English

• Polysubstituted γ-butyrolactones are naturally occurring flavors and fragrances which are widely used in food, beverage, and cosmetics [1]. For example, 3,4-dimethyl substituted lactone, also known as Nicotiana tabacum lactone, exists in sun-cured tobacco leaves, featuring a pleasant herbaceous scent (Scheme 1) [2]. Whisky lactone is a flavor compound in alcoholic beverages that confers a typical coconut character [3]. In addition, many bioactive natural products and pharmaceuticals contain the core structures of polysubstituted γ-butyrolactones [4–10]. Hinokinin has displayed strong anti-inflammatory and antitrypanosomal activity [11–14]. Deoxyelephantopin can be used to treat hepatitis, arthritis, asthma, and cancer [5].

  Scheme 1

  

  Scheme 1.  Representative flavors, drugs, and bioactive natural products containing γ-butyrolactone-based structures.

  DownLoad: Full-Size Img PowerPoint

  It is well known that the specific activity of these chiral compounds, e.g., odor perception and bioactivity, is closely related to their stereo-configuration. Therefore, substantial effort has been devoted to the development of synthetic strategies for these α,β-disubstituted γ-butyrolactones. So far, carbene, transition metal, and organo-catalyzed reactions have been reported in the literature [15]. Typical examples include N-heterocyclic carbene catalyzed dynamic kinetic resolution of β-halo α-keto esters [16], a three-step Knoevenagl condensation-Michael addition-lactonization [17], and a sequential Michael-hemiacetalization-oxidation reaction based on organocatalysts [18].

  Recently, Bobek et al. reported the total synthesis of polysubstituted γ-butyrolactone lignan derivatives through oxime carbonate cyclization from δ-nitro alcohols, delivering diverse γ-butyrolactones at ambient reaction conditions [19].

  Alternatively, enzymatic pathways to these compounds have also attracted attention. Chiral γ-butyrolactones can be obtained by selective oxidation of racemic diols by alcohol dehydrogenase [20]. An impressive method uses enoate (or ene) reductases (ERs) to reduce the carbon-carbon double bond of α,β-unsaturated γ-ketoesters, followed by a stereoselective reduction of corresponding γ-ketoesters with carbonyl reductases (CRs). Classen et al. reported the cascade synthesis of β-methyl-substituted γ-valerolactone via this method [21], with an ee of 98%−99% and a maximum yield of 90%, but the product is limited to methyl-substituted short-chain lactones. Brenna et al. employed this route to synthesize 4a with high stereoselectivity using Old Yellow Enzymes (OYEs, such as OYE1, OYE2, and OYE3 combined with alcohol dehydrogenases [22]. Kumru et al. extended this method to prepare whisky and cognac lactones with moderate to good yields and very good diastereoselectivities [23].

  Although the cascade enzymatic method has demonstrated its feasibility in the synthesis of polysubstituted γ-butyrolactones, the low activity of enzymes and phase separation in batch reaction led to low overall product yields and long reaction time. In particular, the second step of the cascade greatly limits the reaction time due to the imbalance in the activity of the two reactions (Scheme 2a). A key opportunity of multi-enzyme reactions in microreactors is the ability to perform in vitro biosynthetic reactions [24]. A number of continuous flow systems were used to overcome batch incompatibility by Mattey et al., thus allowing for successful biocatalytic cascades [25]. Besides, Gruber et al. provide an overview of recent examples of cascaded microreactors [26]. Therefore, to address these issues, we presented a cascade continuous-flow approach for the synthesis of polysubstituted γ-butyrolactones in this study. We discovered a new ER, SsER from Swingsia samuiensi, through gene mining, and combined it with SsCR (Protein Data Bank: 5GMO) [27]. Directed evolution of these two enzymes was performed to enhance their catalytic performance. In particular, we showed that the two reactions can be regulated in two 3D microfluidic reactors by controlling the flow rates of substrates and enzymes, significantly reducing the reaction time and enhancing the total reaction efficiency (Scheme 2b). The detailed structures and sizes of the two microreactors are shown in Fig. S1 (Supporting information).

