Design and Synthesis of Chiral Zn2+ Complexes Mimicking Natural Aldolases for Catalytic C–C Bond Forming Reactions in Aqueous Solution

Extending carbon frameworks via a series of C–C bond forming reactions is essential for the synthesis of natural products, pharmaceutically active compounds, active agrochemical ingredients, and a variety of functional materials. The application of stereoselective C–C bond forming reactions to the one-pot synthesis of biorelevant compounds is now emerging as a challenging and powerful strategy for improving the efficiency of a chemical reaction, in which some of the reactants are subjected to successive chemical reactions in just one reactor. However, organic reactions are generally conducted in organic solvents, as many organic molecules, reagents, and intermediates are not stable or soluble in water. In contrast, enzymatic reactions in living systems proceed in aqueous solvents, as most of enzymes generally function only within a narrow range of temperature and pH and are not so stable in less polar organic environments, which makes it difficult to conduct chemoenzymatic reactions in organic solvents. In this review, we describe the design and synthesis of chiral metal complexes with Zn2+ ions as a catalytic factor that mimic aldolases in stereoselective C–C bond forming reactions, especially for enantioselective aldol reactions. Their application to chemoenzymatic reactions in aqueous solution is also presented.


Introduction
C-C bond formation is one of the fundamental transformations in organic synthesis. Extension of a carbon framework via a series of C-C bond forming reactions is essential to the synthesis of natural products, pharmaceutically active compounds, active agrochemical ingredients, and related functional materials [1]. One of most important applications of C-C bond forming reactions is one-pot synthesis, whereby reactants are subjected to successive chemical reactions in just one reactor. This methodology is convenient not only in the laboratory, but also in industrial reactions, because lengthy separation and purification processes of the intermediates can be avoided, resulting in time and resource-saving and, eventually, in a more efficient chemical synthesis. Despite the remarkable progress achieved in one-pot multistep synthetic methodologies including enantioselective C-C bond formation in organic solvents, only a few attempts have been made to combine a chemical catalyst and a biocatalyst in a one-pot multistep process, especially in water-containing solvent systems [2].
It is clear that much can be learned from natural enzymes for the design of water-soluble asymmetric catalysts. Typical examples of enzymes that catalyze C-C bond forming reactions in living systems would be aldolases, a class of enzymes that accelerate aldol and retro-aldol reactions in a stereospecific and reversible manner in natural metabolic pathways [3]. For example, fructose 1,6-bis(phosphate) aldolase (FBP-aldolase, EC 4.1.2.13) catalyzes the cleavage of D-fructose 1,6-bis(phosphate) (FBP) to give dihydroxyacetone phosphate (DHAP) and D-glyceraldehyde 3-phosphate (G3P), as well as the reverse formation of FBP from DHAP and G3P (Scheme 1). Natural aldolases can be classified into two groups on the basis of their reaction mechanisms. In the case of class I aldolases, an enamine intermediate is formed between the lysine residue of the enzyme and the carbonyl group of the substrate. In class II aldolases, a zinc(II) ion cofactor acts as a Lewis acid to generate enolates at the active site.   The use of aldolases in organic and bio-organic synthesis has been found to be an effective method for producing aldol products with high stereoselectivities in aqueous solution [4]. Wong et al. reported on a short step synthesis of 1-deoxynojirimycin, a glycosidase inhibitor, in which aldol reaction between DHAP and 1, using FBP-aldolase is conducted to obtain the important intermediate, 2 (Scheme 2a) [5]. In addition, Lerner and co-workers developed catalytic antibodies, which were obtained after screening of polyclonal antibodies for binding with the hapten [6,7]. The aldolase antibodies were found to catalyze aldol reactions by the enamine mechanism, analogous to class I aldolases. The aldol reaction between 3 and hydroxyacetone (HA, 4), catalyzed by aldolase antibody 38C2, has been applied in highly enantioselective total syntheses of (1R, 1'R, 5'R, 7'R)-and (1S, 1'R, 5'R, 7'R)-1-hydroxy-exo-brevicomin (Scheme 2b) [8]. Chiral organocatalysts have recently emerged as a reagent class representing a new methodology for stereoselective aldol reactions, in that they are capable of mimicking class I aldolases [9]. In 2000, List, Lerner, and Barbas reported that L-proline serves as a catalyst for direct aldol reactions of acetone and benzaldehyde derivatives 6 to give aldol adducts such as 7 in good chemical yields and with moderate enantioselectivities (Scheme 3, CTN stands for catalytic turnover number) [10]. It was suggested that L-proline forms an enamine intermediate that react with aldehydes to give aldol products [11].
