Skip to content
Publicly Available Published by De Gruyter February 4, 2015

Template-directed nonenzymatic oligonucleotide synthesis: lessons from synthetic chemistry

  • Albert C. Fahrenbach EMAIL logo

Abstract

The nonenzymatic synthesis of nucleic acids, in particular, RNA, and the template-directed synthesis of artificial organic molecules, such as macrocycles, catenanes and rotaxanes, have both undergone significant development since the last half of the 20th century. The intersection of these two fields affords insights into how template effects can lead to information copying and storage at the molecular level. Mechanistic examples of model template-directed RNA replication experiments as well as those for totally artificial organic template-directed syntheses will be discussed. The fact that templates typically bind to their reacted products more tightly than their unreacted substrates may be a mechanistic feature necessary to store information in the form of nucleic acids. Understanding the mechanisms of nonenzymatic RNA synthesis is not only essential for testing the RNA world hypothesis in the context of the origin of life on Earth and other planetary bodies, but may one day afford chemists the insights to construct their own artificial molecular replicators.

Introduction

What can be learned about the origins of life on Earth from artificial template-directed synthetic protocols? The RNA world hypothesis supposes [1, 2] that before the emergence of DNA or proteins, RNA was performing both of their functions – carrying out enzymatic processes [3, 4] while simultaneously serving as its own template for genetic replication from one generation of “protocells” [5, 6] to the next. At the very origin of life, before any molecular machines had evolved to assist in the replication of genetic material, how could information in the form of RNA have been successfully copied and transmitted? One possibility [7] is that the first protocells capitalized on the ability of RNA to selectively direct the position and orientation of incoming monomers through base pairing and stacking alone, circumventing the need for enzymes. The catalytic enhancements derived from nothing more than these template effects may have initiated the minimal mechanisms needed for Darwinian evolution to occur, a primitive state that eventually gave way to modern systems of genetic replication, wherein the role of RNA remains [8] as a remnant of that time. A fundamental understanding of the mechanisms of nonenzymatic template-directed RNA replication (Scheme 1) will provide key insights into the origins of life on Earth.

Scheme 1 
          Schematic example of template-directed RNA synthesis. The first step involves binding of activated monomers (red) to their complementary positions on the template strand (black). Following the initial binding step, the monomers react sequentially with a rate constant kT enhanced by the template. LG stands for “leaving group.”
Scheme 1

Schematic example of template-directed RNA synthesis. The first step involves binding of activated monomers (red) to their complementary positions on the template strand (black). Following the initial binding step, the monomers react sequentially with a rate constant kT enhanced by the template. LG stands for “leaving group.”

The last 60 years have seen great advances [9–12] in the development of artificial template-directed protocols in synthetic organic chemistry, in particular for generating macrocyclic compounds and mechanically interlocked molecules (MIMs) such as catenanes, rotaxanes and knots. The focus of this PAC review will be the mechanistic features of synthetic organic template-directed protocols as well as those of nonenzymatic nucleic acid replication. Both share the same fundamental characteristics with respect to their mechanisms, and each area of research can provide valuable lessons for improving the state of the art of the other. An understanding of the template-directed mechanisms of synthetic and nucleic acid-based chemistries will help chemists design their own artificial molecular replicators, as well as shed light on the origins of life on Earth. From a planetary science perspective [13, 14], the development of totally artificial template-directed, information-encoding protocols will provide insight into the feasibility of alternative chemistries of life operating on planetary bodies other than the Earth. Selected examples of template-directed mechanisms will be discussed.

Fundamentals of template-directed syntheses

Definitions

The dictionary definition of a template is “anything that determines or serves as a pattern.” In the context of biochemistry, the IUPAC Gold Book defines [15, 16] a template as “the nucleic acid single strand that is copied during replication.” This definition implies the transmission of information – the (complementary) molecular structure is being copied as well as the information encoded within that gives rise to function [17]. In the context of synthetic chemistry, the IUPAC Gold Book does not provide a definition of a template. Paraphrasing the Wikipedia entry for artificial template-directed synthesis, a template uses noncovalent bonds to hold together a substrate or substrates placing their respective reactive sites in close proximity to each other, thereby facilitating a desired covalent reaction. This definition is harmonious with the one provided by Busch [18] in 1992: “a chemical template organizes an assembly of atoms, with respect to one or more geometric loci, in order to achieve a particular linking of atoms.” On the surface, these definitions do not seem to have much to do with the copying of information. Are these two conceptual definitions of a template – information storage and spatial localization and organization of reactants – (Scheme 2) fundamentally different from each other? How do these two different concepts of a molecular template relate?

