NEWS & VIEWS ORG ANIC SYNTHESIS
Better chemistry through radicals An iron catalyst has been developed that mediates bond formation between a wide range of alkene reactants, opening up short synthetic routes to compounds that were previously accessible only through arduous pathways. See Article p.343 STEVEN L. CASTLE
R
eactions that form carbon–carbon (C–C) bonds are essential for synthesi zing complex organic molecules from simple, inexpensive precursors. The value that such molecules have as pharmaceuticals, agrochemicals and materials makes these reactions essential to the practice of organic synthesis. Most classical methods for generat ing C–C bonds rely on reagents that are either strongly basic or strongly acidic, and some require high temperatures to proceed. Such ‘harsh’ reaction conditions are incompatible with many functional groups — the groups of atoms responsible for the properties and reac tivity of molecules. Functional-group incom patibilities are a major nuisance, because they force chemists to design synthetic routes to target molecules that are circuitous rather than direct. In an exciting development reported on page 343 of this issue, Lo et al.1 have developed a C–C bond-forming reaction that provides a promising solution to this problem. Since the 1970s, several C–C bond-forming reactions catalysed by transition-metal com plexes have been developed that use mild (weakly acidic or basic) or neutral reaction conditions, to address the issue of functionalgroup incompatibilities. Although these pro cesses constitute a great advance compared with classical C–C bond-forming methods, the most commonly used catalysts are based on the costly element palladium2. Researchers have therefore begun to explore cheaper alter natives to palladium for these reactions. Iron, with its high natural abundance and low cost, is a logical choice. Many iron-catalysed C–C bond-forming reactions have been discovered in the past decade3. Lo et al. were inspired by the ability of iron catalysts to generate reactive free-rad ical intermediates from alkenes (compounds that contain carbon–carbon double bonds), a property that has been known for more than 20 years4. Earlier this year, some of the authors of the current paper reported an ironcatalysed C–C bond-forming process that joins two alkenes together through the inter mediacy of a radical5. Although useful, this reaction was compatible with only a limited range of functional groups. Lo and colleagues
R2 R1 X Alkene
R3
Catalyst, reducing agent Weak base
O O
O Fe O
R2 R1
R3 + X Radical
C–C bond formation
R4 EWG Acceptor alkene
R2
R4 EWG
R1 R3 X Product
O O
Catalyst
Figure 1 | Iron-catalysed carbon–carbon bond formation with unprecedented functional-group tolerance. Lo et al.1 report an iron catalyst that couples together two alkenes through carbon–carbon (C–C) bond formation. In the presence of a weak base, the catalyst engages a reducing agent to form a species (not shown) that generates a radical from an alkene. This radical adds to an acceptor alkene that incorporates an electron-withdrawing group (EWG), forming a new C–C bond (shown in red in the product). The functional-group tolerance of the reaction derives from the large ligands (purple) bound to the iron atom of the catalyst. R1–R4 represent different carbon-based groups; X represents functional groups containing atoms such as oxygen, nitrogen, sulphur, boron, silicon, fluorine, chlorine, bromine and iodine. The dot on the radical represents a single unpaired electron; broken lines in the catalyst indicate delocalized bonds.
