news & views ENZYMOLOGY
A radical finding
A new family of radical halogenases has been discovered that regio- and stereoselectively chlorinates the unactivated carbon center of indolemonoterpenoid substrates without the prerequisite for the substrate to be bound to a protein carrier.
Rebecca J M Goss & Sabine Grüschow tethered to the prosthetic group of the SyrB1 carrier protein domain at the γ-position8 (Fig. 1); similarly, the halogenase domain in CurA only chlorinates 3-hydroxy-3methylglutarate when it is attached to the downstream carrier protein domain of the same enzyme9. The absolute requirement for protein-bound substrate is repeated throughout the PhyH-type halogenases3,4. Furthermore, these halogenases exhibit specificity toward the carrier protein. The generation of the cyanobacterial hapalindole-type alkaloids remained mysterious for many years, with the primary postulation of their formation being the joining of two putative precursors β-ocimene and (2∙-isocyanoethynyl) indole by a chloronium ion–mediated cationic cascade. However, the authors’ careful observation of the coexistence of deschlorinated as well as chlorinated analogs led them to postulate a late-stage chlorination event. Amongst the putative tailoring enzymes of the biosynthetic machinery responsible for weltwindolinone biosynthesis in Hapalosiphon welwitschii UTEX B1830 (ref. 10), WelO5 has high homology to NHI dioxygenases but no homology to previously characterized NHI halogenases. However, as in the other NHI halogenases, the canonical carboxylate iron ligand, conserved in NHI dioxygenases, is replaced by glycine or alanine (Gly166 in WelO5); the vacated ligand site is known to serve as the docking site for the halogen8. NHI enzymes often require purification under strictly anaerobic conditions to be active. The more amenable WelO5, however, could be heterologously produced and aerobically purified in the apo form (without bound iron); activity could then be reconstituted in vitro by adding a nitrogen-purged Fe(ii)-solution. Chlorination of 12-epi-fischerindole U (Fig. 1) and the partially cyclized derivative 12-epi-hapalindole C proceeded regio- and stereoselectively with a turnover five to tenfold higher than that reported for the PhyH-type halogenases. The introduction of a halogen into a molecule can be highly advantageous, conferring a profound effect upon compound stability, bioavailability and bioactivity; from
nature chemical biology | Advance online publication | www.nature.com/naturechemicalbiology
a synthetic point of view, it can provide a chemically orthogonal handle within a natural product that enables selective functionalization. Synthetic chemical approaches to accessing halogenated materials often use harsh conditions and environmentally detrimental reagents; the processes are also wasteful, as often the regiochemistry of halogenation is not selective. The potential for halogenase enzymes to overcome these problems is evident. The discovery of WelO5 opens up exciting possibilities for biocatalyst development as this enzyme is without many of the unfavorable characteristics of its PhyH counterparts. Most importantly, it is not necessary to attach the substrate to a specific carrier protein. This step demands the presence of an additional enzyme that, when non-native substrates are to be used, will also require adjustments. Furthermore, the specificity of PhyH-type halogenases toward the carrier protein has limited the use of these halogenases in non-native contexts 2OG-FeII_Oxy_3: WelO5 α-KG, Fe(II) O2, NaCl
NC H
Cl
NC H
H
H
N H
PhyH:
N H OH
H2N S
A
CP
O
OH Cl
H2N
SyrB2 α-KG, Fe(II) O2, NaCl
S
A
O
CP
Figure 1 | Pfam classification of and representative reactions catalyzed by NHI halogenases. WelO5 unusually catalyzes late-stage halogenation of a non-enzyme-bound substrate, though both types of enzymes fall within the Cupin superfamily. A, adenylation domain; CP, carrier protein domain of SyrB1; -KG, -ketoglutarate. 1
Debbie Maizels/Nature Publishing Group
npg
© 2014 Nature America, Inc. All rights reserved.
