If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Beijing National Laboratory for Molecular Sciences, Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education, Department of Chemical Biology, College of Chemistry and Molecular Engineering, Synthetic and Functional Biomolecules Center, Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, China
Beijing National Laboratory for Molecular Sciences, Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education, Department of Chemical Biology, College of Chemistry and Molecular Engineering, Synthetic and Functional Biomolecules Center, Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, ChinaAcademy for Advanced Interdisciplinary Studies, Peking University, Beijing, China
1 Professor Xiaoguang Lei, the corresponding author on this paper is a member of the advisory board but this author had no involvement in the peer review process used to assess this work submitted to Tetrahedron Chem. This paper was assessed and the corresponding peer review managed by Dr Chaosheng Luo, a scientific editor working on Tetrahedron Chem.
Xiaoguang Lei
Correspondence
Corresponding author. Beijing National Laboratory for Molecular Sciences, Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education, Department of Chemical Biology, College of Chemistry and Molecular Engineering, Synthetic and Functional Biomolecules Center, Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, China.
1 Professor Xiaoguang Lei, the corresponding author on this paper is a member of the advisory board but this author had no involvement in the peer review process used to assess this work submitted to Tetrahedron Chem. This paper was assessed and the corresponding peer review managed by Dr Chaosheng Luo, a scientific editor working on Tetrahedron Chem.
Affiliations
Beijing National Laboratory for Molecular Sciences, Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education, Department of Chemical Biology, College of Chemistry and Molecular Engineering, Synthetic and Functional Biomolecules Center, Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, ChinaInstitute for Cancer Research, Shenzhen Bay Laboratory, Shenzhen, 518107, China
1 Professor Xiaoguang Lei, the corresponding author on this paper is a member of the advisory board but this author had no involvement in the peer review process used to assess this work submitted to Tetrahedron Chem. This paper was assessed and the corresponding peer review managed by Dr Chaosheng Luo, a scientific editor working on Tetrahedron Chem.
The Diels-Alder (D-A) reaction is one of the most critical chemical transformations to construct C–C bonds with predictable regio- and stereo-selectivities. It has been widely used as a retrosynthetic disconnection in synthetic chemistry. Although significant advances have been made, synthetic challenges remain in how to precisely control the stereochemistry of intermolecular Diels-Alder reaction. Enzymes are well known for their remarkable catalytic efficiency and selectivity compared with chemo-catalysts. Therefore, identifying and designing intermolecular Diels-Alderases that can be used in organic synthesis have received considerable attention from the synthetic community. In this review, we review all the enzymes capable of catalyzing formal intermolecular Diels-Alder reactions in natural product biosynthetic pathways, discuss their catalytic mechanisms in detail, and highlight their synthetic potential in the precise and efficient synthesis of enantiopure D-A products. We also discuss the different strategies that can be used to create new artificial Diels-Alderases, especially RNA-based Diels-Alderases.
], it was not fully described until Prof. Otto Diels and his student, Dr. Kurt Alder, identified the reaction products of cyclopentadiene (1) and quinone (2) to be the cycloadducts 3 and 4 in 1928 (Scheme 1) [
]. As Diels and Alder prophetically envisioned in the original paper, Diels-Alder reactions and hetero-Diels-Alder reactions have been extensively exploited by synthetic chemists to form C–C and C-heteroatom bonds in natural product total synthesis [
Recent advances in applying Diels-Alder reactions involving o-quinodimethanes, aza-o-quinone methides and o-quinone methides in natural product total synthesis.
The greatest challenge for application of the Diels-Alder reaction in organic synthesis is how to precisely control its stereoselectivity. Since the first example of an asymmetric Diels-Alder reaction catalyzed by chiral alkoxyaluminium dichloride was reported in 1979 [
]. Notably, the original discovery of organocatalysis in asymmetric Diels-Alder reaction was recognized by the Nobel Prize in Chemistry in 2021. Besides these traditional chemical approaches to promote Diels-Alder reactions with high stereoselectivity, the discovery and design of enzymes catalyzing Diels-Alder reactions have received increasing attention from synthetic chemists. In this review, we survey the biocatalysts that promote the intermolecular Diels-Alder reactions and highlight the synthetic power of these biocatalysts for the chemo-enzymatic synthesis of D-A products.
