adult acute non-lymphocytic leukemia. Since then, many APN inhibitors such as probestin¹², prebestin¹², AHPA-Val¹³ and amatatin¹⁴ have been developed. Judging from the structure of these compounds, we can see that they all contain the key unit 3-amino-2-hydroxy acyl scaffold, which is important for their biological activity for that it can chelate the Zn²⁺ in the active site of metal-dependent enzymes¹⁵.

The interaction of bestatin with the active site of APN showed that the hydroxyl group and carbonyl group of bestatin belong to the zinc-binding group (ZBG) to chelate the Zn²⁺ in the active site of APN. The phenyl group can insert into the S1 pocket of APN and the Leu residue can insert into the S1′ pocket, while the free amino group can interact with Glu 350. In our previous work, we have designed a series of novel chloramphenicol amine derivatives as APN inhibitors, but the result was not satisfied, most compounds showed less activity than bestatin¹⁶, the most active compound 13b is a pseudotripeptide with the IC₅₀ value to 7.1 μM, which is still less than that of bestatin (IC₅₀ = 3.00 μM). Recently, our group has synthesised a series of novel L-isoserine derivatives as APN inhibitors, most of these compounds showed moderate activities, the most active compound 14b (IC₅₀ = 12.2 μM) showed still less inhibitory activity toward APN than that of bestatin (IC₅₀ = 7.3 μM)^17,18.

Taxol, a well-known anti-cancer drug, also has the 3-amino-2-hydroxy acyl scaffold ((2R, 3S)-3-amino-2-hydroxy-3-phenylpropanoic acid residue). Compared with L-isoserine, (2R, 3S)-3-amino-2-hydroxy-3-phenylpropanoic acid is much more similar with the residue of bestatin (2S, 3R)-3-amino-2-hydroxy-4-phenylbutyl acid except for the chirality and one carbon. There is a phenyl group in C₃ of (2R, 3S)-3-amino-2-hydroxy-3-phenylpropanoic acid, which may be beneficial for APN interaction. So, we used bestatin and 14b as the lead compound, (2R, 3S)-3-amino-2-hydroxy-3-phenylpropanoic acid as the key scaffold, by coupling with different amino acids or other functional groups, and we got a series of novel bi-peptide and tri-peptide derivatives with 3-amino-2-hydroxy-3-phenylpropanoic acid scaffold. Considering the best ZBG hydroxamate group may be good for APN inhibitors, we also converted the carboxyl group of some compounds into hydroxamate group (Figure 1).

Figure 1.  Novel aminopeptidase N inhibitors with 3-amino-2-hydroxy-3-phenylpropanoic acid scaffold.

Results and discussion

Chemistry

The target compounds were synthesised via the route outlined in Schemes 1 and 2. The commercial (2R, 3S)-3-amino-2-hydroxyl-3-phenylpropanoic acid hydrochloride 1 was protected by (Boc)₂O to get the key intermediate compound 2. Compound 2 can be coupled with different methyl ester of amino acids or methyl ester of dipeptides by classical EDCI/HOBt method to provide compounds 3a–3o. The esters of compounds 3a–3o were hydrolysed to carboxylic acid by NaOH/H₂O in methanol, and then converted into the target compounds 5a–5o by deprotecting the Boc group. Compounds 7a–7h containing hydroxymate group were obtained from compounds 3a–3h by reacting with NH₂OK in methanol and then cleaving the Boc group.

As shown in Scheme 2, compounds 9a–9d were obtained from compound 4a by coupling it with different substituted organic aromatic amines and then cleaving the Boc group.

Scheme 1.  Reagents and conditions: (a) (Boc)₂O, NaOH, THF; (b) EDCI, HOBT, DCM, NH₂-AA-COOCH₃; (c) NaOH, CH₃OH; (d) 3N HCl-EtOAc; (e) NH₂OK, CH₃OH;

Scheme 2.  Reagents and conditions: (a) EDCI, HOBT, DCM, NH²R⁴; (b) 3N HCl-EtOAc

Preliminary in vitro evaluation and SAR discussion

The newly synthesised 3-amino-2-hydroxyl-3-phenylpropanoic acid derivatives were assayed for the enzymatic inhibitions both on APN from porcine kidney (Microsomal; Sigma-Aldrich, St. Louis, MO, USA) and matrix metalloproteinases-2 (MMP-2). The results are listed in Table 1. Similar to APN, MMP-2 is also a zinc-dependent metalloproteinase, which is involved in the process of tumour invasion and metastasis¹⁹. The difference between them is that MMP-2 is an endopeptidase, while APN is an exopeptidase. All the target compounds were assayed for the inhibitory activities on APN and MMP-2 to observe the selectivity.