Scheme 2 
            The concept of a molecular template carries with it implications of (1) information storage from the perspective of a molecular biologist, and (2) from the point of view of a supramolecular chemist, a chemical species which uses noncovalent bonds to order two or more substrates in close proximity facitlitating their covalent reaction. These two concepts of a template are related to each other through entropy.
Scheme 2

The concept of a molecular template carries with it implications of (1) information storage from the perspective of a molecular biologist, and (2) from the point of view of a supramolecular chemist, a chemical species which uses noncovalent bonds to order two or more substrates in close proximity facitlitating their covalent reaction. These two concepts of a template are related to each other through entropy.

Information theory, which is often applied [17, 19–22] to oligonucleotide copying, can help provide answers to these questions. The information encoded by a nucleic acid template can be defined [21] as the mutual information, which is equal to the difference in entropy between the initial state, in which the monomeric activated nucleotides are randomly distributed in solution, and the final state wherein they are bound to the template in an ordered sequence through specific noncovalent interactions, i.e., Watson–Crick pairing. In mathematical terms [21]:

I ( T , X ) = Δ S ( T, X )

where T is the sequence of the template strand and X is its complement to be copied, I(T, X) is the mutual information content and ΔS(T, X) is the change in entropy upon monomer binding. The polymerization step via phosphodiester bond formation, which occurs after the initial binding step does not change the sequence of the monomers and so is not considered in the calculation of mutual information. Nevertheless, it is still extremely important for preserving the information after removal of the template. In essence, the information content can be thought as equal to the decrease in entropy brought about by the noncovalent binding of the monomers to the template.

Synthetic template-directed protocols

Synthetic template-directed protocols can be applied [23, 24] in two basic strategies using either thermodynamic or kinetic control. A thermodynamic template influences the thermodynamic distribution of products in a reaction involving totally reversible steps, while a kinetic template favors a specific product by speeding up one or more of the irreversible steps in the reaction pathway leading to that product. The focus of this review is on the kinetic template effect in organic synthesis as this is the most relevant to nonenzymatic RNA replication. For a review [25] of thermodynamically controlled template-directed syntheses, see the reference cited.

Template-directed strategies in organic synthesis have undergone substantial development, particularly in response to the challenge of making macrocycles and MIMs [26] in high yields. Kinetic template-directed protocols have employed a large range of noncovalent interactions to hold together the template and reactive substrates. These templation strategies make use of a wide range of noncovalent bonding interactions, including hydrogen bonding [27, 28], π–π donor–acceptor forces between π-electron-rich and π-electron-poor species [29], metal–ligand coordination [30, 31] and radical pairing [32, 33].

Consider one of the early examples [34] (Fig. 1a) of synthetic template-directed protocols reported [35–40] for crown ethers that makes use of alkali and alkaline earth metals. Following the initial synthesis of crown ethers by Pedersen [41] came the realization that the same metal salts with which these macrocyclic polyethers formed stable complexes could increase the yield of the final ring-closing step. The Mandolini group found that these increased yields [42] were the result of a kinetic effect, i.e., certain metal salts speed up the rate [34] of macrocyclization, while other salts actually slow it down.

Fig. 1 
            (a) The template-directed mechanism of B18C6 synthesis by monovalent metal cations. The metals coordinate to the oligomeric precursor 1, forming the 1⊂M+ complex and placing the phenoxide anion in close proximity to the alkyl bromide. The metal salts accelerate the reaction by binding more tightly to the transition state leading to an enhanced rate constant kT in comparison to the untemplated rate constant k0. (b) Kinetic data showing the rate enhancement for B18C6 synthesis by different monovalent salts as their concentrations are increased. In the case of Li+, the rate is actually decreased as a consequence of the fact that Li+ binds less tightly to the transition state. The “⊂” symbol carries the meaning “complexed with”. Data reprinted with permission from [34]. Copyright 1983 American Chemical Society.
Fig. 1