  Scheme 2

  

  Scheme 2.  Continuous-flow synthesis of chiral polysubstituted γ-butyrolactones via a dual-enzyme cascade catalysis.

  DownLoad: Full-Size Img PowerPoint

  First, enzyme screening was performed with α,β-unsaturated γ-ketoesters (E)−1a and (E)−1b (Fig. 1). The protein sequences of OYE2 [22,28] and NCr_zm [29] were used as the templates to carry out pBLAST searching in UniProt. A library of 13 gene sequences was obtained (sequence S1 in Supporting information), and 7 genes were expressed successfully into soluble recombinant enzymes in Escherichia coli (E. Coli) BL21 (DE3) (Table S1 and Fig. S2 in Supporting information). The specific activity and stereoselectivity of these ERs were assayed against (E)−1a and (E)−1b (Fig. 1 and Table S2 in Supporting information). Among them, SsER (from Swingsia samuiensi) displayed the highest activity of 6.1 U/mg and good (S)-stereoselectivity (93% ee) toward (E)−1a. SsER also exhibited higher activity than the templates OYE2 (1.5 U/mg) and NCr_zm (0.6 U/mg) toward (E)−1a, and comparable stereoselectivity with OYE2 (98% ee) and NCr_zm (93% ee). For (E)−1b, SsER showed higher activity (6.4 U/mg) and excellent (R)-stereoselectivity (99% ee) compared with OYE2 (2.1 U/mg, 98% ee) and NCr_zm (1.4 U/mg, 99% ee). Thus, SsER was chosen for the first ene reduction in the following studies.

  Figure 1

  

  Figure 1.  Specific activity and stereoselectivity of ERs toward (E)−1a and (E)−1b.

  DownLoad: Full-Size Img PowerPoint

  In addition, screening of the optimal CR with high conversion and enantioselectivity toward β-substituted γ-ketoesters was performed with (S)−2a and (R)−2b (Fig. 2). In previous studies, three CRs, including SsCR, SmCR_M5, and SmCR_V4 have been demonstrated to show high activity and stereoselectivity toward γ-ketoacids and γ-ketoester without β-substituted groups [27,30,31]. In this assay, SsCR (Fig. S3 in Supporting information) displayed the highest activity (1.3 and 0.6 U/mg) toward (S)−2a and (R)−2b, respectively. The stereoselectivity was also excellent ((S)−99% ee). However, the activity of SmCR_M5 and SmCR_V4 toward these two substrates was much weaker (Tables S3 and S4 in Supporting information). Therefore, SsCR was selected for further studies. The detailed preparation methods of the α,β-unsaturated γ-ketoesters can be found in the supporting information (Figs. S4 and S5 in Supporting information).

  Figure 2

  

  Figure 2.  Specific activity and stereoselectivity of CRs toward (S)−2a and (R)−2b.

  DownLoad: Full-Size Img PowerPoint

  With the two enzymes in hand, we proceeded to construct an efficient synthetic route to polysubstituted γ-butyrolactones. To maximize the power of the continuous-flow chemistry, the catalytic performance of SsER and SsCR needs further improvement via directed evolution. AlphaFold2 was used to build the three-dimensional protein structure models of SsER and its variants (https://alphafold.com/). (E)−1a was docked into the active pocket of SsER (Figs. 3a and b), and amino acid (a.a.) residues within 8 Å around the carbonyl oxygen of (E)−1a were selected for site-directed saturation mutation using NNK codons (Table S5 and sequence S2 in Supporting information).

  Figure 3

  

  Figure 3.  Directed evolution of SsER. Molecular docking of (E)−1a into the active sites of SsER (a) and SsER-A314S (b). (c) The specific activity of SsER and variants toward (E)−1a and (E)−1b. (d) The specific activity of SsER and SsER-A314S toward different substrates. The time-course conversion of 8 mmol/L (E)−1a (e) and 8 mmol/L (E)−1b (f) by SsER and SsER-A314S. Conditions: (E)−1a and (E)−1b (8 mmol/L), NADP⁺ (0.2 mmol/L), glucose (1.5 equiv.), lyophilized BmGDH (10 mg), purified ERs (0.5 mg for (E)−1a) and 0.1 mg for ((E)−1b), 5% DMSO in 5 mL sodium phosphate buffer (100 mmol/L, pH 6.5), 35 ℃.