Table 1. The structures and in vitro APN inhibitory activities of the target compounds.

Compound	R1	R2	IC50^a(μM)
			APN	MMP-2
5a	−CH2CH(CH3)2	−OH	351.71 ± 0.70	214.87 ± 0.49
5b	−CH(CH3)2	−OH	NA^b	287.42 ± 0.93
5c	−CH3	−OH	342.42 ± 0.45	273.08 ± 0.25
5d	−CH(CH3)CH2CH3	−OH	NA^b	264.1 ± 0.5
5e		−OH	197.57 ± 11.02	254.11 ± 0.25
5f		−OH	NA^b	497.15 ± 0.91
5g		−OH	NA^b	407.25 ± 1.89
5h		−OH	NA^b	561.79 ± 2.55
5i			590.39 ± 10.61	690.44 ± 1.92
5j			512.91 ± 23.91	424.18 ± 1.18
5k			NA^b	185.90 ± 0.36
5l			NA^b	NA^b
5m			NA^b	NA^b
5n			104.53 ± 0.21	NA^b
5o			NA^b	NA^b
7a	−CH2CH(CH3)2	−NHOH	56.71 ± 0.01	64.60 ± 0.79
7b	−CH(CH3)2	−NHOH	76.95 ± 0.43	128.12 ± 0.14
7c	−CH3	−NHOH	NA^b	292.22 ± 2.08
7d	−CH(CH3)CH2CH3	−NHOH	82.89 ± 2.67	207.11 ± 1.32
7e		−NHOH	1.26 ± 0.01	212.97 ± 1.35
7f		−NHOH	44.00 ± 2.32	249.42 ± 0.94
7g		−NHOH	172.87 ± 1.99	298.79 ± 1.48
7h		−NHOH	157.64 ± 0.58	292.35 ± 1.97
9a			NA^b	NA^b
9b			362.76 ± 0.04	NA^b
9c			NA^b	NA^b
9d			NA^b	NA^b
Bestatin			2.55 ± 0.11	163.48 ± 1.13

^aValues are means of three experiments and standard derivation.

^bbNA, no activity.

As shown in Table 1, the results showed that most compounds had moderate MMP inhibitory activity, which was consistent with our early report. But the result of APN inhibition was not satisfied, most compounds did not show expected APN inhibition, and some compounds (such as 5b, 5d, 5f and so on) even had no APN inhibitory activity except for the hydroxamic acid derivatives 7a–7h. Of all these compounds, compound 7e showed to be the best APN inhibitor with the IC₅₀ value to 1.26 μM, which was much better than that of the positive control bestatin (IC₅₀ = 2.55 μM). By analysing the results, we supposed that these compounds may not have the same interaction model with bestatin by using the hydroxyl group and carbonyl group to chelate the zinc ion but using the hydroxamic acid group, this hypothesis could be verified by the docking results (Figure 2). If the phenyl group of 3-amino-2-hydroxyl-3-phenylpropanoic acid could form hydrophobic interaction with the S₁ pocket as bestatin, the distance of the hydroxyl group and carbonyl group to the zinc ion may be unsuitable because of the lack of one carbon, this may explain why compounds 5a–5o and 9a–9d had poor APN inhibitory activity. But compounds 7a–7h showed very good APN inhibitory activity because of the introduction of the strong ZBG hydroxamic acid group.

Figure 2.  The FlexX docking result of 7e with the active site of aminopeptidase N (PDB: 2DQM).

Comparing compounds 7a–7h, we can know that the APN inhibitory activities were different with the substitutions in R₁. Compounds 7a and 7b with the Leu and Val in R₁ had better APN inhibitory activity than the others (such as 7c, 7d, 7g and 7h), this may suggest that the Leu and Val in R₁ is good for APN inhibition. Compound 7e with phenylglycine in R₁ had the best APN inhibition with the IC₅₀ value to 1.26 ± 0.01 μM. Compound 7f with phenylalanine in R₁ also had good APN inhibition but not better than compound 7e with the IC₅₀ value to 44.00 ± 2.32 μM. Comparing with compounds 7e and 7f, we can know that the length of the α-carbon atom to the aromatic ring may be important for APN inhibition. Compound 7g with a hydroxyl group in the para-position of the phenyl group had less inhibitory activity than compound 7f, this may suggest that the additional hydroxyl group may be unsuitable for APN interaction.