(a) The template-directed mechanism of B18C6 synthesis by monovalent metal cations. The metals coordinate to the oligomeric precursor 1, forming the 1⊂M+ complex and placing the phenoxide anion in close proximity to the alkyl bromide. The metal salts accelerate the reaction by binding more tightly to the transition state leading to an enhanced rate constant kT in comparison to the untemplated rate constant k0. (b) Kinetic data showing the rate enhancement for B18C6 synthesis by different monovalent salts as their concentrations are increased. In the case of Li+, the rate is actually decreased as a consequence of the fact that Li+ binds less tightly to the transition state. The “⊂” symbol carries the meaning “complexed with”. Data reprinted with permission from [34]. Copyright 1983 American Chemical Society.

Take the case [34, 43] of the crown ether benzo-18-crown-6 (B18C6). The rate of the final ring-closing step of the o-HOC6H4(OCH2CH2)5Br oligomeric precursor 1 (Fig. 1b) is accelerated by Na+ and K+ ions but not so for Li+ ions. The degree of acceleration also depends on the amount of ion in solution, eventually reaching a plateau regime past a certain concentration, and the metal ions which bind the strongest to the B18C6 product also give the greatest rate accelerations during the syntheses. The observed rate constant kobs varies with concentration of the metal template according to the following equation which can be derived [43] from the mechanism shown in Fig. 1a:

(1) k o b s = k 0 + k T K G S [ M + ] γ 1 + K G S [ M + ] γ  (1)

where k0 is the rate constant for the untemplated reaction, kT is the rate constant for the templated reaction, KGS is the binding constant of the substrate in the ground state and γ is an activity coefficient. The metal ions act to coordinate to the O atoms and organize the oligoethers through multiple noncovalent [M+···O] bonds, in addition to electrostatic interactions, ordering the ether chains in an optimal conformation in which to react and form the final B18C6 product. As a result, the B18C6 macrocyclic polyether is then preorganized [44] to bind to that metal ion which templated its synthesis – a fact which implies template-enhanced rates can be coupled to the formation of preorganized products that consequently bind the template more tightly than the substrate. These studies helped to provide a quantitative kinetic understanding of the template effect.

Nonenzymatic template-directed replication of nucleic acids

The replication of RNA without the aid of enzymes or any other information-encoded molecule is an essential [45] premise of the RNA world hypothesis. The template-directed replication of RNA occurs on a preformed RNA strand. Complementary, activated nucleotide monomers or oligomers react with each other forming 3′-5′ phosphodiester linkages, although 2′-5′ linkages [46] can also result. The most well-studied experimental procedures employ [47] 5′-phosphoro-2-methylimidazolides as activated monomers at pH∼8 in the presence of MgCl2 and NaCl. For more information about nonenzymatic RNA [1, 2, 7, 47, 48] and nucleic acid replication [49, 50] in general, see the references cited.

Take for example one of the earliest reported [51] nonenzymatic nucleotide replication experiments (Fig. 2) by Naylor and Gilham in 1966. Using the more available deoxy substrates and templates at the time, they first showed that thymidine oligonucleotides associated with polyadenylic acid. Then, they conducted an experiment which demonstrated that the carbodiimide-activated ligation between two T6 thymidine hexanucleotides, forming a T12 dodecanucleotide, is facilitated by the presence of an A12 dodecanucleotide template. If no A12 template is added to the solution, then the T12 product is not observed. In other words, the template has the effect of increasing the rate of ligation, by first binding to the two T6 hexanucleotides and organizing them for reaction.

Fig. 2 
            The template-directed ligation of two DNA thymidine hexamers T6  bound to a dodecamer A12  template. The first step involves Watson–Crick pairing of two T6  hexamers onto the A12  template forming the A12 ⊂2T6  complex, followed by addition of water-soluble carbodiimide derivatives which activate the 5′-phosphate of the T6  hexamer. The 3′-hydroxyl of the adjacent T6  hexamer can attack at an enhanced rate by virtue of being bound to the template.
Fig. 2

The template-directed ligation of two DNA thymidine hexamers T6 bound to a dodecamer A12 template. The first step involves Watson–Crick pairing of two T6 hexamers onto the A12 template forming the A122T6 complex, followed by addition of water-soluble carbodiimide derivatives which activate the 5′-phosphate of the T6 hexamer. The 3′-hydroxyl of the adjacent T6 hexamer can attack at an enhanced rate by virtue of being bound to the template.