2. Natural Diels-Alderases
With the rapid development of omics technologies and bioinformatics, more and more Diels-Alderases (D-Aases) have been discovered since the 2000s [
]. Unlike chemists who often manipulate intermolecular Diels-Alder reactions as a retrosynthetic disconnection, it seems that nature tends to harness the intramolecular transformation to increase the structural complexity in the late-stages of biosynthetic pathways [
Solanapyrone synthase, a possible Diels-Alderase and iterative type I polyketide synthase encoded in a biosynthetic gene cluster from Alternaria solani.
Biosynthesis of versipelostatin: identification of an enzyme-catalyzed [4+2]-cycloaddition required for macrocyclization of spirotetronate-containing polyketides.
] scaffolds. Moreover, D-Aases that catalyze intramolecular hetero-Diels-Alder reactions have also been discovered in the biosynthetic pathways of prenylated indole alkaloids (e.g., PhqE [
]). Compared with intramolecular D-Aases, biocatalysts for intermolecular reactions are more synthetically attractive but rare in nature. So far, only five different families of enzymes that catalyze intermolecular (hetero)Diels-Alder reactions have been identified in natural product biosynthetic pathways. Despite the controversies regarding the definition of Diels-Alderases, we call them intermolecular D-Aases based on the formal Diels-Alder reaction they catalyze regardless of their catalytic mechanisms.
2.1 Macrophomate synthase (MPS)
Macrophomate synthase (MPS), which catalyzes multiple transformations from 5-acety-4-methoxy-6-methyl-2-pyrone (5) and oxaloacetate (6) to macrophomate (7), was identified by Watanabe and co-workers in the 2000s [
Characterization, sequence, and expression in Escherichia coli of the novel enzyme catalyzing unusual multistep transformation of 2-pyrones to benzoates.
]. Based on feeding experiments and enzymatic assays using mimics of 2-pyrone (5) and D-A product 8, Oikawa and co-workers have proposed a plausible intermolecular Diels–Alder reaction between 2-pyrone (5) and pyruvate (9), as well as decarboxylation and dehydration reactions in the biosynthesis of macrophomate (Scheme 2) [
]. To further understand its catalytic mechanism, Ose and co-workers have solved the 1.70 Å resolution crystal structure of MPS with pyruvate (9) and Mg2+ bound in the catalytic pocket [
]. They proposed that the Lewis acidity of the magnesium ion promotes decarboxylation of oxalacetate (6) to form the enolate anion (9), which is stabilized by Mg2+ as well as several nearby residues (G210, E211, and R101). This MPS-9 complex has enough space in the catalytic pocket to accommodate the diene 2-pyrone (5), which can be fixed in place through two hydrogen bonds between the carbonyl oxygen of 2-pyrone (5) and Arg 101, and the C5-acyl oxygen and Tyr 169. In this way, the direction of approach and orientation of the 2-pyrone are well controlled by the enzyme, defining the absolute configuration of the D-A product 8, which then undergoes decarboxylation and dehydration reactions to give macrophomate (7) as the final product (Scheme 2).
Scheme 2The synthesis of macrophomate with a detailed illustration of individual reaction steps.
On the other hand, when Jorgensen and co-workers modeled the enzyme-catalyzed [4 + 2] cycloaddition using the crystal structure of MPS via quantum mechanical/molecular mechanical (QM/MM) computations combined with Monte Carlo simulations and free-energy perturbation (FEP) calculations [
] (Scheme 2). Later, Hilvert's laboratory demonstrated that macrophomate synthase, structurally related to 2-dehydro-3-deoxygalactarate aldolase, also functionally serves as an aldolase to form β-hydroxy phenylpropionate from oxaloacetate and benzaldehyde, which further supports the stepwise Michael-aldol mechanism [
Riboflavin synthase is another possible intermolecular Diels-Alderase that catalyzes the dismutation of two molecules of 6,7-dimethyl-8-ribityllumazine (12) to form riboflavin (13) via the pentacyclic intermediate compound Q (14) [
]. In the proposed mechanism, one molecule of 6,7-dimethyl-8-ribityllumazine (12) firstly tautomerizes to form 15, which then undergoes intermolecular hydride transfer from itself to another molecule of 12, yielding diene 16 and dienophile 17. Next, regioselective Diels-Alder reaction between 16 and 17 generates intermediate 14, which can be further transformed into riboflavin (13) and 18 after two elimination steps (Scheme 3) [
]. However, this proposed Diels-Alder mechanism is not supported by density functional theory (DFT) calculations from Houk and co-workers. They found that, instead of the pathway, mentioned above, the nucleophilic addition mechanism is the lowest energy pathway yielding riboflavin [
Theoretical exploration of the mechanism of riboflavin formation from 6,7-dimethyl-8-ribityllumazine: nucleophilic catalysis, hydride transfer, hydrogen atom transfer, or nucleophilic addition?.