Compound 7e showed the most inhibitory activity toward APN. To investigate the interaction of these compounds with APN, the docking studies of the representative compound 7e were carried out via Sybyl/Sketch module and optimised using Powell’s method with the Tripos force field with convergence criterion set at 0.05 kcal/ (Å mol) and assigned with Gasteiger–Hückel method. The docking study performed using Sybyl7.3/FlexX module, the residues in a radius of 7.0 Å around Bestatin in the co-crystal structure (PDB code: 2DQM) were selected as the active site. Other docking parameters implied in the program were kept default. The docking result was shown in Figure 2. As shown in Figure 2, 3-amino-2-hydroxyl-3-phenylpropanoic acid scaffold of compound 7e inserted into the S1 pocket, and the hydroxymate group chelated with the zinc ion in the active site, while the phenyl group of L-phenylglycine moiety interacted with the S1′ pocket. For a further and detail understanding of the binding mode of 7e with APN, a two-dimensional picture was also created with the program LIGPLOT, which was shown in Figure 3; we can see that the phenyl group of 3-amino-2-hydroxyl-3-phenylpropanoic acid could form hydrophobic interaction with the S₁ pocket formed by Arg⁸²⁵, Tyr³⁷⁶, Met²⁶⁰ and Tyr³⁸¹. The free amino group could form hydrogen bond with Arg⁸²⁵ at the distance of 2.64 Å. The phenyl group of L-phenylglycine moiety could form hydrophobic interaction with the S1′ pocket formed by Arg²⁹³, Asp³⁷² and His²⁹⁷. The two oxygen atoms of hydroxamic acid could chelate with the zinc ion of APN. Additionally, the hydroxyl group of hydroxamic acid could also form hydrogen bond with His²⁹⁷ and His³⁰¹ at the distance of 2.70 Å and 2.84 Å.

Figure 3.  The docking result of 7e with aminopeptidase N showed by LIGPLOT.

Although the computed information partially supported our assumption, the exact binding mode of the 3-amino-2-hydroxyl-3-phenylpropanoic acid derivatives with APN should be obtained from further X-ray crystal studies.

APN is highly expressed on the surface of human ovary clear cell carcinoma cell ES-2²⁰; to study the inhibitory activity of the target compounds toward human APN, we also tested the in vitro enzymatic inhibitions of compound 7e toward APN on ES-2. The results showed that compound 7e is also a dose-dependent APN inhibitor in the concentration range of 0.39–400 μM (Figure 4). The IC₅₀ value of compound 7e is 30.19 ± 1.02 μM, which is still better than that of bestatin (IC₅₀= 60.61 ± 0.1 μM).

Figure 4.  Inhibition of aminopeptidase N (APN)/CD13 activity on ES-2 induced by 7e or bestatin. OVCA cells (1 × 10⁵) seeded in 96-well plates were exposed to various concentrations of 7e or bestatin at 37°C, and the substrate L-leucine-p-nitroanilide was added. APN/CD13 activity was estimated by measuring absorbance at 405 nm using a microplate reader after 60 min of incubation. The inhibition rate was calculated by comparing to the control (without drug exposure). The bars indicate means ± standard deviation (n = 3).

Additionally, the effects of compound 7e on the proliferation of human ovary clear cell carcinoma cell ES-2 compared with bestatin were further assessed by using MTT assay, which were shown in Figure 5. The results showed that the anti-proliferative effects of compound 7e against ES-2 cell were slightly better than that of bestatin with the IC₅₀ value of 0.51 ± 0.08 mM versus 0.57 ± 0.04 mM.

Figure 5.  In vitro anti-proliferative activities of compound 7e and bestatin against ES-2 cells. The bars indicate means ± standard deviation (n = 3).

Conclusion

In summary, we have described the synthesis and properties of a series of novel (2R,3S)-3-amino-2-hydroxyl-3-phenylpropanoic acid derivatives as APN inhibitors. Because of the lack of one carbon as compared with 3-amino-2-hydroxyl-4- phenylbutyl acid, most of the target compounds did not show expected APN inhibitory activities, but the introduction of hydroxamate group to the compounds could greatly improve the inhibitory activities. The most effective compound 7e, not only had better inhibitory activity against APN both from porcine kidney and human but also showed slightly better anti-proliferative effects toward ES-2 cells than that of bestatin. Compound 7e could be used as a lead compound for further development of small molecular peptidomimetic APN inhibitors. Though the preliminary enzymatic results are not very encouraging, it still provides useful clues for further APN inhibitors exploitation.

Supplementary data

Supplementary data (Experimental details, and spectroscopic data for all the compounds and procedures for the activity assays were provided.).

Declaration of interest

This work was supported by National High Technology Research and Development Program of China (863 project; Grant No. 2007AA02Z314), National Nature Science Foundation of China (No. 90713041 and No. 30772654) and National Science and Technology Major Project (Grant No. 2009ZX09103-118).