As the direct aldol reaction catalyzed by L-proline was first reported, a variety of organocatalysts for aldol reactions have been reported [12]. Representative examples are shown in Scheme 4. Maruoka and co-workers successfully extended the concept of amino acid catalysis to novel catalysts containing a binaphthyl or biphenyl axial chirality (e.g., 8) [13]. The use of only 0.1 mol % of 8 in acetone afforded aldol products such as 7b with high yields and enantioselectivities (Scheme 4a). Barbas et al. reported on asymmetric aldol reactions between cyclohexanone 9 and benzaldehydes 6 in water catalyzed by a combination of a lipophilic diamine 10 and trifluoroacetic acid (TFA) [14] (Scheme 4b).
Hayashi's group developed a proline-based catalyst, diarylprolinol 12, for direct crossed-aldol reactions of acetaldehyde in DMF [15] (Scheme 4c). Barbas   A number of excellent studies on chiral metal catalysts for stereoselective aldol reactions have also been reported [17]. Examples of Zn 2+ catalysts for direct asymmetric aldol reactions include Et 2 Zn/linked BINOL 19 developed by Shibasaki [18], Trost's Zn 2+ -semi crown ether 22 [19], and so forth, which can be considered as class II aldolase mimics functioning in organic solvents (Scheme 5). Both catalysts showed high catalytic activities on α-hydroxyketones (18 and 21) by dinuclear Zn 2+ coordination sites for the enolate of α-hydroxyketones and aldehydes. To date, natural and artificial catalysts that possess both functionalities, i.e., Schiff-base forming part and Zn 2+ site, have scarcely been reported. Some Zn 2+ complexes of proline derivatives have been demonstrated to have the ability to catalyze direct asymmetric aldol reactions in aqueous media, thus mimicking class I and class II aldolases (Scheme 6). Reymond, Darbre, et al. showed that the 1:2 complex of Zn 2+ and L-proline 24 (Zn(L-Pro) 2 ) catalyzes aldol reactions of acetone and dihydroxyacetone 14 in aqueous system [20,21]. The use of 5 mol % of 24 gave aldol products from acetone and 6b in quantitative yields, with 56% ee (R), at room temperature (Scheme 6a). They suggested that in the aldol reactions of these substrates, 24 forms enamine species with acetone, but that it forms Zn 2+ -enolate intermediates with 14. Mlynarski's group reported on Zn 2+ complexes of a C 2 -symmetric chiral ligands containing two amino acid units, such as 25 [22]. These catalysts showed high reactivities and enantioselectivities in catalytic aldol reactions between acetone, cyclohexanone 9, and hydroxyacetone 4 and several aldehydes in aqueous systems (up to 99% ee) (Scheme 6b).