U
ntil as recently as the early 1970s, halogenated natural products were considered as infrequently occurring biosynthetic curiosities or artifacts of extraction and isolation1. Today, over 5,000 halogenated metabolites have been isolated and identified2. The enzyme catalysts responsible for the installation of halogens into natural products use different chemical mechanisms that generate and use electrophilic, nucleophilic or radical halogenation species. The mechanism used is tailored to the reactivity of the substrate and the halogen3,4. The insertion of a chlorine or bromine into an unactivated, aliphatic C-H bond is particularly chemically demanding; recent studies have identified a family of nonheme iron-dependent (NHI) halogenases that perform this transformation through the generation of a radical species. Until now, all of the NHI halogenases that had been characterized required the substrate to be covalently linked to a carrier protein. In this issue, Hillwig and Liu5 describe the identification of a new family of radical halogenases that regio- and stereoselectively chlorinate the unactivated carbon center of 12-epi-fischerindole without the prerequisite for the substrate to be protein bound. The most abundant mechanism of biosynthetic halogenation involves the generation of an electrophilic species by either a metal- or flavin-dependent enzyme, enabling halogenation at nucleophilic positions such as electron-rich aromatic rings or olefins4. A small family of halogenases use halide nucleophiles to halogenate S-adenosylmethionine and are the only enzymes known so far that are capable of inserting fluorine into organic molecules6. NHI halogenases, in contrast, catalyze halogenation through a radical mechanism allowing for the insertion of a halogen into an unactivated, aliphatic C-H bond3,4. They are closely related to NHI dioxygenases of the PhyH family, and they share the same mechanism to create the reactive ferryl-oxo species7 with the consumption of α-ketoglutarate as co-substrate. For example, the NHI halogenase SyrB2 chlorinates l-threonine
news & views alkaloids, is not chlorinated by WelO5, which suggests a certain degree of substrate specificity. This is certainly an aspect that will require further investigations into WelO5 and related enzymes. It will be exciting to see how far future studies will be able to push the development of halogenases for use in biocatalysis. Good candidates for the halogenation of electron-rich, activated carbon centers have already been identified in flavin- or vanadium-dependent halogenases; the discovery of WelO5, a radical halogenase, represents a potentially very useful tool for the halogenaton of unactivated alkyl groups. ■ Rebecca J.M. Goss and Sabine Grüschow are at the School of Chemistry, University of St. Andrews, St. Andrews, Fife, UK. e-mail: [email protected]
Published online 14 September 2014 doi:10.1038/nchembio.1649 References
1. Fowden, L. & Robinson, R. Proc. R. Soc. Lond. B Biol. Sci. 171, 5–18 (1968). 2. Gribble, G.W. Alkaloids Chem. Biol. 71, 1–165 (2012). 3. Smith, D.R.M., Grüschow, S. & Goss, R.J.M. Curr. Opin. Chem. Biol. 17, 276–283 (2013). 4. Butler, A. & Sandy, M. Nature 460, 848–854 (2009). 5. Hillwig, M.L. & Liu, X. Nat. Chem. Biol. doi:10.1038/ nchembio.1625 (14 September 2014) 6. Deng. H. & O’Hagan, D. Curr. Opin. Chem. Biol. 12, 582–592 (2008). 7. Krebs, C., Galonic Fujimori, D., Walsh, C.T. & Bollinger, J.M. Acc. Chem. Res. 40, 484–492 (2007). 8. Blasiak, L.C., Vaillancourt, F.H., Walsh, C.T. & Drennan, C.L. Nature 440, 368–371 (2006). 9. Gu, L. et al. Nature 459, 731–735 (2009). 10. Hillwig, M.L. et al. ChemBioChem 15, 665–669 (2014).
Competing financial interests
The authors declare no competing financial interests.
npg
© 2014 Nature America, Inc. All rights reserved.
to evolutionary closely related systems as the protein-protein interactions that govern this specificity are still poorly understood. The need for carrier protein attachment of the substrate has also hitherto rendered NHI halogenases unsuitable for enzyme engineering. WelO5 is without this drawback, and furthermore it processes a compound substantially more complex than most of the known NHI halogenase substrates. Its potentially larger binding pocket will perhaps accommodate other substrates. However, though WeIO5 can process two distinctly different natural substrates, demonstrating a level of flexibility, it remains to be seen whether the wild-type enzyme can chlorinate or indeed brominate non-native substrates. Hillwig and Liu5 report that hapalindole J, a Fischerella muscicola metabolite with a different ring system than the H. welwitschii
2
nature chemical biology | Advance online publication | www.nature.com/naturechemicalbiology
A radical finding
A new family of radical halogenases has been discovered that regio- and stereoselectively chlorinates the unactivated carbon center of indolemonoterpenoid substrates without the prerequisite for the substrate to be bound to a protein carrier.