  DownLoad: Full-Size Img PowerPoint

  Finally, variants SsER-A314S and A314S-F269L were obtained through plate screening. As shown in Table S6 (Supporting information), the specific activity of SsER-A314S-F269L toward (E)−1a and (E)−1b reached 11.6 and 8.0 U/mg, respectively. The specific activity of SsER-A314S toward (E)−1a and (E)−1b were enhanced to 10.4 U/mg (Fig. 3c). In addition, for (E)−1a, the kinetic analysis showed that the K_m of SsER-A314S did not change (15 µmol/L) compared with that of SsER, but the k_cat increased (Table S7, Figs. S6 and S7 in Supporting information). Therefore, the total catalytic efficiency of SsER-A314S (k_cat/K_m) is 1.4-fold higher than that of SsER. For (E)−1b, the K_m of SsER-A314S was reduced from 18 µmol/L to 5.7 µmol/L. Its k_cat was also increased, resulting in a 4.5-fold increase in the catalytic efficiency. Moreover, the activity of SsER-A314S toward a variety of substrates was all improved relative to SsER (Fig. 3d). SsER-A314S showed relatively high activity for α,β-saturated γ-ketoesters in the E-configuration, while it showed lower activity for the Z-configuration. Molecular docking of (E)−1a with SsER-A314S showed that the hydrogen bond distance formed between the ester group of the substrate and the Tyr136 residue of SsER was shortened from 2.2 Å to 1.8 Å after mutation (Figs. 3a and b). Therefore, the decreased distances involved in the hydride attack may contribute to the enhanced k_cat. In addition, the substrate was well stabilized during the reaction, which is also favorable for achieving smaller K_m. The total effect resulted in the enhanced catalytic efficiency of SsER-A314S.

  Subsequently, the time-course conversion of (E)−1a and (E)−1b catalyzed by SsER and SsER-A314S were determined, respectively. It can be seen that the rates of reactions catalyzed by SsER-A314S have been significantly improved compared with the wild type SsER (Figs. 3e and f). SsER-A314S achieved 100% conversion of 8 mmol/L (E)−1a and (E)−1b within 40 min and 6 h, respectively. Finally, we prepared a variety of α- and β-substituted chiral γ-ketoesters with SsER-A314S (Fig. 4). SsER-A314S showed good to excellent stereoselectivity (73%−99% ee) and high yield toward alkyl substituted α- and β-unsaturated γ-ketoesters ((E)−1a-(E)−1g)). In addition, SsER-A314S can also accept the α-aromatic substituted substrate (3f) and showed moderate enantioselectivity (59%) and isolated yield (74%). In addition, the half-lives of SsER-A314S were 577 h, 347 h, 90 h, and 39 min at 30 ℃, 35 ℃, 40 ℃ and 50 ℃, respectively, which were up to 4-fold higher than those of SsER, indicating that the variant displayed improved thermostability (Fig. S8 in Supporting information).

  Figure 4

  

  Figure 4.  Substrate scope. Chiral α- or β-substituted γ-ketoesters were synthesized by the variant SsER-A314S. Isolated yield and ee were determined by ¹H NMR and chiral GC. General procedure unless otherwise stated: 1a-1g (50 mmol/L), 20 mL sodium phosphate buffer (pH 6.5, 100 mmol/L), NADP⁺ (0.2 mmol/L), lyophilized E. coli/SsER-A314S (10 mg/mL), lyophilized BmGDH (2 mg/mL) and glucose (75 mmol/L) at 35 ℃ for 12 h.