With the backgrounds described above, we became interested in the reactivities and enantioselectivities of novel asymmetric catalysts, each dual-functionalized with an enamine-forming amino group and a Zn 2+ complex of macrocyclic polyamines such as cyclen ( [12]aneN 4 ) or [15]aneN 5 . The development of enantioselective aldol reactions was envisaged, as they are one of the most important C-C bond forming reactions for producing β-hydroxy carbonyl compounds bearing two new stereogenic centers at the α-and β-positions of the carbonyl groups [17,26,27]. It should be noted that these reactions are performed mainly in aqueous solvents, which are considered to have enormous potential as a reaction mediums and are critical for chemoenzymatic reactions [28][29][30]. One of the advantages of these aldol reactions is that they would be applicable to the one-pot synthesis of biorelevantly important compounds by combination with enzymatic reactions. These results are reviewed below. The initially designed and synthesized Zn 2+ complexes for stereoselective direct aldol reactions include 31 (L-ZnL 3 prepared from L-prolyl-pendant [15] (Figure 1). The ligands for Zn 2+ complexes 31-37 (L-ZnL 3 -D-ZnL 8 ) were synthesized from tetrakis(tert-butyloxycarbonyl)- [15]aneN 5 (4-Boc- [15]aneN 5 ) or tris(tert-butyloxycarbonyl)- [12]aneN 4 (3-Boc- [12]aneN 4 ) with N-protected amino-acid derivatives [31]. The Zn 2+ complexes were prepared in situ immediately prior to use by reacting the acid-free ligands with Zn 2+ ions.

Enantioselective Aldol Reactions in Aqueous Media Catalyzed by Chiral Zn 2+ Complexes
On the basis of the data on the Zn 2+ complexation behavior of the chiral ligands, described above, the aldol reaction between acetone and 2-chlorobenzaldehyde 6a in the presence of the chiral catalysts in DMSO/acetone or acetone/H 2 O systems was examined. The results are summarized in Table 1. Most of the reactions in Table 1 were carried out at 37 °C as an enzyme model study.
In the acetone/H 2 O system, L-proline gave 7a in 22% yield and with 48% ee (R) at 37 °C (entry 4), both of which were lower than those in the DMSO/acetone system (entry 1). In contrast, 31 (L-ZnL 3 ) gave better yield (43% yield with a nearly racemic adduct) at 37 °C (entry 5) than in DMSO/acetone (entry 2). Interestingly, 32 (L-ZnL 4 ) gave 7a in good yield (73%) and enantioselectivity (80% ee (R)) in entry 6. Metal-free 29 (L-L 4 ) gave only the racemic aldol product (entry 7), whereas the Cd 2+ and Cu 2+ complexes of L-L 4 (41a and 41b) promoted aldol reactions only to a negligible extent (entries 8 and 9). It had previously been reported that the Lewis acidity of the Zn 2+ -cyclen complex is higher than that of Cd 2+ -cyclen [33]. We therefore concluded that the Lewis acidity of Zn 2+ is an important factor for this enantioselective aldol reaction.
These results suggest that primary or secondary amino groups on the side chains are important and that the amino acid portions of 31-33 (L-ZnL 3 -L-ZnL 5 ), 36 (L-ZnL 8 ) and 37 (D-ZnL 8 ), function as bases for the deprotonation of acetone activated by Zn 2+ , rather than as Schiff-base forming units.  [31,34]); f Zn 2+ complexes, L-CdL 4 , and L-CuL 4 were formed in situ; g Isolated Zn 2+ complexes were used; h L-L 4 was extracted with CHCl 3 from aq. NaOH (pH > 12) prior to use.
Next, 53 and 54 (ZnL 21 and ZnL 22 ), which contained dipeptide side chains, were prepared based on the assumption that the presence of more hydrophobic and hydrogen bonding functionalities around the Zn 2+ site would improve their catalytic activity. However, these Zn 2+ complexes resulted in low chemical and optical yields (entries 15 and 16), thus, suggesting that one amino-acid side chain is suitable for aldol reactions catalyzed by the ZnL series. Table 4 summarizes the aldol reactions between acetone and various benzaldehydes (6b-c and 6f-g), as catalyzed by 45 (L-ZnL 13 ) and 49 (L-ZnL 17 ) in acetone/H 2 O (9/1). When 4-chlorobenzaldehyde 6c was used, 7c was obtained in good yield and 94% ee (R) and 95% ee (R), respectively (entries 3 and 4) and, when 4-, 3-, and 2-nitrobenzaldehydes (6b, 6f, and 6g) were used as acceptors, high chemical and optical yields were similarly observed (entries 1, 2, and 5-8).  In order to examine the issue of whether amino groups of L-proline, 32 (L-ZnL 4 ), and 33 (L-ZnL 5 ) form Schiff-bases with substrates, we initially carried out a UV titration of acetone with L-proline and 32 (L-ZnL 4 ). However, the UV spectral change was negligible.