thus set out to develop an improved method that would exhibit broad functional-group tolerance. The researchers hypothesized that the restricted functional-group tolerance of their original method was caused by the small ligand molecules bound to the iron atom. Accord ingly, they prepared and evaluated several iron catalysts that have large ligands. This revealed that a catalyst bearing three bulky diisobu tyrylmethane ligands effectively mediated the formation of a C–C bond between two alkenes in the presence of a reducing agent (Fig. 1). Further investigation established the benefi cial effect of a weakly basic additive (disodium phosphate), although its specific role in the reaction is unclear. Because radical-mediated reactions proceed under mild conditions and typically involve uncharged intermediates, their functionalgroup tolerance generally exceeds that exhib ited by other types of organic reaction, which often involve harsh conditions and charged intermediates6,7. But Lo and co-workers’ reac tion exhibits unprecedented functional-group tolerance, even for a process involving radicals. Specifically, atoms such as sulphur, boron, chlorine, bromine and iodine can remain
3 3 2 | NAT U R E | VO L 5 1 6 | 1 8 / 2 5 D E C E M B E R 2 0 1 4
© 2014 Macmillan Publishers Limited. All rights reserved
attached to the radical-bearing carbon atom, emerging unscathed at the end of the reaction; bonds between carbon and these atoms are fre quently cleaved during radical processes. The new reaction therefore allows compounds with functional groups containing these atoms to be accessed in a direct and straightforward manner. Another shortcoming of many C–C bondforming reactions is the need to rigorously exclude air and moisture from them. Reactions that cannot tolerate the presence of oxygen or water require cumbersome procedures that are difficult to reproduce with precision. Because the new iron-catalysed alkene coupling pro ceeds efficiently in the presence of water and air, it is conducted using a simple, user-friendly protocol. As a result, anyone with basic train ing in organic synthesis should be able to suc cessfully perform this reaction and generate reproducible results. All organic reactions have limitations, and Lo and colleagues’ alkene coupling is no excep tion. The substrate scope of the alkene that acts as a radical precursor (the green alkene in Fig. 1) is exceptionally broad, but there are some constraints to the structure of the other ‘acceptor’ alkene. Currently, bulky acceptor
NEWS & VIEWS RESEARCH alkenes — those with large groups at the R4 position shown in Figure 1 — are not viable coupling partners. Further fine-tuning of the catalyst structure and reaction conditions might uncover a solution to this problem. By facilitating the linking of two alkenes through carbon–carbon bond formation, Lo and co-workers’ reaction will allow the direct generation of valuable, structurally complex organic molecules from simpler precursors.
What is more, the iron catalyst is readily prepared from fairly inexpensive ingredients. This method therefore has the potential to transform the way in which chemists think about constructing complicated molecules. ■ Steven L. Castle is in the Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84602, USA. e-mail: [email protected]
SYNTHETIC B IO LO GY
Toehold gene switches make big footprints The development of RNA-based devices called toehold switches that regulate translation might usher in an era in which protein production can be linked to almost any RNA input and provide precise, low-cost diagnostics. SIMON AUSLÄNDER & MARTIN FUSSENEGGER
A
fundamental tool of synthetic biology is a type of genetic device that controls the expression of target genes in a trigger-inducible manner, and so can be used to predictably and robustly program cellular behaviour. The number of such gene switches is growing, and switches have been success fully used in combination with other compo nents, such as enzymes to assemble metabolic pathways that produce biofuels1 and thera peutic drugs2, and in designer cells that have the potential to correct metabolic diseases3–5. But the design of circuits of interconnecting switches is often complicated by the fact that each switch is made of natural components and is sensitive to its own predetermined trigger compound. A strategy that produces compatible gene switches tailored to desired trigger compounds would enable the switches to be easily assembled in combination, increas ing the precision and complexity with which cellular behaviour can be programmed. Writ ing in Cell, Green et al.6 describe a method for generating gene switches that can indeed be tailored to desired RNA inputs. RNA is gathering momentum as a control device for synthetic biology. RNAs are modu lar, programmable and versatile. Furthermore, the specific sequence of each RNA dictates which molecules it can interact with and what functions its structure confers. The primary RNA sequence is determined by the sequential arrangement of different nucleotides, and this sequence can be engineered so that it forms secondary RNA structures internally or with complementary DNA or RNA molecules. One such structure is the hairpin loop, which
comprises two base-paired sequences ending in an unpaired loop. Secondary structures can affect the translation of messenger RNA, and so can be exploited to regulate protein produc tion from genes of interest. Translation of mRNA occurs in a complex molecular machine called the ribosome. The ribosome contains a small and a large subunit, both of which are composed of a mixture of ribosomal RNAs and proteins. In bacteria, mRNAs are recruited to the ribosome through their ribosome-binding site (RBS) — a sequence that binds to the small subunit to initiate translation.
a Expression off
1. Lo, J. C., Gui, J., Yabe, Y., Pan, C.-M. & Baran, P. S. Nature 516, 343–348 (2014). 2. de Meijere, A. & Diederich, F. (eds) Metal-Catalyzed Cross-Coupling Reactions 2nd edn (Wiley, 2004). 3. Nakamura, E. et al. Org. React. 83, Ch. 1, 1–209 (2014). 4. Kato, K. & Mukaiyama, T. Chem. Lett. 21, 1137–1140 (1992). 5. Lo, J. C., Yabe, Y. & Baran, P. S. J. Am. Chem. Soc. 136, 1304–1307 (2014). 6. Rowlands, G. J. Tetrahedron 65, 8603–8655 (2009). 7. Rowlands, G. J. Tetrahedron 66, 1593–1636 (2010).