Sorbicillinoid natural products are a family of hexaketide metabolites isolated from fungi with complex and highly oxygenated skeletons and intriguing biological activities [
]. Retrosynthetically, they are thought to originate from the highly reactive intermediate sorbicilinol (19) via epoxidation (such as epoxysorbicilinol (20)), Diels-Alder reaction (such as bisorbicilinol (21)), or Michael addition (such as bisvertinolone (22)(Scheme 4). The biosynthetic gene cluster of sorbicilinoid in the fungus Penicillium chrysogenum E01-10/3 was discovered by Cox et al. in 2014, which encodes seven putative open-reading frames (ORF), including a FAD-dependent monooxygenase (SorC), an NR-iPKS (SorB), a highly-reducing iPKS (SorA), and a second FAD-dependent oxidoreductase (SorD) [
]. Based on bioinformatic analysis, sorbicillin (23) was produced by two PKSs, i.e., SorA and SorB, via chain elongation, methylation, reduction, and Knoevenagel cyclization from acetyl CoA and S-adenosyl methionine (SAM) as shown in Scheme 4. In vitro assays demonstrated that SorC catalyzed the oxidative dearomatization of sorbicinlin (23) to give sorbicilinol (19), which is highly reactive and undergoes spontaneous dimerization to generate 21 via an intermolecular Diels-Alder reaction [
]. Besides the evidence mentioned above from chemical synthesis, Gulder et al. have also shown that sorbicilinol (19), which was generated in vitro by incubation of sorbicillin (23) with SorC, would undergo spontaneous dimerization to form bisorbicilinol (21) and bisvertinolone (22) [
]. Taken together, these synthetic results suggest that SorA, B, and C are sufficient for the fungi to produce dimeric sorbicillinoids from sorbicilinol (19), with the role of SorD remaining functionally elusive.
Scheme 4The proposed biosynthetic pathway of sorbicillinoid natural products in T. reesei QM6a.
To elucidate the precise role of SorD in sorbicillinoids biosynthesis, Skellam and co-workers heterologously reconstituted the sorbicillinoid biosynthetic pathway in Aspergillus oryzae NSAR1 from Trichoderma reesei QM6a. cDNA templates [
]. As expected, A. oryzae transformed with sorAB produced sorbicillin (23), and adding sorC to the heterologous expression system resulted in sorbicilinol (19) production. However, in contrast to the previous chemo-enzymatic synthesis of dimeric sorbicillinoids, no formation of any dimeric sorbicillinoids was observed in any of the + sorABC transformants, indicating that an in vivo catalyst is indispensable for the dimerization under physiological conditions. Introducing the sorD gene to the + sorABC transformants resulted in the production of dimerized sorbicillinoids such as bisorbicilinol (21), bisvertinol (24), and epoxysorbicilinol (20), suggesting that the flavin-dependent monooxygenase SorD in T. reesei QM6a might be a multi-functional enzyme that catalyzes intermolecular Diels–Alder and Michael dimerization reactions, as well as the epoxidation of sorbicillinol (19) in sorbicillinoid biosynthesis (Scheme 4).
2.4 EupfF
An intermolecular hetero-Diels-Alder reaction (HDA) is proposed to play an important role in the biosynthesis of natural products containing a tetrahydropyran moiety, as evidenced by numerous successful biomimetic total syntheses of natural products such as (−)-xyloketal [
] (Fig. 1). These biomimetic syntheses typically involve in situ generation of ortho-quinone methide or an α,β-unsaturated ketone via enzymatic/chemical benzylic oxidation or dehydration from alcohols, followed by a catalyst-free intermolecular hetero-Diels-Alder reaction.
Fig. 1Hetero-Diels-Alder reaction in total biomimetic syntheses of natural products.