As described in the Introduction, it was suggested that the mechanism of the aldolase antibodies involves enamine formation as the result of a reaction between the ε-amino group of Lys and the ketone substrate [7]. In this scenario, it was reported that enaminone between the Lys residue of the antibody and β-diketone exhibit a strong UV absorption at 316 nm (ε 316 ~ 15,000 M −1 ·cm −1 ).

Stopped-Flow Experiments to Determine the Rates of ZnL-(acac) − Complexation
Herein, stopped-flow experiments were performed to more-precisely determine the rates of formation of ZnL-(acac) − complex. The increase in UV/Vis absorption of acac (0.2 mM) with 33 (L-ZnL 5 , 3 mM) at 294 nm was monitored and a rate constant of 6.18 (±0.03) × 10 −2 s −1 was calculated from the resulting curve by fitting to a single exponential equation. Similar studies of the complexation between acac and 45 (L-ZnL 13 ) and 49 (L-ZnL 17 ) afforded rate constants of 9.03 (±0.07) × 10 −2 s −1 and 7.47 (±0.05) × 10 −2 s −1 , respectively, which were almost the same as that of 33 (L-ZnL 5 ). These formation rates of ZnL-(acac) − complex are about 1.4 × 10 5 higher than that for enaminone 55.  Figure 4). On the other hand, in the presence of 32 (L-ZnL 4 ), it was found that 6a was converted quantitatively in 24 h at 37 °C (plain curve in Figure 4). The initial reaction rate of the aldol reaction between acetone and 6a via 58 is approximately 10 times higher than that via 57, indicating that the Zn 2+ -enolate is more reactive than enamine species. Scheme 11. Comparison of the reactivity of Zn 2+ -enolate and enamine intermediates by 1 H-NMR spectroscopy.  . A Zn 2+ ion is 6-coordinated by these nitrogen atoms of a cyclen ring, one nitrogen atom of the side chain, and two oxygen atoms of the NO 3 anion. All of the hydrogen atoms and one NO 3 anion are omitted for clarity.

X-ray Crystal Structure of 48 (L-ZnL 16 )
Initially, it was assumed that the NH 2 group of the amino acid moiety would coordinate weakly or not at all to Zn 2+ center (Scheme 8). X-ray crystal structure analyses of 48 (L-ZnL 16 ) disclosed that the Zn 2+ was coordinated not only by three nitrogen atoms (N(5), N(8), and N(11)) of cyclen and a NO 3 anion, but also by the nitrogen atom (N(14)) of β-naphthylalanyl moiety, as shown in Figure 5. The approximate Zn-N bond lengths are 2.12 Å for Zn-N(5), 2.14 Å for Zn-N(8), 2.15 Å for Zn-N(11), and 2.09 Å for Zn-N(14), thus, implying that these Zn-N coordinate-bond lengths are almost identical. In this structure, the NO 3  Zn 2+ -bound NO 3 anion is replaced by H 2 O in aqueous solution, based on our previous findings [23,24]. The Zn 2+ center in 50 (L-ZnL 18 ) was also coordinated by the nitrogen atom of diphenylalanyl moiety and by one water molecule [35].

Proposed Mechanism for the Aldol Reaction of Acetone Catalyzed by Chiral Zn 2+ Complexes
Proposed reaction mechanisms for the ZnL-catalyzed aldol reactions of acetone and benzaldehydes 6 based on the aforementioned results are shown in Scheme 12. Our initial hypothesis involved path A, in which the amine group of the side chain in Zn 2+ complexes deprotonated the α-proton of acetone, which was activated by coordination to the Lewis acidic Zn 2+ center, as shown in 60, to generate the Zn 2+ -enolate complex 61. However, given the data reported herein, this pathway appears to be less plausible, as the basicity of the amino side chain would be lowered by its coordination to the Zn 2+ center, as observed in the X-ray crystal structure ( Figure 5).

