Catalytic Hydrolysis of Phosphate Monoester by Supramolecular Complexes Formed by the Self-Assembly of a Hydrophobic Bis(Zn2+-cyclen) Complex, Copper, and Barbital Units That Are Functionalized with Amino Acids in a Two-Phase Solvent System

We previously reported on the preparation of supramolecular complexes by the 2:2:2 assembly of a dinuclear Zn2+-cyclen (cyclen = 1,4,7,10-tetraazacyclododecane) complex having a 2,2′-bipyridyl linker equipped with 0~2 long alkyl chains (Zn2L1~Zn2L3), 5,5-diethylbarbituric acid (Bar) derivatives, and a copper(II) ion (Cu2+) in aqueous solution and two-phase solvent systems and their phosphatase activities for the hydrolysis of mono(4-nitrophenyl) phosphate (MNP). These supermolecules contain Cu2(μ-OH)2 core that mimics the active site of alkaline phosphatase (AP), and one of the ethyl groups of the barbital moiety is located in close proximity to the Cu2(μ-OH)2 core. The generally accepted knowledge that the amino acids around the metal center in the active site of AP play important roles in its hydrolytic activity inspired us to modify the side chain of Bar with various functional groups in an attempt to mimic the active site of AP in the artificial system, especially in two-phase solvent system. In this paper, we report on the design and synthesis of new supramolecular complexes that are prepared by the combined use of bis(Zn2+-cyclen) complexes (Zn2L1, Zn2L2, and Zn2L3), Cu2+, and Bar derivatives containing amino acid residues. We present successful formation of these artificial AP mimics with respect to the kinetics of the MNP hydrolysis obeying Michaelis–Menten scheme in aqueous solution and a two-phase solvent system and to the mode of the product inhibition by inorganic phosphate.


Introduction
The regulation of protein phosphorylation and dephosphorylation is an important process in terms of cellular signal transduction, apoptosis, and the cell cycle among other issues. The phosphorylation of proteins is mediated by protein kinases, such as tyrosine kinases which are related to signal transduction pathways in living cells [1]. Indeed, several molecular targeted drugs that target tyrosine kinases in cancer cells and inhibit their action have been developed for the treatment of cancer [2,3]. On the other hand, dephosphorylation is promoted by the action of protein phosphatases such as alkaline phosphatase (AP) which contains two Zn 2+ ions in its active center [4][5][6][7][8][9][10]. Although some artificial phosphatases that mimic the active center of metallophosphatases have been reported [11][12][13][14][15][16][17][18][19][20][21][22], very few of them function as catalysts for the hydrolysis of a phosphate monoester such as mono(4-nitrophenyl) phosphate (MNP). Hence, dephosphorylation by artificial catalysts that mimic protein phosphatases remains a great challenge. It has been described that chemically synthesized enzyme models have several disadvantages including, (i) a lack of long-range interactions between catalysts and substrate and/or between functional groups in their active sites, (ii) the synthesis of such compounds is time-consuming, and (iii) they are active only in organic solvents in many cases [23]. Indeed, the synthesis of artificial enzyme models that are constructed by covalent bonds require long and complex synthetic routes especially for their functionalization, as a result, only a limited number of structures have been reported. These drawbacks, however, could be overcome by a supramolecular strategy utilizing the self-assembly of the artificial and functionalized molecular building blocks that are readily available .