  DownLoad: Full-Size Img PowerPoint

  Similarly, SsCR was engineered to enhance its catalytic activity toward (R)−2b by engineering its active pocket (Figs. 5a and b, sequence S3 in Supporting information). First, the specific activity and k_cat/K_m of the variant SsCR-V1 toward (R)−2b were 8.1 and 10 times higher than those of SsCR, respectively (Tables S8 and S9, and Fig. S9 in Supporting information). However, it displayed a slight substrate inhibition. The second variant, SsCR-V2, did not reduce substrate inhibition. According to a previous report [32], we introduced the mutation L211H into V2. The final variant V3 successfully eliminated the substrate inhibition and its specific activity against (R)−2b reached 1.3 U/mg. Moreover, the activity of SsCR-V3 toward various substrates was enhanced compared with the wild-type SsCR (Fig. 5c). In particular, V3 achieved 41 U/mg toward (R)−2f, which was 13.7-fold higher than that of SsCR. Molecular docking of (R)−2b with SsCR-V3 showed that the distance from the amino group of Ser125 to the carbonyl oxygen atom of the carboxyl group of the (R)−2b was shortened after mutation from 3.1 Å for SsCR to 2.7 Å (Figs. 5a and b). Therefore, the decreased distances involved in the hydride attack and the proton transfer may contribute to enhancing catalytic efficiency. In addition, the half-lives of SsCR-V3 were 238 and 36 h at 30 ℃ and 35 ℃, respectively, which were enhanced 3.2- and 24-fold compared with those of SsCR, indicating that the engineered enzyme had higher thermostability (Fig. S10 in Supporting information).

  Figure 5

  

  Figure 5.  Directed evolution of SsCR. Molecular docking of (R)−2b into the active sites of SsCR (a) and SsCR-V3 (b). (c) The specific activity of SsCR and V3 toward different substrates. (d) The time-course conversion of 20 mmol/L (R)−2b by SsCR, V1, and V3. Reaction condition: (R)−2b (20 mmol/L), NADP⁺ (0.2 mmol/L), Glucose (1.5 equiv.), lyophilized BmGDH (10 mg), CRs (0.4 mg), 5% DMSO in 2 mL sodium phosphate buffer (100 mmol/L, pH 6.5), 30 ℃.

  DownLoad: Full-Size Img PowerPoint

  To further verify the improvement in catalytic performance, the conversion of (R)−2b by SsCR and its variants (V1 and V3) was performed. The reaction rate was significantly enhanced after evolution. SsCR-V3 reached 98% conversion within 3 h, while SsCR converted only 64% (R)−2b at the same time (Fig. 5d). These results indicated that V3 was suitable for the asymmetric reduction of γ-ketoesters.

  With SsER-A314S and SsCR-V3 in hand, we performed a one-pot two-step sequential reaction to synthesize Nicotiana tabacum lactone ((3S,4S)−4a) and whisky lactone ((3R,4S)−4b) in a preparative scale (Table 1). For (3S,4S)−4a, the first ene reduction took 6 h to reach substrate conversion > 99%, and then SsCR-V3 was added into the reaction to achieve 90% conversion within 20 h. The total reaction time was 26 h, generating (3S,4S)−4a at 56% yield with excellent optical purity (99% ee) and good diastereoselectivity (87% de). For (3R,4S)−4b, SsER-A314S also converted the substrate (E)−1a > 99% within 6 h in the first step, however, SsCR-V3 catalyzed the carbonyl reduction even slower, reaching 87% conversion with 99% ee and 98% de after 36 h. The total reaction time reached 42 h and the final yield was 53%.

  Table 1

  

  Table 1.  One-pot synthesis of optically pure γ-butyrolactones.^a

  DownLoad: CSV

  These results showed that the second carbonyl reduction is the rate-limiting step, which is consistent with the previous report [23]. However, the two steps cannot be combined in a single-step one-pot reaction because SsCR-V3 can also catalyzes the α,β-unsaturated γ-ketoesters, generating allylic alcohols which are not accepted by SsER-A314S. Furthermore, the reaction's high substrate loading (50 mmol/L) was essential for obtaining scaled-up preparation. However, phase separation was observed because of the low substrate solubility in water, which also limited the overall reaction rate.