Scheme 12.
Proposed mechanism for the aldol reaction of acetone catalyzed by chiral Zn 2+ complexes. Herein, two other possibilities were considered, namely, paths B and C. It has been reported that Zn 2+ -bound HO − and alkoxide species can act as bases and nucleophiles [23,24]. As shown in Scheme 7, the Zn 2+ -bound HO − in the class II aldolase model complex 27b (ZnL 2 (OH − )) is considered to deprotonate the α-proton to the carbonyl group of the phenacyl side chain to give the Zn 2+ -enolate form 27c (Zn(H -1 L 2 )) in aqueous solution [25]. Accordingly, it is hypothesized in path B that the OH − of 59b deprotonates the α-proton of acetone with the aid of the Lewis acidic Zn 2+ in 63 and generates the Zn 2+ -enolate intermediate 64.
Path C shows another possibility, in which the Zn 2+ -bound OH − deprotonates the NH 2 group to give rise to a NH − species 62, which deprotonates acetone, thus resulting in the formation of Zn 2+ -enolate intermediate 64. However, this is unlikely because the pK a value of the amine group would be over 30, which is much higher than those of Zn 2+ -bound H 2 O and alcohol moieties in typical Zn 2+ -cyclen complexes (pK a value: 7~9) [23,24]. Table 6. Results for an asymmetric aldol reaction between 4 and 6b catalyzed by 32 (L-ZnL 4 ) and 33 (L-ZnL 5 ). A consideration of these points allows us to conclude that path B is the most plausible among the three possibilities, although we do not completely rule out the path A scenario. Then, Zn 2+ -enolate intermediate 64 reacts with an acceptor aldehyde through six-membered transition state 65 to afford 66, from which the aldol product is released and 59 is regenerated. The enantioselectivity could be explained in terms of the Zimmerman-Traxler-type transition state 65 [36]. We assume that the Zn 2+ in 65 is 5-(or 6-) coordinated by four nitrogen atoms of ligand, the enolate of the acetone, and the carbonyl group of the aldehyde, by analogy with the 5-(or 6-)coordinated Zn 2+ in 32-(acac) − complex 56 in Scheme 10, and that enolate predominantly attacks at the Re-face of the aldehydes.
Whereas 32 (L-ZnL 4 ) afforded the aldol products anti-67, from 4 and 6b, and syn-16, from 14 and 6b, with poor enantioselectivities, 33 (L-ZnL 5 ) gave the syn forms of 67 and 16 with moderate enantioselectivities (Tables 6 and 7). These results suggest that a primary amine unit on the side chain is required for enantiodiscrimination in aldol reactions between α-hydroxyketones, such as 4 and 14 and aldehydes [16]. It was assumed that the enantioselectivity in aldol reactions of 14 is influenced by two hydroxyl groups.

Introduction for One-Pot Chemoenzymatic Synthesis
Chemoenzymatic synthesis by the combined use of chemical catalysis and biocatalysis is a powerful methodology for the multistep synthesis of biologically important compounds, drugs, and so on. Organic reactions by artificial catalysts are generally conducted in organic solvents, as many organic molecules, such as reagents, catalysts, and intermediates, are usually not stable or soluble in water. In contrast, enzymatic reactions in living systems are conducted in aqueous solvents, because most of enzymes are functional only within a narrow range of temperature and pH levels and not so stable in less polar organic environments. This discrepancy makes it difficult to conduct one-pot chemoenzymatic reactions in an organic environment. Recently, the development of bioorthogonal reactions are growing as chemical reactions that neither interact with nor interfere with a biological system under physiological conditions [40] and it could be applied to chemoenzymatic synthesis.