In this context, we previously reported on the formation of the supramolecular complex 8a, by the 2:2:2 assembly of a bis(Zn 2+ -cyclen) complex (cyclen = 1,4,7,10-tetraazacyclododecane) containing 2,2 -bipyridyl (bpy) linker 1 (Zn 2 L 1 ), a dianion of barbital 4a (Bar), and a copper(II) ion in an aqueous solution which is stabilized by non-covalent bonds (Scheme 1a) [19,24]. The X-ray crystal structure of the supermolecule 8a revealed that it contains a Cu 2 (µ-OH) 2 core, which resembles the active centers of metallophosphatases such as AP, and that one of the ethyl groups of the Bar unit is located in close proximity to the Cu 2 (µ-OH) 2 core [24]. More importantly, 8a accelerates the hydrolysis of MNP, although the yield is low, possibly due to product inhibition by inorganic phosphate (HPO 4 2− ), a byproduct of the MNP hydrolysis. We also reported on the formation of the hydrophobic supramolecular complexes 9a-c and the amphiphilic supramolecular complexes 10a-c (Scheme 1a) by the 2:2:2 assembly of a hydrophobic bis(Zn 2+ -cyclen) complex containing two long alkyl chains 2 (Zn 2 L 2 ) and a complex that contained one long alkyl chain 3 (Zn 2 L 3 ), respectively, with Bar 2− derivatives and Cu 2+ in a two-phase solvent system (CHCl 3 /H 2 O) [53,54]. As shown in Scheme 2, we expected that the hydrophobic or amphiphilic 2:2:2 complexes 9 or 10 would be formed mainly in the organic layer, and that product inhibition by HPO 4 2− could be avoided, since the hydrophilic HPO 4 2− would be released into the aqueous layer, thus permitting the Cu 2 (µ-OH) 2 core to be regenerated. The results showed that 9 and 10 accelerated the hydrolysis of MNP and one of most important findings was that the hydrolysis of MNP by 9 and 10 in the two-phase solvent system obeyed Michaelis-Menten kinetics, suggesting that these reaction systems consisting of a supramolecular complex in two-phase solvent system closely mimic the active sites of AP. However, catalytic activity was observed only in the case of 10 (the catalytic turnover numbers (CTN) were 3~4 [54]), but not in 9 [53]. In addition, the product inhibition of 9 and 10 by inorganic phosphate (HPO 4 2− ) was not competitive, although it is well known that HPO 4 2− inhibits natural AP in a competitive manner. These results raised next question of how to design and synthesize good and appropriate models for natural enzymes such as AP based on a supramolecular strategy.   Scheme 2. Concept of the hydrophobic supramolecular complexes 9 and 10 and the hydrolysis of mono(4-nitrophenyl) phosphate (MNP) in a two-phase solvent system, in which the product inhibition by HPO4 2is avoided.
It is generally accepted that amino acids in the active site of AP contribute to the hydrolysis of phosphomonoesters; for example, the arginine residue captures and fixes the conformation of the phosphomonoester, and the serine residue attacks the phosphorous to promote the hydrolysis [4][5][6][7][8][9][10]. These well-grounded assumptions prompted us to introduce amino acids (alanine, serine, arginine, phenylalanine, and tyrosine) into the side chains of the Bar units for the easy and versatile construction of a combinatorial library of supramolecular phosphatases (Scheme 3). For example, functionalization with arginine (Arg) unit would be expected to stabilize the transition state that is produced in the hydrolysis of MNP (Scheme 3a) by electrostatic interactions. Second, the serine (Ser) residue would participate in a nucleophilic attack on the phosphorous of the MNP (Scheme 3b). Third, a phenylalanine (Phe) unit was introduced to support π-π interactions with MNP for a stronger complexation (Scheme 3c). In this paper, we report on the combinatorial construction of 2:2:2 supramolecular complexes 15-17 from dinuclear Zn 2+ -cyclen complexes 1 (Zn2L 1 ), 2 (Zn2L 2 ), or 3 (Zn2L 3 ) with the Bar derivatives that are functionalized with amino acids 11d-p as well as their synthetic intermediates 11a-c and Cu 2+ (Scheme 1b), their catalytic activity for the hydrolysis of MNP, and the results of a Michaelis-Menten kinetic study. These findings indicate that hydrophobic active sites of the supramolecular phosphatases well mimic the mode of the substrate (MNP) recognition and the product inhibition (by inorganic phosphate), and both of hydrophobic and hydrophilic property of active sites are important for the catalytic turnover of the system. Scheme 2. Concept of the hydrophobic supramolecular complexes 9 and 10 and the hydrolysis of mono(4-nitrophenyl) phosphate (MNP) in a two-phase solvent system, in which the product inhibition by HPO 4 2− is avoided.