In the biosynthetic study of a tropolone-sesquiterpene family of meroterpenoids, Hu et al. reported the first D-Aase that catalyzes an intermolecular hetero-Diels-Alder reaction [
]. In this case, the reactive tropolone o-quinone methide (25) was generated from stipitaldehyde (27) via EupfE-catalyzed reduction, followed by dehydration of 26. The dehydration of 26 was spontaneous in the reaction buffer and accelerated in the presence of EupfF. When 27 was incubated with the dienophile 28 without EupfF, the spontaneous hetero-Diels-Alder reaction gave two D-A products 29 and 30 in a 1:1 ratio. In contrast, no D-A product was observed in the reaction buffer containing 27 and 31. However, neosetophomone B (32) was produced as the sole product when EupfF was added. Therefore, EupfF was proved to promote the intermolecular hetero-Diels−Alder reaction stereoselectively (Scheme 5).
Scheme 5The biosynthetic pathways of neosetophomone B and related tropolone sesquiterpenoids.
]. One homologous protein of EupfF from Acremonium strictum, AsR5, recognizes the benzylic alcohol 33 (tautomer of 34) and α-humulene (35) as substrates to give xenovulene B (36). Another homolog of EupfF from the biosynthetic pathway of pycnidione (37) in Leptobacillium sp., PycR1, catalyzes two tandem HDA reactions between 25 and 35 to yield di-tropolone meroterpenoid 38 as the sole product in a highly stereoselective way that was different from the biomimetic synthesis of pycnidinone [
Total synthesis and computational investigations of sesquiterpene-tropolones ameliorate stereochemical inconsistencies and resolve an ambiguous biosynthetic relationship.
Moraceous plants are rich sources for isolating a new family of D-A type natural products, containing more than 100 different members with diverse biological activities, including anti-diabetic [
]. These molecules originate from enzymatic intermolecular Diels-Alder reactions of a chalcone dienophile moiety and four different dehydroprenyl diene moieties [i.e., flavonoids, chalcones, stilbenes, and 2-aryl-benzofuran] [
] as shown in Fig. 2. Since these D-A products exist in enantiopure form, chiral catalyst(s) is (are) believed to be involved in their biosynthesis, as in the case of the asymmetric biomimetic total syntheses of kuwanon I [
To hunt for the novel intermolecular Diels-Alderase(s) in moraceous plants, we applied a biosynthetic intermediate probe (BIP)-based target identification strategy [
], together with activity-guided protein purification and transcriptional analysis. We identified two FAD-dependent enzymes (MaMO and MaDA) from cell callus of Morus alba as the critical enzymes in the biosynthetic pathway leading to chalcomoracin (39) [
]. Enzymatic assays suggested that MaMO functionally serves as an oxidase that produces diene 40 from moracin C (41). At the same time, MaDA endo- and enantioselectively catalyzed intermolecular Diels-Alder reaction between diene 40 and dienophile morachalcone A (42) to give the D-A product chalcomoracin (39) as shown in Scheme 7. Despite sharing high sequence identity (62%) with MaMO, MaDA shows no oxidative activity with moracin C (41), suggesting that it is a standalone D-Aase and different from previously reported intermolecular enzymes, which are multifunctional enzymes.
Scheme 7The biosynthetic pathway of chalcomoracin.
MaDA was also found to recognize different dehydroprenyl diene moieties to enantioselectively produce other types of D–A natural products with endo configurations, such as deoxyartonin I, kuwanon J, and kuwanol E (their structures are shown in Fig. 2) [
]. The structural diversity of the D-A natural products in Morus plants may thus originate from the substrate promiscuity of MaDA. However, endo D–A adducts without a prenyl group in the dienophile moiety, such as mulberrofuran C (43), or D–A natural products with exo configurations, such as mongolicin F (44), were not accessible via MaDA-catalyzed Diels-Alder reactions, suggesting that other Diels-Alderases with different substrate scope and endo/exo selectivity exist in Morus plants and await identification and characterization. Using MaDA as a query, 41 homologous proteins were identified in the genome of Morus notabilis using the protein Basic Local Alignment Search Tool (pBLAST). Guided by phylogenetic analysis, three homologous genes (MaDA-1, MaDA-2, and MaDA-3) were preferentially chosen for gene cloning, hetero-expression, and activity tests. The biochemical results demonstrate that MaDA-1 is a new endo-selective Diels-Alderase with broader dienophile scope than MaDA since it can recognize not only dienophile 42 but also 45 to produce D-A products chalcomoracin (39) and mulberrofuran C (43) (Scheme 8). More importantly, MaDA-2 and MaDA-3 were shown to be the enzymes that are responsible for the biosynthesis of mongolicin F (44) and mulberrofuran J (46) (Scheme 8), making them the first two exo-selective intermolecular Diels-Alderases from nature [
To understand how these homologous FAD-dependent enzymes control endo and exo Diels-Alder pathways in the Diels–Alder reactions, the crystal structures of MaDA (PDB ID: 6JQH) and MaDA-3 (PDB ID: 7E2V) were solved to 2.30 and 2.94 Å, respectively. Docking calculations with the computed endo transition state (TS) in the active site of MaDA, followed by molecular dynamics (MD) simulations, were performed, revealing the binding mode shown in Fig. 3A. The computational results suggest that R443 forms a hydrogen bonding interaction with the carbonyl oxygen of the dienophile and probably promotes the Diels-Alder reaction by lowering the energy of the lowest unoccupied molecular orbital (LUMO) of the dienophile. Indeed, DFT theozyme calculations suggest that this hydrogen bonding interaction decreases the gap between the highest occupied molecular orbital (HOMO) of the diene and lowest unoccupied molecular orbital (LUMO) of the dienophile by 0.8 eV, leading to a reduction of the activation free energy barrier by 1.9 kcal mol−1. Mutagenesis experiments suggest that R428 is the critical catalytic residue since the MaDA-R428A variant completely lost activity. Rotation of the diene from endo to exo resulted in a steric clash with V177, as shown in the model in Fig. 3A.