Rebecca J M Goss & Sabine Grüschow tethered to the prosthetic group of the SyrB1 carrier protein domain at the γ-position8 (Fig. 1); similarly, the halogenase domain in CurA only chlorinates 3-hydroxy-3methylglutarate when it is attached to the downstream carrier protein domain of the same enzyme9. The absolute requirement for protein-bound substrate is repeated throughout the PhyH-type halogenases3,4. Furthermore, these halogenases exhibit specificity toward the carrier protein. The generation of the cyanobacterial hapalindole-type alkaloids remained mysterious for many years, with the primary postulation of their formation being the joining of two putative precursors β-ocimene and (2∙-isocyanoethynyl) indole by a chloronium ion–mediated cationic cascade. However, the authors’ careful observation of the coexistence of deschlorinated as well as chlorinated analogs led them to postulate a late-stage chlorination event. Amongst the putative tailoring enzymes of the biosynthetic machinery responsible for weltwindolinone biosynthesis in Hapalosiphon welwitschii UTEX B1830 (ref. 10), WelO5 has high homology to NHI dioxygenases but no homology to previously characterized NHI halogenases. However, as in the other NHI halogenases, the canonical carboxylate iron ligand, conserved in NHI dioxygenases, is replaced by glycine or alanine (Gly166 in WelO5); the vacated ligand site is known to serve as the docking site for the halogen8. NHI enzymes often require purification under strictly anaerobic conditions to be active. The more amenable WelO5, however, could be heterologously produced and aerobically purified in the apo form (without bound iron); activity could then be reconstituted in vitro by adding a nitrogen-purged Fe(ii)-solution. Chlorination of 12-epi-fischerindole U (Fig. 1) and the partially cyclized derivative 12-epi-hapalindole C proceeded regio- and stereoselectively with a turnover five to tenfold higher than that reported for the PhyH-type halogenases. The introduction of a halogen into a molecule can be highly advantageous, conferring a profound effect upon compound stability, bioavailability and bioactivity; from
nature chemical biology | Advance online publication | www.nature.com/naturechemicalbiology
a synthetic point of view, it can provide a chemically orthogonal handle within a natural product that enables selective functionalization. Synthetic chemical approaches to accessing halogenated materials often use harsh conditions and environmentally detrimental reagents; the processes are also wasteful, as often the regiochemistry of halogenation is not selective. The potential for halogenase enzymes to overcome these problems is evident. The discovery of WelO5 opens up exciting possibilities for biocatalyst development as this enzyme is without many of the unfavorable characteristics of its PhyH counterparts. Most importantly, it is not necessary to attach the substrate to a specific carrier protein. This step demands the presence of an additional enzyme that, when non-native substrates are to be used, will also require adjustments. Furthermore, the specificity of PhyH-type halogenases toward the carrier protein has limited the use of these halogenases in non-native contexts 2OG-FeII_Oxy_3: WelO5 α-KG, Fe(II) O2, NaCl
NC H
Cl
NC H
H
H
N H
PhyH:
N H OH
H2N S
A
CP
O
OH Cl
H2N
SyrB2 α-KG, Fe(II) O2, NaCl
S
A
O
CP
Figure 1 | Pfam classification of and representative reactions catalyzed by NHI halogenases. WelO5 unusually catalyzes late-stage halogenation of a non-enzyme-bound substrate, though both types of enzymes fall within the Cupin superfamily. A, adenylation domain; CP, carrier protein domain of SyrB1; -KG, -ketoglutarate. 1
Debbie Maizels/Nature Publishing Group
npg
© 2014 Nature America, Inc. All rights reserved.