It is generally accepted that amino acids in the active site of AP contribute to the hydrolysis of phosphomonoesters; for example, the arginine residue captures and fixes the conformation of the phosphomonoester, and the serine residue attacks the phosphorous to promote the hydrolysis [4][5][6][7][8][9][10]. These well-grounded assumptions prompted us to introduce amino acids (alanine, serine, arginine, phenylalanine, and tyrosine) into the side chains of the Bar units for the easy and versatile construction of a combinatorial library of supramolecular phosphatases (Scheme 3). For example, functionalization with arginine (Arg) unit would be expected to stabilize the transition state that is produced in the hydrolysis of MNP (Scheme 3a) by electrostatic interactions. Second, the serine (Ser) residue would participate in a nucleophilic attack on the phosphorous of the MNP (Scheme 3b). Third, a phenylalanine (Phe) unit was introduced to support π-π interactions with MNP for a stronger complexation (Scheme 3c). In this paper, we report on the combinatorial construction of 2:2:2 supramolecular complexes 15-17 from dinuclear Zn 2+ -cyclen complexes 1 (Zn 2 L 1 ), 2 (Zn 2 L 2 ), or 3 (Zn 2 L 3 ) with the Bar derivatives that are functionalized with amino acids 11d-p as well as their synthetic intermediates 11a-c and Cu 2+ (Scheme 1b), their catalytic activity for the hydrolysis of MNP, and the results of a Michaelis-Menten kinetic study. These findings indicate that hydrophobic active sites of the supramolecular phosphatases well mimic the mode of the substrate (MNP) recognition and the product inhibition (by inorganic phosphate), and both of hydrophobic and hydrophilic property of active sites are important for the catalytic turnover of the system. Micromachines 2019, 9, x 5 of 24 Scheme 3. Predicted effect of new barbital derivatives functionalized with amino acids on the hydrolysis of MNP. (a) It is considered that Arg stabilizes the transition state of the MNP hydrolysis by the electrostatic interactions between them, (b) Ser is supposed to undergo the nucleophilic attack to the phosphorous of MNP, (c) Phe is supposed to contribute to the π-π interactions with MNP.

General Information
All reagents and solvents were of the highest commercial quality and were used without further purification, unless otherwise noted. Anhydrous N,N-dimethylformamide (DMF) was obtained by distillation from calcium hydride. MNP was purchased from Nacalai Tesque (Kyoto, Japan). All aqueous solutions were prepared using deionized and distilled water. The Good's buffer reagents (Dojindo, pKa at 20 °C) were obtained from commercial sources: HEPES (2-(4-(2-hydroxyethyl)-1piperazinyl))ethanesulfonic acid, pKa = 7.6). For the measurement of UV/Vis spectra, CHCl3 was purchased from Nacalai Tesque (Kyoto, Japan). UV/Vis spectra were recorded on a JASCO V-550 spectrophotometer with quartz cuvettes (path length: 10 mm). IR spectra were recorded on a Perkin-Elmer attenuated total reflectance (ATR)-IR spectrometer 100 at room temperature. Melting points were measured on a Yanaco MP-J3 Micro Melting Point apparatus and are uncorrected. 1 H-(300 and 400 MHz) and 13 C-(75 and 100 MHz) NMR spectra at 25 ± 0.1 °C were recorded on a JEOL Always 300 spectrometer and a JEOL Lamda 400 spectrometer. Tetramethylsilane (TMS) was used as the internal reference for 1 H-and 13 C-NMR measurements in CDCl3 and CD3OD. 3-(Trimethylsilyl)propionic-2,2,3,3-d4 acid sodium salt (TSP) was used as the external reference for 1 Hand 13 C-NMR measurements in D2O. Mass spectra was recorded on a JEOL JMS-700 and Varian 910-MS spectrometer. Elemental analyses were performed on a Perkin-Elmer CHN 2400 analyzer. Optical rotations were measured with a JASCO-P-1030 digital polarimeter in 50 mm cells using the D line of sodium (589 nm). Thin-layer chromatography (TLC) and silica gel column chromatography was performed using Merck Silica gel 60 F254 plate or Fuji Silysia Chemical CHROMATOREX NH-TLC PRATE, and Fuji Silysia Chemical FL-100D or Fuji Silysia Chemical CHROMATOREX NH chromatography Silica Gel, respectively.