To date, a number of examples of one-pot processes have been developed, based on chemocatalytic tandem reactions, multienzymatic reactions, and biotechnological reactions [41][42][43]. It is assumed that one-pot processes involving chemical catalysts and biocatalysts take full advantage of the productivity of chemical catalysts and the chemo-, region-, and stereoselectivity of biocatalysts [2]. For example, Gröger et al. reported on an enantioselective synthesis of ethyl (S)-3-aminobutanoic acid 74 by means of combinations of aza-Michel reaction between 70 and 71, and the kinetic resolution via aminolysis catalyzed by Candida antarctica lipase to give 72, followed by hydrolysis and hydrogenation (Scheme 13a) [44]. Chiral 1,3-diols derivatives of aromatic compounds such as 76 were synthesized by enantioselective aldol reactions of acetone with 6c using organocatalysts 75 and successive enantioselective reduction of ADH and NADH system (Scheme 13b) [45]. Chiral biaryl-containing alcohols, 78 and 80, were synthesized in very high optically yields using a combination of cross-coupling reactions of 77 with 79 promoted by palladium catalysts and enantioselective enzymatic reduction (Scheme 13c) [46]. Scheme 13. (a) Sequential and modular synthesis of enantiomerically pure β-amino acids; (b) Sequential and modular synthesis of chiral 1,3-diols with two stereogenic centers; (c) Combination of a palladium-catalyzed cross-coupling with an asymmetric biotransformation.
It was expected that the one-pot chemoenzymatic synthesis by the combined use of chiral Zn 2+ catalysts and enzymes in an aqueous solvent would be useful for the selective synthesis of all of the possible stereoisomers of 1,3-diols 76 in a one-pot manipulation involving enantioselective aldol reactions of acetone with benzaldehydes to give 7 using chiral Zn 2+ complexes, and the successive enantioselective reduction of 7 to give 76 using oxidoreductases with the regeneration of the NADH (reduced form of nicotinamine adenine dinucleotide) cofactor (Scheme 14) [54].

Scheme 14.
One-pot synthesis of optically active 1,3-diols 76 by chemoenzymatic synthesis in an aqueous solvent, involving enantioselective aldol reactions of acetone with 6 catalyzed by chiral Zn 2+ complexes (ZnL) and the successive enzymatic reduction of 7 with regeneration of the NADH cofactor.

Enantioselective Reductions of β-Hydroxyketones 7a-c Using Oxidoreductase
For the reduction of β-hydroxyketones 7a-c, we first chose Baker's yeast alcohol dehydrogenase (ADH) and oxidoreductases from Saccharomyces cerevisiae (S. cerevisiae) and Lactobacillus kefir (L. kefir) as these enzymes have been reported to catalyze the enantioselective reduction of 4-phenyl-4-hydroxy-2-butanone [55]. However, Baker's yeast ADH and S. cerevisiae ADH were not very effective for the reduction of 7a (entries 1 and 2 in Table 9). In entry 3, it was found that the ADH from L. kefir is effective for the stereoselective reduction of 7a, albeit this enzyme produces only the (R)-form of 76a.
We, thus, decided to test the "Chiralscreen ® OH" kit, which is available from Daicel Co., Ltd, Niigata, Japan, and contains a library of recombinant NADH-dependent oxidoreductases [56]. Generally, oxidoreductases require an equivalent amount of NAD(P)H (reduced form) for activity. The reductases in "Chiralscreen ® OH" themselves can reduce NAD + (oxidized form) to NADH using 2-propanol as a hydride source, so that the concentration of NADH can be reduced to a catalytic amount. It has been also reported that "Chiralscreen ® OH" can be used to catalyze the reduction of a variety of ketone even if the solubility of the substrate is low in aqueous solution [56]. The results of the stereoselective reduction of racemic 7a-c using "Chiralscreen ® OH" enzymes in 100 mM phosphate buffer (pH 7.2) at 30 °C were listed in entries 4-12. Among the nine enzymes of "Chiralscreen ® OH" tested (E001, 021, 031, 039, 041, 051, 057, 092, and 119), four enzymes such as E001, E031, E039, and E092, were found to be effective for the reduction of 7a (entries 4-7) [34]. In general, E001 and E039 gave better chemical yields than E031 and E092 (entries 4 and 6 versus entries 5 and 7). Interestingly, it was found that E001 and E039 give (S)-and (R)-forms of 76a (99% ee), respectively, with respect to the stereogenic center (position 3) in 76a. It should also be noted that the anti/syn ratios of 76a were almost 1:1, thus indicating that kinetic resolution negligibly occurred (except for E092 in entry 7). It was also found that 7b and 7c are converted into 76b and 76c by E001 and E039, respectively (entries [8][9][10][11], and the reduction of (S)-7a (90% ee) with E001 exclusively gave anti-76a (1S, 3S) as the main product (entry 12).