U
ntil as recently as the early 1970s, halogenated natural products were considered as infrequently occurring biosynthetic curiosities or artifacts of extraction and isolation1. Today, over 5,000 halogenated metabolites have been isolated and identified2. The enzyme catalysts responsible for the installation of halogens into natural products use different chemical mechanisms that generate and use electrophilic, nucleophilic or radical halogenation species. The mechanism used is tailored to the reactivity of the substrate and the halogen3,4. The insertion of a chlorine or bromine into an unactivated, aliphatic C-H bond is particularly chemically demanding; recent studies have identified a family of nonheme iron-dependent (NHI) halogenases that perform this transformation through the generation of a radical species. Until now, all of the NHI halogenases that had been characterized required the substrate to be covalently linked to a carrier protein. In this issue, Hillwig and Liu5 describe the identification of a new family of radical halogenases that regio- and stereoselectively chlorinate the unactivated carbon center of 12-epi-fischerindole without the prerequisite for the substrate to be protein bound. The most abundant mechanism of biosynthetic halogenation involves the generation of an electrophilic species by either a metal- or flavin-dependent enzyme, enabling halogenation at nucleophilic positions such as electron-rich aromatic rings or olefins4. A small family of halogenases use halide nucleophiles to halogenate S-adenosylmethionine and are the only enzymes known so far that are capable of inserting fluorine into organic molecules6. NHI halogenases, in contrast, catalyze halogenation through a radical mechanism allowing for the insertion of a halogen into an unactivated, aliphatic C-H bond3,4. They are closely related to NHI dioxygenases of the PhyH family, and they share the same mechanism to create the reactive ferryl-oxo species7 with the consumption of α-ketoglutarate as co-substrate. For example, the NHI halogenase SyrB2 chlorinates l-threonine
news & views alkaloids, is not chlorinated by WelO5, which suggests a certain degree of substrate specificity. This is certainly an aspect that will require further investigations into WelO5 and related enzymes. It will be exciting to see how far future studies will be able to push the development of halogenases for use in biocatalysis. Good candidates for the halogenation of electron-rich, activated carbon centers have already been identified in flavin- or vanadium-dependent halogenases; the discovery of WelO5, a radical halogenase, represents a potentially very useful tool for the halogenaton of unactivated alkyl groups. ■ Rebecca J.M. Goss and Sabine Grüschow are at the School of Chemistry, University of St. Andrews, St. Andrews, Fife, UK. e-mail: [email protected]
Published online 14 September 2014 doi:10.1038/nchembio.1649 References
1. Fowden, L. & Robinson, R. Proc. R. Soc. Lond. B Biol. Sci. 171, 5–18 (1968). 2. Gribble, G.W. Alkaloids Chem. Biol. 71, 1–165 (2012). 3. Smith, D.R.M., Grüschow, S. & Goss, R.J.M. Curr. Opin. Chem. Biol. 17, 276–283 (2013). 4. Butler, A. & Sandy, M. Nature 460, 848–854 (2009). 5. Hillwig, M.L. & Liu, X. Nat. Chem. Biol. doi:10.1038/ nchembio.1625 (14 September 2014) 6. Deng. H. & O’Hagan, D. Curr. Opin. Chem. Biol. 12, 582–592 (2008). 7. Krebs, C., Galonic Fujimori, D., Walsh, C.T. & Bollinger, J.M. Acc. Chem. Res. 40, 484–492 (2007). 8. Blasiak, L.C., Vaillancourt, F.H., Walsh, C.T. & Drennan, C.L. Nature 440, 368–371 (2006). 9. Gu, L. et al. Nature 459, 731–735 (2009). 10. Hillwig, M.L. et al. ChemBioChem 15, 665–669 (2014).
Competing financial interests
The authors declare no competing financial interests.
npg
© 2014 Nature America, Inc. All rights reserved.
to evolutionary closely related systems as the protein-protein interactions that govern this specificity are still poorly understood. The need for carrier protein attachment of the substrate has also hitherto rendered NHI halogenases unsuitable for enzyme engineering. WelO5 is without this drawback, and furthermore it processes a compound substantially more complex than most of the known NHI halogenase substrates. Its potentially larger binding pocket will perhaps accommodate other substrates. However, though WeIO5 can process two distinctly different natural substrates, demonstrating a level of flexibility, it remains to be seen whether the wild-type enzyme can chlorinate or indeed brominate non-native substrates. Hillwig and Liu5 report that hapalindole J, a Fischerella muscicola metabolite with a different ring system than the H. welwitschii
2
nature chemical biology | Advance online publication | www.nature.com/naturechemicalbiology