5,5-Bis[3-(toluenesulfonyl)oxypropyl]barbituric acid (18)
A solution of 4c (698 mg, 2.86 mmol) and DMAP (35 mg, 0.29 mmol) in pyridine (5 mL) was stirred at 0 °C for 5 min, to which p-toluenesulfonyl chloride (1.53 g, 8.00 mmol) was slowly added, Scheme 3. Predicted effect of new barbital derivatives functionalized with amino acids on the hydrolysis of MNP. (a) It is considered that Arg stabilizes the transition state of the MNP hydrolysis by the electrostatic interactions between them, (b) Ser is supposed to undergo the nucleophilic attack to the phosphorous of MNP, (c) Phe is supposed to contribute to the π-π interactions with MNP.

General Information
All reagents and solvents were of the highest commercial quality and were used without further purification, unless otherwise noted. Anhydrous N,N-dimethylformamide (DMF) was obtained by distillation from calcium hydride. MNP was purchased from Nacalai Tesque (Kyoto, Japan). All aqueous solutions were prepared using deionized and distilled water. The Good's buffer reagents (Dojindo, pK a at 20 • C) were obtained from commercial sources: HEPES (2-(4-(2-hydroxyethyl)-1-piperazinyl))ethanesulfonic acid, pK a = 7.6). For the measurement of UV/Vis spectra, CHCl 3 was purchased from Nacalai Tesque (Kyoto, Japan). UV/Vis spectra were recorded on a JASCO V-550 spectrophotometer with quartz cuvettes (path length: 10 mm). IR spectra were recorded on a Perkin-Elmer attenuated total reflectance (ATR)-IR spectrometer 100 at room temperature. Melting points were measured on a Yanaco MP-J3 Micro Melting Point apparatus and are uncorrected. 1 H-(300 and 400 MHz) and 13 C-(75 and 100 MHz) NMR spectra at 25 ± 0.1 • C were recorded on a JEOL Always 300 spectrometer and a JEOL Lamda 400 spectrometer. Tetramethylsilane (TMS) was used as the internal reference for 1 H-and 13 C-NMR measurements in CDCl 3 and CD 3 OD. 3-(Trimethylsilyl)propionic-2,2,3,3-d 4 acid sodium salt (TSP) was used as the external reference for 1 H-and 13 C-NMR measurements in D 2 O. Mass spectra was recorded on a JEOL JMS-700 and Varian 910-MS spectrometer. Elemental analyses were performed on a Perkin-Elmer CHN 2400 analyzer. Optical rotations were measured with a JASCO-P-1030 digital polarimeter in 50 mm cells using the D line of sodium (589 nm). Thin-layer chromatography (TLC) and silica gel column chromatography was performed using Merck Silica gel 60 F254 plate or Fuji Silysia Chemical CHROMATOREX NH-TLC PRATE, and Fuji Silysia Chemical FL-100D or Fuji Silysia Chemical CHROMATOREX NH chromatography Silica Gel, respectively.