The reversible nature of this binding interaction is exploited by a class of engineered RNA-based gene switches called riboregula tors, which contain an ‘anti-RBS sequence’ that binds to the RBS to form a hairpin loop7, thus preventing the mRNA from accessing the ribosome and lowering the rate of trans lation6. The anti-RBS sequence is located in the target mRNA itself, in a region that will not be translated into protein, upstream of the site where translation begins. Riboregulators are switched by a ‘trigger sequence’ that inter acts with and disrupts the hairpin, forming an alternative RNA structure that permits RBS–ribosome binding. Depending on the presence or absence of the trigger RNA, target gene expression can therefore be switched on or off. However, because typical riboregulators must fit into the upstream mRNA region and bind to the RBS, they can be designed for only a limited number of trigger sequences. Green and colleagues have developed a more diverse type of riboregulator, which they call a toehold switch. Toehold riboregulators are designed to interact with the region around the protein-coding start site of each mRNA instead of the RBS, but are not complemen tary to the start site itself (Fig. 1). Furthermore,
b Expression on Small subunit
RBS
Large subunit Ribosome Stranddisplacement reaction
Translational start site mRNA Toehold sequence
Protein-coding sequence
Trigger RNA
Figure 1 | The design of toehold switches. a, Green et al.6 have designed an RNA-based device, called a toehold switch, that can regulate translation of bacterial messenger RNA in response to the presence or absence of any desired ‘trigger’ RNA. Toehold switches are located upstream of the site at which translation begins. The switch has an exposed single-stranded region called the toehold sequence that is designed to be complementary to the trigger RNA. To be translated, bacterial mRNA must bind to the small ribosomal subunit through a ribosome-binding site (RBS), but, if the trigger RNA is absent, the presence of the toehold switch causes the formation of a hairpin structure that blocks RBS–ribosome binding, thereby preventing translation. b, The presence of the trigger RNA causes a strand-displacement reaction that breaks up the hairpin structure, exposes the RBS to the ribosome and induces translation. 1 8 / 2 5 D E C E M B E R 2 0 1 4 | VO L 5 1 6 | NAT U R E | 3 3 3
© 2014 Macmillan Publishers Limited. All rights reserved
Better chemistry through radicals An iron catalyst has been developed that mediates bond formation between a wide range of alkene reactants, opening up short synthetic routes to compounds that were previously accessible only through arduous pathways. See Article p.343 STEVEN L. CASTLE
R
eactions that form carbon–carbon (C–C) bonds are essential for synthesi zing complex organic molecules from simple, inexpensive precursors. The value that such molecules have as pharmaceuticals, agrochemicals and materials makes these reactions essential to the practice of organic synthesis. Most classical methods for generat ing C–C bonds rely on reagents that are either strongly basic or strongly acidic, and some require high temperatures to proceed. Such ‘harsh’ reaction conditions are incompatible with many functional groups — the groups of atoms responsible for the properties and reac tivity of molecules. Functional-group incom patibilities are a major nuisance, because they force chemists to design synthetic routes to target molecules that are circuitous rather than direct. In an exciting development reported on page 343 of this issue, Lo et al.1 have developed a C–C bond-forming reaction that provides a promising solution to this problem. Since the 1970s, several C–C bond-forming reactions catalysed by transition-metal com plexes have been developed that use mild (weakly acidic or basic) or neutral reaction conditions, to address the issue of functionalgroup incompatibilities. Although these pro cesses constitute a great advance compared with classical C–C bond-forming methods, the most commonly used catalysts are based on the costly element palladium2. Researchers have therefore begun to explore cheaper alter natives to palladium for these reactions. Iron, with its high natural abundance and low cost, is a logical choice. Many iron-catalysed C–C bond-forming reactions have been discovered in the past decade3. Lo et al. were inspired by the ability of iron catalysts to generate reactive free-rad ical intermediates from alkenes (compounds that contain carbon–carbon double bonds), a property that has been known for more than 20 years4. Earlier this year, some of the authors of the current paper reported an ironcatalysed C–C bond-forming process that joins two alkenes together through the inter mediacy of a radical5. Although useful, this reaction was compatible with only a limited range of functional groups. Lo and colleagues
R2 R1 X Alkene
R3
Catalyst, reducing agent Weak base
O O
O Fe O
R2 R1
R3 + X Radical
C–C bond formation
R4 EWG Acceptor alkene
R2
R4 EWG
R1 R3 X Product
O O
Catalyst
Figure 1 | Iron-catalysed carbon–carbon bond formation with unprecedented functional-group tolerance. Lo et al.1 report an iron catalyst that couples together two alkenes through carbon–carbon (C–C) bond formation. In the presence of a weak base, the catalyst engages a reducing agent to form a species (not shown) that generates a radical from an alkene. This radical adds to an acceptor alkene that incorporates an electron-withdrawing group (EWG), forming a new C–C bond (shown in red in the product). The functional-group tolerance of the reaction derives from the large ligands (purple) bound to the iron atom of the catalyst. R1–R4 represent different carbon-based groups; X represents functional groups containing atoms such as oxygen, nitrogen, sulphur, boron, silicon, fluorine, chlorine, bromine and iodine. The dot on the radical represents a single unpaired electron; broken lines in the catalyst indicate delocalized bonds.