One-Pot Chemoenzymatic Synthesis of Optically Active 1,3-Diols 76a-c from Acetone and Benzaldehydes 6a-c
To examine the one-pot synthesis of 1,3-diols 76a-c from acetone and benzaldehydes 6a-c, simultaneous one-pot aldol-reduction reactions was attempted in a mixture containing acetone, 6a-c, Zn 2+ complexes (36 (L-ZnL 8 ) and 37 (D-ZnL 8 )), NADH, and "Chiralscreen ® OH"-E001 or E039 (Scheme 15). It was assumed that acetone, which would be generated as a byproduct of the regeneration of NADH from 2-propanol and NAD + , could be used as the substrate of aldol reaction, so that the amount of acetone could be reduced and the aldol-reduction cycle could proceed in just one reactor. However, this system did not work, because an excess amount of acetone was required to promote the aldol reaction, during which the Chiralscreen ® enzymes were inactivated. As a result, only aldol product 7 was obtained with the negligible formation of 76. Scheme 15. Scheme of initially attempted one-pot synthesis of optically active 1,3-diols 76 by chemoenzymatic synthesis in an aqueous solvent.
Then, step-wise chemoenzymatic reaction was performed, as summarized in Scheme 16. The enantioselective aldol reaction between acetone and 6 with 36 (L-ZnL 8 ) or 37 (D-ZnL 8 ) (10 mol %) was conducted in acetone/H 2 O to give 7. The reaction mixture was then diluted with phosphate buffer (100 mM, pH 7.2), and the enzyme, NAD + , and 2-propanol were added for the reduction of the aldol product 7.

Conclusions
We reviewed the design and synthesis of chiral Zn 2+ complexes comprising chiral amino acids and Zn 2+ -cyclen complexes, inspired by two classes of natural aldolases. The combined findings indicate that these Zn 2+ complexes are efficient catalysts for asymmetric aldol reactions of acetone with benzaldehydes in water-containing solvent systems and that Zn 2+ complexes that contain appropriate hydrophobic and bulky side chains give high chemical and optical yields (up to 97% yield and 96% ee). Mechanistic studies including UV/Vis titrations of ZnL with acac and X-ray crystal structure analysis of the Zn 2+ complexes indicate that these catalysts accelerate the aldol reactions via a Zn 2+ -enolate intermediate, which is generated by the cooperative functions of Zn 2+ ion of Zn 2+ complexes that activate ketone substrates as Lewis acids and the Zn 2+ -bound OH − that deprotonates the α-proton of ketones.
One of the advantages of these aldol reactions is that they are applicable to the one-pot synthesis of biorelevantly important compounds such as the optically active 1,3-diol 76 by using a combination of the enantioselective aldol reactions catalyzed by chiral Zn 2+ complexes and successive reduction by the recombinant oxidoreductase system "Chiralscreen ® OH". Typical examples include the one-pot chemoenzymatic synthesis from acetone and 6a with 36 (L-ZnL 8 ) and E001 to afford (1R, 3S)-76a in 88% yield with 96% ee. Using these methodologies, all of the possible stereoisomers of 76a-c can be obtained when the appropriate ZnL aldol catalyst and oxidoreductases are used.
We conclude that these results afford useful information concerning the design, synthesis, and mechanistic study of new artificial catalysts for catalytic enantioselective aldol reactions. Further catalyst design promises to lead, not only to the development of efficient stereoselective reactions, but also to the development of more practical and useful chemoenzymatic and biocompatible synthesis methods.