5,5-Bis(3-azidopropyl)barbituric acid (11a)
A solution of 18 (850 mg, 1.54 mmol) and sodium azide (280 mg, 4.31 mmol) in DMF (4 mL) was stirred at 80 • C for 5 h. After allowing the solution to cool to room temperature, the reaction mixture was poured into H 2 O and the resulting mixture was extracted with Et 2 O (40 mL × 3). The combined organic layer was washed with brine, dried over Na 2 SO 4 , and concentrated under reduced pressure. The resulting residue was purified by silica gel column chromatography (CHCl 3 /MeOH = 1/0 to 10/1) to give 11a as a white solid (434 mg, 96% yield). 1

Synthesis of Barbital Derivatives
The barbital derivatives that were functionalized with amino acid residues were synthesized as shown in Scheme 4. Compound 4c was synthesized from diethyl malonate via 4b according to our previous paper [24] and was reacted with p-toluenesulfonyl chrolide to afford the ditosylate 18. The reaction of 18 with sodium azide gave compound 11a. Our attempt to carry out Staudinger reactions [55,56] of 11a with the o-phosphynothioesters of amino acids resulted in failure. Therefore, the reduction of the azide groups of 11a were carried out to obtain the diamino intermediate 11c. It should be noted that two amino groups of 11c produced by the reduction of 11a were directly protected with a Boc group in situ for easy purification as a Boc-protected compound 11b and the subsequent deprotection of the Boc groups afforded 11c. The condensation reactions of 11c with Boc-protected amino acids in DMF in the presence of PyBOP and DIEA gave 11d-h. The deprotection of the Boc groups by treatment with HCl afforded the corresponding compounds 11i-m in quantitative yield.
The barbital derivative equipped with serine 11n was obtained from 11c and Fmoc-l-Ser(OtBu)-OH and then reacted with piperidine and TFA to give compound 11o and 11p, respectively.

Complexation Behavior of 1 (Zn2L 1 ) with Barbital Derivatives and Cu 2+ by UV/Vis Titrations.
In order to examine the formation of the 2:2 and 2:2:2 supramolecular complexes from the bis(Zn 2+ -cyclen) complexes 1 (Zn2L 1 ) and 2 (Zn2L 2 ) with the synthesized Bar units and Cu 2+ as shown in Scheme 1, UV/Vis titrations of the bis(Zn 2+ -cyclen) complexes with 11d and 11i, and then with Cu(NO3)2·3H2O were attempted. While the titrations of 2 in the two-phase solvent system (CHCl3/50 mM HEPES buffer (pH 7.4, I = 0.1 (NaNO3)) with 11d and 11i and Cu(ClO4)2·6H2O were unsuccessful, the UV/Vis titrations of 1 with 11d and 11i were successfully carried out in 10 mM HEPES buffer (pH 7.4, I = 0.1 (NaNO3)) at 37 °C. As shown in Figure 1a and b, 80 μM of 1 has an absorption maxima (λmax) at 287 nm, which increased upon the addition of 11d and decreased upon the addition of 11i, reaching a plateau at a 1:1 ratio, suggesting a 2:2 assembly of 1 with 11d or 11i as same as that with 4a. In addition, as shown in Figure 1c

Complexation Behavior of 1 (Zn 2 L 1 ) with Barbital Derivatives and Cu 2+ by UV/Vis Titrations
In order to examine the formation of the 2:2 and 2:2:2 supramolecular complexes from the bis(Zn 2+ -cyclen) complexes 1 (Zn 2 L 1 ) and 2 (Zn 2 L 2 ) with the synthesized Bar units and Cu 2+ as shown in Scheme 1, UV/Vis titrations of the bis(Zn 2+ -cyclen) complexes with 11d and 11i, and then with Cu(NO 3 ) 2 ·3H 2 O were attempted. While the titrations of 2 in the two-phase solvent system (CHCl 3 /50 mM HEPES buffer (pH 7.4, I = 0.1 (NaNO 3 )) with 11d and 11i and Cu(ClO 4 ) 2 ·6H 2 O were unsuccessful, the UV/Vis titrations of 1 with 11d and 11i were successfully carried out in 10 mM HEPES buffer (pH 7.4, I = 0.1 (NaNO 3 )) at 37 • C. As shown in Figure 1a,b, 80 µM of 1 has an absorption maxima (λ max ) at 287 nm, which increased upon the addition of 11d and decreased upon the addition of 11i, reaching a plateau at a 1:1 ratio, suggesting a 2:2 assembly of 1 with 11d or 11i as same as that with 4a. In addition, as shown in Figure 1c

Location of Complexes 13 and 16 in the Two-Phase Solvent System, as Determined by UV/Vis Spectra.