  The continuous-flow process carried out automatically in a microreactor can improve biocatalysis performance, making large-scale bioproduction more cost-effective due to significantly reduced reaction volume, improved mass transfer, shorter reaction time, and improved space-time yield [33]. To further increase the overall reaction rate, we performed the preparative reaction in two 3D microfluidic reactors in a continuous-flow manner (Fig. 6). In the first microfluidic reactor, the double bond reduction proceeded with > 99% conversion in 30 min, thanks to the solved substrate (5 mmol/L) and the high enzyme activity of SsER-A314S. Then, the intermediate α,β-substituted γ-ketoesters were injected into the second microreactor together with SsCR-V3 and converted into the final products. Considering the relatively lower activity of SsCR-V3, the second reaction was cycled a couple of times to improve the total conversion. As expected, the second reaction took 26 min to reach 93% conversion for (3S,4S)−4a with the same stereoselectivity (99% ee and 87% de). The total reaction time was reduced to 56 min (Table 2), and the space-time yield reached approximately 5.0 mmol L⁻¹ h⁻¹, which was 3 times higher than that of the batch reaction (1.7 mmol L⁻¹ h⁻¹). For (3R,4S)−4b, the total reaction time was reduced to 40 min without any appreciable loss of optical purity, giving a space-time yield of 7.4 mmol L⁻¹ h⁻¹ which was 7.4 times higher than that of the batch reaction (1.0 mmol L⁻¹ h⁻¹). The reduction in reaction time can be attributed to the elimination of phase separation and enhanced mixing of reactants in the 3D microfluidic reactors. Although the continuous-flow was run at a low substrate concentration, the automatic operation and independent control over the flow rates allow for the parallel synthesis of γ-butyrolactones. In particular, future improvements in enzyme catalytic activity will further enhance the productivity of the continuous-flow process, while the batch reaction benefits little from this because the serious phase separation remains at high substrate loading.

  Figure 6

  

  Figure 6.  Continuous-flow synthesis of chiral polysubstituted γ-butyrolactones via a dual-enzyme cascade.

  DownLoad: Full-Size Img PowerPoint

  Table 2

  

  Table 2.  Continuous-flow synthesis of γ-butyrolactones.^a

  DownLoad: CSV

  In conclusion, the improvement of efficiency and productivity of cascade enzymatic processes relies on the combination of biocatalyst engineering and process optimization. Herein, we engineered an ene reductase SsER and a carbonyl reductase SsCR, which displayed improved activity and thermostability compared with the wild-type enzymes. We proved that the optically pure Nicotiana tabacum lactone ((3S,4S)−4a) and whisky lactone ((3R,4S)−4b) can be synthesized efficiently from α,β-unsaturated ketoesters in the continuous-flow process, delivering products with 3 and 7.4 times higher space-time yields compared with the one-pot, two-step batch reaction. The integration of cascade enzyme catalysis into cutting-edge continuous microfluidic technology will benefit the general in vitro multi-enzyme catalytic systems in the future.

  Declaration of competing interest

  The authors declare that they have no known competitive financial interests or personal relationships that may affect the work of this report.

  Acknowledgments

  This work was financially sponsored by the National Key Research and Development Program of China (No. 2021YFC2102804) and the National Natural Science Foundation of China (No. 22078096).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108896.

  
  
• 1. [1]

     A. Świzdor, A. Panek, N. Milecka-Tronina, et al., Int. J. Mol. Sci. 13 (2012) 16514–16543. doi: 10.3390/ijms131216514

  2. [2]

     B. Kimland, J. Aasen, S.O.A. lmqvist, P. Arpino, C.R. Enzell, Phytochemistry 12 (1973) 835–847. doi: 10.1016/0031-9422(73)80688-0

  3. [3]

     E. Masson, R. Baumes, C.L. Guernev, J.L. Puech, J. Agric, Food Chem. 48 (2000) 4306–4309. doi: 10.1021/jf0002950

  4. [4]

     V.B. Mathema, Y.S. Koh, B.C. Thakuri, M. Sillanpää, Inflammation 35 (2012) 560–565. doi: 10.1007/s10753-011-9346-0

  5. [5]

     F.A. Kabeer, R. Prathapan, Pharmacologia 5 (2014) 272–285. doi: 10.5567/pharmacologia.2014.272.285

  6. [6]

     A.A. Beeran, N. Maliyakkal, M.C. Rao, N. Udupa, Pharmacogn. Mag. 11 (2015) 257–268. doi: 10.4103/0973-1296.153077

  7. [7]

     C. Li, H. Wu, Y. Yang, J. Liu, Z. Chen, Oncol. Lett. 15 (2018) 9673–9680.

  8. [8]

     C.L. Li, H.Z. Wu, Y.P. Huang, et al., Pharmacotherapy 6 (2008) 401–409.