thus set out to develop an improved method that would exhibit broad functional-group tolerance. The researchers hypothesized that the restricted functional-group tolerance of their original method was caused by the small ligand molecules bound to the iron atom. Accord ingly, they prepared and evaluated several iron catalysts that have large ligands. This revealed that a catalyst bearing three bulky diisobu tyrylmethane ligands effectively mediated the formation of a C–C bond between two alkenes in the presence of a reducing agent (Fig. 1). Further investigation established the benefi cial effect of a weakly basic additive (disodium phosphate), although its specific role in the reaction is unclear. Because radical-mediated reactions proceed under mild conditions and typically involve uncharged intermediates, their functionalgroup tolerance generally exceeds that exhib ited by other types of organic reaction, which often involve harsh conditions and charged intermediates6,7. But Lo and co-workers’ reac tion exhibits unprecedented functional-group tolerance, even for a process involving radicals. Specifically, atoms such as sulphur, boron, chlorine, bromine and iodine can remain
3 3 2 | NAT U R E | VO L 5 1 6 | 1 8 / 2 5 D E C E M B E R 2 0 1 4
© 2014 Macmillan Publishers Limited. All rights reserved
attached to the radical-bearing carbon atom, emerging unscathed at the end of the reaction; bonds between carbon and these atoms are fre quently cleaved during radical processes. The new reaction therefore allows compounds with functional groups containing these atoms to be accessed in a direct and straightforward manner. Another shortcoming of many C–C bondforming reactions is the need to rigorously exclude air and moisture from them. Reactions that cannot tolerate the presence of oxygen or water require cumbersome procedures that are difficult to reproduce with precision. Because the new iron-catalysed alkene coupling pro ceeds efficiently in the presence of water and air, it is conducted using a simple, user-friendly protocol. As a result, anyone with basic train ing in organic synthesis should be able to suc cessfully perform this reaction and generate reproducible results. All organic reactions have limitations, and Lo and colleagues’ alkene coupling is no excep tion. The substrate scope of the alkene that acts as a radical precursor (the green alkene in Fig. 1) is exceptionally broad, but there are some constraints to the structure of the other ‘acceptor’ alkene. Currently, bulky acceptor
NEWS & VIEWS RESEARCH alkenes — those with large groups at the R4 position shown in Figure 1 — are not viable coupling partners. Further fine-tuning of the catalyst structure and reaction conditions might uncover a solution to this problem. By facilitating the linking of two alkenes through carbon–carbon bond formation, Lo and co-workers’ reaction will allow the direct generation of valuable, structurally complex organic molecules from simpler precursors.