The distribution of 13 and 16 in the organic phase and the aqueous phase of the two-phase solvent system was determined from the UV/Vis absorption spectra of both layers, as described in our previous report [54]. For the in situ formation of 13 and 16, 11d and 11f were selected as more hydrophobic barbital units, and 4a, 11i, and 11k were selected as more hydrophilic barbital units. Two-phase solutions of Zn2L 2 (2) alone (40 μM), 2:2 complex (20 μM) of Zn2L 2 and barbital derivatives (6a, 13d, 13f, 13i, and 13k), and 2:2:2 complex (20 μM) of Zn2L 2 , barbital derivatives, and Cu 2+ (9a, 16d,  16f, 16i, and 16k) were prepared in CHCl3/50 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO3) (1/1), incubated for 18 h at 37 °C, and centrifuged (2000 rpm × 10 min) at room temperature. Pictures of each complex and their distribution ratios are summarized in Figure 2 and Table 1. The findings indicate that 2 is located mostly in the organic layer (ca. 99%), and the addition of barbital derivatives did not significantly affect the distribution of 6a, 13d, 13f, 13i, and 13k (ca. 96%->99%), implying that these complexes are distributed mainly in the organic layer. Since the addition of Cu 2+ quenched the emission from the bpy units, it was not possible to determine the distribution of 9a, 16d, 16f, 16i, and  16k in both layers. These behaviors of 6a, 13d, 13f, 13i, and 13k are different from 7a prepared from 3 (Zn2L 3 ) and 4a (Bar), which were distributed both in the aqueous and the organic layers [54].

Hydrolysis of MNP by 2:2:2 Complexes in a Two-Phase Solvent System
We conducted the hydrolysis of MNP (100 µM) in the single-phase solvent system (10 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO 3 )) at 37 • C in the presence of 8a or 15i-m (prepared from 1, 11i-m, and Cu 2+ ) (100 µM) [24] and in the two-phase solvent system (CHCl 3 /50 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO 3 )) (2/8) at 37 • C in the presence of 9a or 16i-m (prepared from 2, 11i-m, and Cu 2+ ) (20 µM in the total solution including the aqueous phase and the CHCl 3 phase). The results showed that the MNP hydrolysis in the presence of 15i-m was much slower than that of 8a in the single-phase solvent system, as shown in Figure 3. We conducted the hydrolysis of MNP (100 μM) in the single-phase solvent system (10 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO3)) at 37 °C in the presence of 8a or 15i-m (prepared from 1, 11i-m, and Cu 2+ ) (100 μM) [24] and in the two-phase solvent system (CHCl3/50 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO3)) (2/8) at 37 °C in the presence of 9a or 16i-m (prepared from 2, 11i-m, and Cu 2+ ) (20 μM in the total solution including the aqueous phase and the CHCl3 phase). The results showed that the MNP hydrolysis in the presence of 15i-m was much slower than that of 8a in the single-phase solvent system, as shown in Figure 3. As shown in Figure 4a, the activity of 16i-m (20 μM) was similar to that of 9a in the two-phase solvent system with negligible catalytic turnover (note that [16i-m] = 20 μM), suggesting that more hydrophobic supermolecules have higher hydrolysis activity in the two-phase solvent system than in the single-phase solvent system. Interestingly, the MNP hydrolysis activity for 16d-h (prepared from 2, 11d-h, and Cu 2+ ) was higher than that of 16i-m under the same conditions, as shown in Figure  4b, and the hydrolysis yields for 16d and 16f were in excess of 20% after 1 day and in excess of 25%-40% after 7 days, indicating that these complexes had catalytic activities. Moreover, the activities of 16a and 16b were similar to that for 16d-h ( Figure 5), suggesting that the azide and Boc-protected amino groups on the Bar unit provide a similar effect on the hydrolysis of MNP as those of the protected amino acid residues in 16d-g. As shown in Figure 4a, the activity of 16i-m (20 µM) was similar to that of 9a in the two-phase solvent