Fig. 3The proposed binding models of MaDA-TS(endo) (A) and MaDA-3-TS(exo) (B).
Similarly, we also proposed a binding model for the exo transition state with MaDA-3, as shown in Fig. 3B. No hydrogen bonding interaction was formed between R443 and the carbonyl oxygen of the dienophile. Instead, a critical cation–π interaction between R294 in the active site of MaDA-3 and the 2,4-dihydroxy phenyl substituent of the dienophile was observed, and this cation–π interaction can decrease the HOMO–LUMO gap by 0.7 eV and reduce the free energy barrier of the exo D–A reaction by 2.9 kcal mol−1 in the DFT theozyme calculations. R294 is evolutionally conserved in the two exo-selective Diels-Alderases (MaDA-2 and 3). However, for the two endo-selective Diels-Alderases (MaDA and MaDA-1), the amino acid at the 294 site is G instead of R. Moreover, R294G mutation in MaDA-3 resulted in a dramatic decrease in exo selectivity, while endo activity increased concomitantly, and vice versa. Thus, R294 appears to be an important residue contributing to exo selectivity via a cation–π interaction. Rotation of the diene from exo to endo is believed to be disfavored due to the steric clash between the diene moiety and FAD, as shown in the model in Fig. 3B.
3. Artificial Diels-Alderases
Due to the importance of Diels-Alder cycloadditions in synthetic chemistry, creating artificial enzymes promoting intermolecular Diels-Alder reactions with high selectivity has gained considerable attention. To that end, artificial protein-based Diels-Alderases have been successfully generated using the mammalian immune system [
]. Based on the mechanistic studies, it seems that natural D-Aases, antibody and de novo designed enzymes all activate the dienophiles or dienes via hydrogen bonding interactions, resulting in rate accelerations. However, a unique cation-π interaction between the natural enzyme MaDA-3 and the dienophile 42 was found to promote the enzymatic D-A reaction, which has been rarely found in antibody and de novo designed enzymes. On the other hand, metal coordination is the key interaction for artificial metalloenzymes to promote the D-A reaction, while this interaction is hardly used by natural D-Aases to promote D-A reactions.
Besides protein-based Diels-Alderases, artificial RNA motifs have also been isolated as Diels-Alderase ribozymes from combinatorial RNA libraries. Since a recent review has comprehensively summarized the different strategies to make artificial protein-based Diels-Alderases [
], we will only focus on Diels-Alderase ribozymes in this review.