What is more, the iron catalyst is readily prepared from fairly inexpensive ingredients. This method therefore has the potential to transform the way in which chemists think about constructing complicated molecules. ■ Steven L. Castle is in the Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84602, USA. e-mail: [email protected]
SYNTHETIC B IO LO GY
Toehold gene switches make big footprints The development of RNA-based devices called toehold switches that regulate translation might usher in an era in which protein production can be linked to almost any RNA input and provide precise, low-cost diagnostics. SIMON AUSLÄNDER & MARTIN FUSSENEGGER
A
fundamental tool of synthetic biology is a type of genetic device that controls the expression of target genes in a trigger-inducible manner, and so can be used to predictably and robustly program cellular behaviour. The number of such gene switches is growing, and switches have been success fully used in combination with other compo nents, such as enzymes to assemble metabolic pathways that produce biofuels1 and thera peutic drugs2, and in designer cells that have the potential to correct metabolic diseases3–5. But the design of circuits of interconnecting switches is often complicated by the fact that each switch is made of natural components and is sensitive to its own predetermined trigger compound. A strategy that produces compatible gene switches tailored to desired trigger compounds would enable the switches to be easily assembled in combination, increas ing the precision and complexity with which cellular behaviour can be programmed. Writ ing in Cell, Green et al.6 describe a method for generating gene switches that can indeed be tailored to desired RNA inputs. RNA is gathering momentum as a control device for synthetic biology. RNAs are modu lar, programmable and versatile. Furthermore, the specific sequence of each RNA dictates which molecules it can interact with and what functions its structure confers. The primary RNA sequence is determined by the sequential arrangement of different nucleotides, and this sequence can be engineered so that it forms secondary RNA structures internally or with complementary DNA or RNA molecules. One such structure is the hairpin loop, which
comprises two base-paired sequences ending in an unpaired loop. Secondary structures can affect the translation of messenger RNA, and so can be exploited to regulate protein produc tion from genes of interest. Translation of mRNA occurs in a complex molecular machine called the ribosome. The ribosome contains a small and a large subunit, both of which are composed of a mixture of ribosomal RNAs and proteins. In bacteria, mRNAs are recruited to the ribosome through their ribosome-binding site (RBS) — a sequence that binds to the small subunit to initiate translation.
a Expression off
1. Lo, J. C., Gui, J., Yabe, Y., Pan, C.-M. & Baran, P. S. Nature 516, 343–348 (2014). 2. de Meijere, A. & Diederich, F. (eds) Metal-Catalyzed Cross-Coupling Reactions 2nd edn (Wiley, 2004). 3. Nakamura, E. et al. Org. React. 83, Ch. 1, 1–209 (2014). 4. Kato, K. & Mukaiyama, T. Chem. Lett. 21, 1137–1140 (1992). 5. Lo, J. C., Yabe, Y. & Baran, P. S. J. Am. Chem. Soc. 136, 1304–1307 (2014). 6. Rowlands, G. J. Tetrahedron 65, 8603–8655 (2009). 7. Rowlands, G. J. Tetrahedron 66, 1593–1636 (2010).
The reversible nature of this binding interaction is exploited by a class of engineered RNA-based gene switches called riboregula tors, which contain an ‘anti-RBS sequence’ that binds to the RBS to form a hairpin loop7, thus preventing the mRNA from accessing the ribosome and lowering the rate of trans lation6. The anti-RBS sequence is located in the target mRNA itself, in a region that will not be translated into protein, upstream of the site where translation begins. Riboregulators are switched by a ‘trigger sequence’ that inter acts with and disrupts the hairpin, forming an alternative RNA structure that permits RBS–ribosome binding. Depending on the presence or absence of the trigger RNA, target gene expression can therefore be switched on or off. However, because typical riboregulators must fit into the upstream mRNA region and bind to the RBS, they can be designed for only a limited number of trigger sequences. Green and colleagues have developed a more diverse type of riboregulator, which they call a toehold switch. Toehold riboregulators are designed to interact with the region around the protein-coding start site of each mRNA instead of the RBS, but are not complemen tary to the start site itself (Fig. 1). Furthermore,
b Expression on Small subunit
RBS
Large subunit Ribosome Stranddisplacement reaction
Translational start site mRNA Toehold sequence
Protein-coding sequence
Trigger RNA
Figure 1 | The design of toehold switches. a, Green et al.6 have designed an RNA-based device, called a toehold switch, that can regulate translation of bacterial messenger RNA in response to the presence or absence of any desired ‘trigger’ RNA. Toehold switches are located upstream of the site at which translation begins. The switch has an exposed single-stranded region called the toehold sequence that is designed to be complementary to the trigger RNA. To be translated, bacterial mRNA must bind to the small ribosomal subunit through a ribosome-binding site (RBS), but, if the trigger RNA is absent, the presence of the toehold switch causes the formation of a hairpin structure that blocks RBS–ribosome binding, thereby preventing translation. b, The presence of the trigger RNA causes a strand-displacement reaction that breaks up the hairpin structure, exposes the RBS to the ribosome and induces translation. 1 8 / 2 5 D E C E M B E R 2 0 1 4 | VO L 5 1 6 | NAT U R E | 3 3 3
© 2014 Macmillan Publishers Limited. All rights reserved