  9. [9]

     M. Sabanero, L. Quijano, T. Rios, R. Trejo, Planta Med. 61 (1995) 185–186. doi: 10.1055/s-2006-958046

  10. [10]

      X.G. Lu, M. L, J.L. Wei, et al., Biomed. Pharmacother. 88 (2017) 102–108. doi: 10.1016/j.biopha.2017.01.041

  11. [11]

      R.D. Silva, G.H.B.D. Souza, A.A.D. Silva, et al., Bioorg. Med. Chem. Lett. 15 (2005) 1033–1037. doi: 10.1016/j.bmcl.2004.12.035

  12. [12]

      T.C. Lima, R. Lucarini, A.C. Volpe, et al., Med. Chem. Lett. 27 (2017) 176–179. doi: 10.1016/j.bmcl.2016.11.081

  13. [13]

      V.A.D. Souza, R.D. Silva, A.C. Pereira, et al., Bioorg. Med. Chem. Lett. 15 (2005) 303–307. doi: 10.1016/j.bmcl.2004.10.079

  14. [14]

      J. Saraiva, C. Vega, M. Rolon, et al., Parasitol. Res. 100 (2007) 791–795. doi: 10.1007/s00436-006-0327-4

  15. [15]

      J. Hur, J. Jang, J. Sim, Int. J. Mol. Sci. 22 (2021) 2769. doi: 10.3390/ijms22052769

  16. [16]

      C.G. Goodman, M.M. Walker, J.S. Johnson, J. Am. Chem. Soc. 137 (2015) 122–125. doi: 10.1021/ja511701j

  17. [17]

      T.M. Khopade, A.D. Sonawane, J.S. Arora, R.G. Bhat, Adv. Synth. Catal. 35 (2017) 3905–3910.

  18. [18]

      P. Mahto, N.K. Rana, K. Shukla, et al., Org. Lett. 21 (2019) 5962–5966. doi: 10.1021/acs.orglett.9b02094

  19. [19]

      K.B. Bobek, N.S. Ezzat, B.S. Jones, et al., Org. Lett. 25 (2023) 31–36. doi: 10.1021/acs.orglett.2c03727

  20. [20]

      F. Boratyński, K. Dancewicz, M. Paprocka, B. Gabryś, C. Wawrzeńczyk, PLoS One 11 (2016) e0146160. doi: 10.1371/journal.pone.0146160

  21. [21]

      T. Classen, M. Korpak, M. Schölzel, J. Pietruszka, ACS Catal. 4 (2014) 1321–1331. doi: 10.1021/cs5000262

  22. [22]

      E. Brenna, F.G. Gatti, D. Monti, et al., J. Mol. Catal. B 114 (2015) 77–85. doi: 10.1016/j.molcatb.2014.12.011

  23. [23]

      C. Kumru, T. Classen, J. Pietruszka, ChemCatChem 10 (2018) 4917–4926. doi: 10.1002/cctc.201801040

  24. [24]

      M.Y. Lee, A. Srinivasan, B. Ku, J.S. Dordick, Biotechnol. Bioeng. 83 (2003) 20–28. doi: 10.1002/bit.10642

  25. [25]

      A.P. Mattey, G.J. Ford, J. Citoler, Angew. Chem. Int. Ed. 60 (2021) 18660–18665. doi: 10.1002/anie.202103805

  26. [26]

      P. Gruber, M.P.C. Marques, B. O'Sullivan, et al., Biotechnol. J. 12 (2017) 1700030. doi: 10.1002/biot.201700030

  27. [27]