In the ‘RNA world’ hypothesis, RNA is proposed to catalyze a wide variety of chemical reactions that lead to the appearance of life. This assumption has stimulated interest in identifying ribozymes with new catalytic functions using in vitro selection (SELEX) from combinational RNA libraries. Eaton and co-workers reported the first Diels-Alderase ribozyme [
]. The RNA molecules were composed of a contiguous 100-nucleotide randomized region, flanked by constant sequence segments to allow amplification and other enzymatic manipulations. Each uridine moiety in the RNA was modified with a 5-pyridylmethylcarboxamidyl group by substituting 5-pyridylmethylcarboxamid-UTP (47) for UTP (48) during transcription. The introduction of the 5-pyridylmethylcarboxamidyl group was intended to furnish additional interactions (i.e., hydrogen bonding, hydrophobic and dipolar) and provide new metal coordination sites. To allow for rapid and efficient RNA molecule selection, the RNAs were chemically linked to an acyclic diene through a flexible polyethylene glycol (PEG) linker, generating a library of RNA-PEG-diene (49) as shown in Scheme 9. Rounds of in vitro selection for Diels-Alder activity were conducted by incubating varying RNA-PEG-dienes with the biotinylated dienophile (50) in the presence of transition metals. RNA with the desired Diels-Alder activity (51) was covalently modified with biotin and then partitioned away from unreacted RNA using streptavidin binding and denaturing polyacrylamide electrophoresis. The isolated RNAs were reverse-transcribed and amplified to generate a new DNA library, which was then transcribed in vitro to RNA sequences for the next round of selection.
Scheme 9Protocol for in vitro selection of RNA Diels-Alderases from combinational RNA libraries.
After 12 rounds of in vitro selection, the RNA library with enhanced activity was cloned and bidirectionally sequenced. Eight non-clonally derived families of RNAs were identified as Diels-Alderases. Interestingly, one consensus sequence, UUCUAACGCG, was found in five of these sequences. Biochemical assays [
] were also performed to characterize these ribozymes. The Diels-Alderase activity was completely dependent on the presence of both the pyridyl-modified uridine and Cu2+, indicating that the pyridyl–Cu2+ complex likely serves as a Lewis-acid catalyst. Compared with the uncatalyzed reaction, this RNA-catalyzed Diels-Alder reaction is 800-fold faster.
Jäschke and coworkers used a similar strategy to identify unmodified RNAs that function as Diels-Alderases [
]. After 10 rounds of in vitro selection, a total of 42 different sequences were identified that accelerated the Diels-Alder reaction between a biotinylated maleimide and an RNA-tethered anthracene by up to 18,500-fold. All these ribozymes contain an unpaired GGAG sequence at the 5′-terminus. Another two consensus sequences, UGCCA (B1) and AAUACU (B2), which were found in 90% of the ribozymes, had a large bulge in their secondary structure. Below the bulge, a double-stranded helix (helix II) comprising 4 variable base pairs (bp) is formed. Above the bulge, another helix (helix III) with variable length and occasional mismatches and a loop (loop 2) are observed. A third helix (helix I) composed of at least 5 base pairs was found near helix II. The ends of helix I are closed by another loop (loop l), which has variable size (n ≥ 4) and sequence (Scheme 10). A sequence with variable nucleotides and size (14–90) is located at the 3′ end of each ribozyme. Deleting this sequence generates a common structural motif shared by most ribozymes. Interestingly, this structural motif, containing 3 consensus sequences, 3 helixes and 2 loops, showed the same catalytic activity as the untruncated parent, indicating that this structural motif contains all the elements necessary for the Diels-Alderase activity. Removal of helix I and loop 1 from this structural motif to generate a smaller oligoribonucleotide didn't affect its Diels-Alderase activity either, indicating that they are also not functionally crucial for the Diels-Alder ribozymes.
Scheme 10The procedure to identify unmodified Diels-Alderase ribozymes and their secondary structure.
The RNA-catalyzed Diels-Alder reaction between a maleimide and an RNA-tethered diene is interesting but synthetically less useful due to low atom economy. Moreover, the RNA cannot be defined as a true catalyst since it is permanently changed during the Diels-Alder reaction. However, in their further work on Diels-Alderase ribozymes, Jäschke and coworkers created ribozymes that act as true catalysts of a Diels-Alder reaction between two small, organic substrates [
]. As the previous study, a 49-mer oligoribonucleotide containing all the elements from the standard structural motif of the Diels-Alderase ribozymes was synthesized. The chemical reaction of different substituted maleimides and anthracenes was performed in the absence or presence of the 49-mer oligoribonucleotide. The results showed that 9-(hydroxylmethyl) anthracene (51) was the minimum diene and that the minimum dienophile was ethyl-substituted maleimide (52) (Fig. 4). As R2 becomes bulkier, the enzymatic Diels-Alder reaction occurs faster (Fig. 4). Kinetic characterization of the ribozyme was also performed using diene 53 and dienophile 54. The kcat was determined to be 21 min−1 (Table 1) and the Michaelis constants (KM) for diene 53 and dienophile 54 were 370 μM and 8 mM. Chiral HPLC were used to analyze the D-A product of 55 and 56, giving ee value as high as 95%. When the D-nucleotides were replaced with L-nucleotides, the enantioselectivity was completely reversed, as expected (−95% ee).