system with negligible catalytic turnover (note that [16i-m] = 20 µM), suggesting that more hydrophobic supermolecules have higher hydrolysis activity in the two-phase solvent system than in the single-phase solvent system. Interestingly, the MNP hydrolysis activity for 16d-h (prepared from 2, 11d-h, and Cu 2+ ) was higher than that of 16i-m under the same conditions, as shown in Figure 4b, and the hydrolysis yields for 16d and 16f were in excess of 20% after 1 day and in excess of 25%-40% after 7 days, indicating that these complexes had catalytic activities. Moreover, the activities of 16a and 16b were similar to that for 16d-h ( Figure 5), suggesting that the azide and Boc-protected amino groups on the Bar unit provide a similar effect on the hydrolysis of MNP as those of the protected amino acid residues in 16d-g. We conducted the hydrolysis of MNP (100 μM) in the single-phase solvent system (10 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO3)) at 37 °C in the presence of 8a or 15i-m (prepared from 1, 11i-m, and Cu 2+ ) (100 μM) [24] and in the two-phase solvent system (CHCl3/50 mM HEPES buffer (pH 7.4) with I = 0.1 (NaNO3)) (2/8) at 37 °C in the presence of 9a or 16i-m (prepared from 2, 11i-m, and Cu 2+ ) (20 μM in the total solution including the aqueous phase and the CHCl3 phase). The results showed that the MNP hydrolysis in the presence of 15i-m was much slower than that of 8a in the single-phase solvent system, as shown in Figure 3. As shown in Figure 4a, the activity of 16i-m (20 μM) was similar to that of 9a in the two-phase solvent system with negligible catalytic turnover (note that [16i-m] = 20 μM), suggesting that more hydrophobic supermolecules have higher hydrolysis activity in the two-phase solvent system than in the single-phase solvent system. Interestingly, the MNP hydrolysis activity for 16d-h (prepared from 2, 11d-h, and Cu 2+ ) was higher than that of 16i-m under the same conditions, as shown in Figure  4b, and the hydrolysis yields for 16d and 16f were in excess of 20% after 1 day and in excess of 25%-40% after 7 days, indicating that these complexes had catalytic activities. Moreover, the activities of 16a and 16b were similar to that for 16d-h ( Figure 5), suggesting that the azide and Boc-protected amino groups on the Bar unit provide a similar effect on the hydrolysis of MNP as those of the protected amino acid residues in 16d-g.     MNP by 16e, 16j, 16n, 16o, and 16p 100 μM and [16e, 16j, 16n, 16o, and 16p] = 20 μM). The findings suggest that 16e and 16n consisting of the fully protected Ser have a higher activity than 16j, 16o, and 16p in which the Bar units contain fully or partially deprotected Ser.       The hydrolysis of MNP (100 μM) by the amphiphilic supramolecular complexes 17d or 17f (prepared from 3, 11d, or 11f, and Cu 2+ ) (20 μM) was also examined. As shown in Figure 7, the hydrolytic activity of 17d and 17f were higher than 9a, but slightly lower than 16d and 16f.  Figure 8 (the results for 9a are included in all of the graphs as standard data), in which 16d-g formed from the N-protected barbital units 11d-g showed higher hydrolysis activities. It should be noted that 16f demonstrated the highest MNP hydrolysis yield as shown in Figure 8b, and the catalytic turnover number (CTN) was more than 2 after 1 day. The CTN values for 9a, 10a, 16b, 16dn, 17d, and 17f at [MNP] =1000 μM, which are summarized in Figure 9, imply that the CTNs of 16b, 16e, 16f, 16g, 17d, and 17f are over 2, while these values are smaller than that of our previously reported 10a (~4) [54]. The hydrolysis of MNP (100 µM) by the amphiphilic supramolecular complexes 17d or 17f (prepared from 3, 11d, or 11f, and Cu 2+ ) (20 µM) was also examined. As shown in Figure 7, the hydrolytic activity of 17d and 17f were higher than 9a, but slightly lower than 16d and 16f. The hydrolysis of MNP (100 μM) by the amphiphilic supramolecular complexes 17d or 17f (prepared from 3, 11d, or 11f, and Cu 2+ ) (20 μM) was also examined. As shown in Figure 7, the hydrolytic activity of 17d and 17f were higher than 9a, but slightly lower than 16d and 16f.  