      Y.Q. Zhang, T.T. Feng, Y.F. Cao, et al., ACS Catal. 11 (2021) 10487–10493. doi: 10.1021/acscatal.1c02221

  28. [28]

      A. Goffeau, B.G. Barrell, H. Bussey, et al., Science 274 (1996) 546–567. doi: 10.1126/science.274.5287.546

  29. [29]

      M. Hall, C. Stueckler, B. Hauer, et al., Eur. J. Org. Chem. 2008 (2008) 1511–1516. doi: 10.1002/ejoc.200701208

  30. [30]

      T. Wang, X.Y. Zhang, Y.C. Zheng, Y.P. Bai, Chem. Commun. 57 (2021) 10584–10587. doi: 10.1039/D1CC04542C

  31. [31]

      M. Chen, X.Y. Zhang, C.G. Xing, et al., ChemCatChem 11 (2019) 2600. doi: 10.1002/cctc.201900382

  32. [32]

      Y.P. Shang, Qi. Chen, A.T. Li, et al., J. Biotech. 308 (2020) 141–147. doi: 10.1016/j.jbiotec.2019.12.011

  33. [33]

      Y. Wu, W.Q. Chen, Y.Q. Zhao, et al., Chin. Chem. Lett. 26 (2015) 334–338. doi: 10.1016/j.cclet.2014.11.013

• 

  Scheme 1  Representative flavors, drugs, and bioactive natural products containing γ-butyrolactone-based structures.

  下载: 全尺寸图片 幻灯片
  

  Scheme 2  Continuous-flow synthesis of chiral polysubstituted γ-butyrolactones via a dual-enzyme cascade catalysis.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Specific activity and stereoselectivity of ERs toward (E)−1a and (E)−1b.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Specific activity and stereoselectivity of CRs toward (S)−2a and (R)−2b.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Directed evolution of SsER. Molecular docking of (E)−1a into the active sites of SsER (a) and SsER-A314S (b). (c) The specific activity of SsER and variants toward (E)−1a and (E)−1b. (d) The specific activity of SsER and SsER-A314S toward different substrates. The time-course conversion of 8 mmol/L (E)−1a (e) and 8 mmol/L (E)−1b (f) by SsER and SsER-A314S. Conditions: (E)−1a and (E)−1b (8 mmol/L), NADP⁺ (0.2 mmol/L), glucose (1.5 equiv.), lyophilized BmGDH (10 mg), purified ERs (0.5 mg for (E)−1a) and 0.1 mg for ((E)−1b), 5% DMSO in 5 mL sodium phosphate buffer (100 mmol/L, pH 6.5), 35 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Substrate scope. Chiral α- or β-substituted γ-ketoesters were synthesized by the variant SsER-A314S. Isolated yield and ee were determined by ¹H NMR and chiral GC. General procedure unless otherwise stated: 1a-1g (50 mmol/L), 20 mL sodium phosphate buffer (pH 6.5, 100 mmol/L), NADP⁺ (0.2 mmol/L), lyophilized E. coli/SsER-A314S (10 mg/mL), lyophilized BmGDH (2 mg/mL) and glucose (75 mmol/L) at 35 ℃ for 12 h.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Directed evolution of SsCR. Molecular docking of (R)−2b into the active sites of SsCR (a) and SsCR-V3 (b). (c) The specific activity of SsCR and V3 toward different substrates. (d) The time-course conversion of 20 mmol/L (R)−2b by SsCR, V1, and V3. Reaction condition: (R)−2b (20 mmol/L), NADP⁺ (0.2 mmol/L), Glucose (1.5 equiv.), lyophilized BmGDH (10 mg), CRs (0.4 mg), 5% DMSO in 2 mL sodium phosphate buffer (100 mmol/L, pH 6.5), 30 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Continuous-flow synthesis of chiral polysubstituted γ-butyrolactones via a dual-enzyme cascade.

  下载: 全尺寸图片 幻灯片

  Table 1.  One-pot synthesis of optically pure γ-butyrolactones.^a

  下载: 导出CSV

  Table 2.  Continuous-flow synthesis of γ-butyrolactones.^a

  下载: 导出CSV
•