Fig. 4The Diels-Alder reactions are catalyzed by a Diels-Alderase ribozyme.
]. The hydrophobic pocket is precisely complementary to the reaction product and governs the stereoselectivity (Fig. 5) Mutagenesis experiments were also performed to systematically probe the relevance of crystallographically observed ground-state interactions for catalytic function. According to the results of these experiments, mutations that destabilize the binding pocket kill the enzymatic activity completely and hydrogen bonding interactions between two nucleotides (U17 and G9) and the product appear to promote the Diels-Alder reaction by 25-fold (Fig. 5) [
Fig. 5The tertiary structure of the 49-mer Diels-Alderase ribozyme complexed with the D-A product. The D-A product is in green in the active pocket, and Mg2+ ions are shown as green balls. The yellow dashed lines indicate the hydrogen-bonding interactions between the carbonyl oxygen of the D-A product and two hydrogen bond donors from RNA (the 2-NH2 group of G9 or the 2′-OH group of U17).
4. Chemo-enzymatic synthesis of D-A products using Diels-Alderases
Although artificial Diels-Alderases have been successfully developed by different strategies, their synthetic application in the chemo-enzymatic synthesis of complex natural products is rare due to the limited substrate scope or imperfect stereoselectivity. Due to the inherited advantage in stereoselectivity and catalytic efficiency, naturally occurring Diels-Alderases have been considered as a powerful synthetic tool in chemo-enzymatic syntheses of the corresponding D-A products [
]. The resulting chalcone 59 was transformed into flavonoid 60 by the iodine-mediated oxidative cyclization. Deprotection of 60 followed by reprotection with acetyl group gave 62. Regioselective iodination of 62 was achieved using benzyltrimethylammonium dichloroiodate (BTMA•ICl2) to yield the critical synthetic intermediate 63. After optimization of experimental conditions, the diene precursor 64 was obtained via a palladium-catalyzed cross-coupling reaction.
Another diene precursor, 65, was synthesized following a similar synthetic route. With these two diene precursors in hand, diene 66 or 67 were generated in situ by mild hydrolysis and then added to the reaction mixture containing morachalcone A (42) and MaDA. Due to the high enantioselectivity and endo-specificity of the enzymatic Diels-Alder reaction, endo-products artonin I and dideoxyartonin I were obtained in 90% and 38% yield over two steps, respectively, and in nearly optically pure form (Scheme 11).
Scheme 11Chemo-enzymatic syntheses of artonin I and dideoxyartonin I.
To take advantage of their substrate promiscuity, endo-selective MaDA-1 and exo-selective MaDA-3 were used for the chemo-enzymatic synthesis of D-A natural products (e.g., endo-chalcomoracin and its exo-isomer mongolicin F) as well as their analogs containing non-natural dienes or/and dienophiles [
]. D-A products containing unnatural dienes (68–79) with endo-configuration were obtained as the main product in 10%–87% isolated yields using MaDA-1 as the catalyst. On the other hand, the corresponding exo-isomers were selectively produced in the presence of MaDA-3 in isolated yields of 14–88% (Fig. 6).
Fig. 6The unnatural D-A products synthesized using MaDA-1 and MaDA-3.
Besides unnatural dienes, MaDA-1 and MaDA-3 can also recognize unnatural dienophiles and produce D-A products 80–84 except for exo-80 and endo-84. These data suggest that both wild-type MaDA-1 and MaDA-3 have limitations with some substrates that will require further protein engineering to overcome. Moreover, D-A products (85 and 86) containing both unnatural dienes and dienophiles were also accessible using wild-type MaDA-1 and MaDA-3 with very high enantioselectivities (≥98%), albeit in low isolated yield. (Fig. 6).
5. Discussion
Driven by the synthetic importance of D-A cycloadditions, especially the intermolecular reactions, the discovery or design of enzymes that formally catalyze this chemical transformation with high efficiency and stereoselectivity has received continuous attention since the late 1990s. Since the discovery of the first stand-alone Diels-Alderase, SpnF [
], in 2011, genome mining has been proven to be a powerful method to identify new Diels-Alderases from microbial natural product biosynthetic pathways. In contrast, discovery of Diels-Alderases from plants using genome mining strategies is more challenging due to their huge genomes and the scarcity of biosynthetic gene clusters. In this case, other omics strategies such as transcriptome analyses, proteomic technologies, and metabolome analyses are alternative methods to narrow down candidate genes for function characterization. Despite significant achievements in this field, naturally-occurring intermolecular Diels-Alderases are still very limited compared with intramolecular Diels-Alderases. Recently, two groups have individually reported that intramolecular Diels-Alderases (TbtD [
]) could also perform stereospecific intermolecular D-A reaction on non-natural substrates, opening a new window for discovering more new-to-nature enzymatic intermolecular D-A reactions from the reported intramolecular Diels-Alderases via substrate engineering or protein engineering.