Figure 8 (the results for 9a are included in all of the graphs as standard data), in which 16d-g formed from the N-protected barbital units 11d-g showed higher hydrolysis activities. It should be noted that 16f demonstrated the highest MNP hydrolysis yield as shown in Figure 8b, and the catalytic turnover number (CTN) was more than 2 after 1 day. The CTN values for 9a, 10a, 16b, 16dn, 17d, and 17f at [MNP] =1000 μM, which are summarized in Figure 9, imply that the CTNs of 16b, 16e, 16f, 16g, 17d, and 17f are over 2, while these values are smaller than that of our previously reported 10a (~4) [54].  Figure 8 (the results for 9a are included in all of the graphs as standard data), in which 16d-g formed from the N-protected barbital units 11d-g showed higher hydrolysis activities. It should be noted that 16f demonstrated the highest MNP hydrolysis yield as shown in Figure 8b, and the catalytic turnover number (CTN) was more than 2 after 1 day. The CTN values for 9a, 10a, 16b, 16d-n, 17d, and 17f at [MNP] =1000 µM, which are summarized in Figure 9, imply that the CTNs of 16b, 16e, 16f, 16g, 17d, and 17f are over 2, while these values are smaller than that of our previously reported 10a (~4) [54].

Entry
Cat. a V max (µM min −1 ) K m (µM) k (min −1 ) b K i (µM) K m /K i CTN c site of AP, as shown in Figure 11b and 11c. Therefore, we conclude that the hydrophobicity of supramolecular complexes is important in terms of improving the stabilization of the supermolecule-MNP complexes (i.e., the ES complexes) that have smaller Km values and the effective extraction of MNP from the aqueous layer. Namely, the hydrophobic active site of the artificial supramolecular complexes is important for mimicking the hydrophobic active site of natural AP and hydrophobicity/hydrophilicity balance is important for catalytic activity.

Conclusions
In conclusion, we report on the design, and synthesis of Bar building units that are functionalized with amino acids and related units, with which many supramolecular complexes are formed by the 2:2:2 self-assembly of 2 (Zn 2 L 2 ) and Cu 2+ in a two-phase solvent system based on the structure of the active site of AP. The ease of functionalizing the barbital units allowed us to construct a series of various supramolecular complexes which were then used to assess their catalytic activity for the hydrolysis of MNP and related studies of the reaction mechanism. A comparison of the hydrolysis of MNP in a single-phase system and that in the two-phase system in the presence of 8, 9, 15, or 16 strongly suggest that two-phase solvent system contributes to the improvement in the activity of MNP hydrolysis by reducing the product inhibition by HPO 4 2− . In addition, the hydrolysis of MNP by the hydrophobic supramolecular complexes follows Michaelis-Menten kinetics, and functionalization with Boc-protected phenylalanine (16f) results in a higher V max and a lower K m , and a greater k value than the corresponding values for 9a. The K m /K i value for 16f (ca. 5.4) is close to that for AP (ca. 2.3) and competitive inhibition by inorganic phosphate was observed, unlike for our previously reported complexes such as 10a-c. We therefore conclude that the distribution of the Cu 2 (µ-OH) 2 active site of the supermolecules in both layers in the two-phase solvent system contributes to the improvement of catalytic turnover, and the formation of hydrophobic Cu 2 (µ-OH) 2 site mainly in organic layer appropriately mimics the binding mode of AP with the substrates (for the E-S complexation) and the inhibitors (product inhibition). These findings should be highly useful for the design of more efficient catalysts in reference to biochemistry including enzymatic reactions. Funding: This research received no external funding.