As discussed in the Section 2, the naturally occurring inter-molecular Diels-Alderases are phylogenetically highly diverse without any apparent sequence conservation, indicating their independent evolution from distinct protein families. It seems that nature tends to choose biosynthetic enzymes catalyzing the formation of highly reactive dienes as scaffolds to modify them for product diversification via intermolecular D-A reactions evolutionally. Considering that the diene moiety already binds to the active site, it is more likely to evolutionally generate another space accommodating the dienophile moiety than de novo evolution of an active site that both diene and dienophile can simultaneously bind to. In contrast, two different approaches to artificially create intermolecular Diels-Alderases are used by scientists [
]: 1) selecting and designing an enzyme with a proper binding pocket that can stabilize the transition state of the selected D-A reaction (via immunization, computational design or RNA selection); 2) selecting a protein cavity to accommodate a cofactor to catalyze the reaction (artificial metalloenzyme). However, the initially designed or selected enzymes are usually featured with low catalytic efficiency or poor stereoselectivity and require further structure optimization via protein engineering to be used in chemical synthesis, which mimics how naturally occurring Diels-Alderases evolve.
In summary, this review highlights the recent remarkable discoveries of intermolecular Diels-Alderases and provides a future perspective on the potential applications of Diels-Alderase in biocatalysis. We can envision that more and more interesting new enzymes will be discovered and developed to shape our capacity to make functional organic molecules more efficiently.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work is funded by grants from the National Key Research and Development Program of China (2017YFA0505200 to X.L.; 2021YFC2102902 to L.G.), the National Natural Science Foundation of China Grant (21625201, 21961142010, 91853202 to X.L.; 22101009 to L.G.), the Beijing Outstanding Young Scientist Program, China (BJJWZYJH0120191000 1001 to X.L.), and a special research grant from Novartis (Switzerland) for developing new biocatalysis. We also thank W.-Y. Li for proofreading the manuscript.
Recent advances in applying Diels-Alder reactions involving o-quinodimethanes, aza-o-quinone methides and o-quinone methides in natural product total synthesis.
Solanapyrone synthase, a possible Diels-Alderase and iterative type I polyketide synthase encoded in a biosynthetic gene cluster from Alternaria solani.
Biosynthesis of versipelostatin: identification of an enzyme-catalyzed [4+2]-cycloaddition required for macrocyclization of spirotetronate-containing polyketides.
Characterization, sequence, and expression in Escherichia coli of the novel enzyme catalyzing unusual multistep transformation of 2-pyrones to benzoates.
Theoretical exploration of the mechanism of riboflavin formation from 6,7-dimethyl-8-ribityllumazine: nucleophilic catalysis, hydride transfer, hydrogen atom transfer, or nucleophilic addition?.
Total synthesis and computational investigations of sesquiterpene-tropolones ameliorate stereochemical inconsistencies and resolve an ambiguous biosynthetic relationship.
00009-2&id=ga1.jpg){kind=link}
00009-2&id=sc1.jpg){kind=link}
00009-2&id=sc2.jpg){kind=link}
00009-2&id=sc3.jpg){kind=link}
00009-2&id=sc4.jpg){kind=link}
00009-2&id=gr1.jpg){kind=link}
00009-2&id=sc5.jpg){kind=link}
00009-2&id=sc6.jpg){kind=link}
00009-2&id=gr2.jpg){kind=link}
00009-2&id=sc7.jpg){kind=link}
00009-2&id=sc8.jpg){kind=link}
00009-2&id=gr3.jpg){kind=link}
00009-2&id=sc9.jpg){kind=link}
00009-2&id=sc10.jpg){kind=link}
00009-2&id=gr4.jpg){kind=link}
00009-2&id=gr5.jpg){kind=link}
00009-2&id=sc11.jpg){kind=link}
00009-2&id=gr6.jpg){kind=link}