In a more quantitative study, Kanavarioti and coworkers investigated [52] to what extent a template accelerates the rate of formation of a phosphodiester linkage between two nucleotides. They looked at the rate of guanosine dimer formation (Fig. 3a) from guanosine 5′-phosphate-2-methylimidazolide (2-MeImpG) in the absence and presence of polycytidylate (PolyC) as a template. In the absence of the PolyC template, two molecules of 2-MeImpG form three different dimers – 3′-5′ linked, 2′-5′ linked and pyrophosphate linked. The average rate constant of untemplated dimerization k0 for these three processes was determined to be 4.2 × 10–2 M–1 h–1, and the plot of the initial rate of total dimer formation versus the total monomer concentrations reveals a straight line. In the presence of a PolyC template (Fig. 3b), the majority of the dimers are 3′-5′ linked. A plot of the initial rate of dimer formation versus the total concentration of 2-MeImpG no longer gives a linear relationship, but instead gives a curve which is likely sigmoidal – a consequence of cooperative binding of the 2-MeImpG monomers to the PolyC template. By evaluating the data against various different mechanistic models which make the assumption of cooperativity, the authors concluded [52] that the intrinsic rate constant for dimerization on the template kT is 18 × 10–2 h–1, an increase of about 4.3 fold over the untemplated reaction. The ratio of the initial rates of reaction for 3′-5′ linked dimerization on the PolyC template (9 × 10–3 M h–1) and in solution (6.3 × 10–5 M h–1) at 0.1 M 2-MeImpG and 0.05 M PolyC is about 150-fold under these conditions.

Fig. 3 
            (a) Dimerization of the activated 2-MeImpG results in a mixture of 2′-5′, 3′-5′ and pyrophosphate linked products at an average rate constant k0. (b) Dimerization of 2-MeImpG in the presence of a PolyC template results in predominantly 3′-5′ linked dimers along with an enhanced rate constant kT.
Fig. 3

(a) Dimerization of the activated 2-MeImpG results in a mixture of 2′-5′, 3′-5′ and pyrophosphate linked products at an average rate constant k0. (b) Dimerization of 2-MeImpG in the presence of a PolyC template results in predominantly 3′-5′ linked dimers along with an enhanced rate constant kT.

Kinetic templates bind the transition state

The role of a kinetic template is to select the desired reaction pathway by lowering the kinetic barrier to formation of the respective product. In this sense, the template behaves in a fashion similar to a catalyst or an enzyme. When an enzyme binds the substrate(s), and then facilitates a reaction, it does so by binding more tightly to the transition state than the ground state. This idea of enzyme catalysis by tighter transition-state binding was first hypothesized [53] by Linus Pauling in 1948 and later was verified by experiment. According to transition state theory [43], the ratio of the catalyzed rate constant kc to the uncatalyzed rate constant k0 is equal to the binding constant of the enzyme to the transition state K divided by the binding constant to the ground state KGS, or:

(2) K = k c k 0 K G S  (2)

If an enzyme is to accelerate the rate of a reaction, it must bind the transition state more tightly than the ground state. Likewise, so must a template, (Fig. 4).

Fig. 4 
          Reaction coordinate profiles depicting the action of a template T on a substrate S leading to product P. The reaction profiles before and after template binding to the substrate are shown. The template acts to accelerate the reaction by binding more tightly to the transition state, ΔG(K‡), than to the ground state, ΔG(KGS).
Fig. 4

Reaction coordinate profiles depicting the action of a template T on a substrate S leading to product P. The reaction profiles before and after template binding to the substrate are shown. The template acts to accelerate the reaction by binding more tightly to the transition state, ΔG(K), than to the ground state, ΔG(KGS).

The Mandolini group [34] demonstrated experimentally this concept, that an efficient template also will bind more tightly to the transition state than the ground state, and this is the reason for the accelerated rate of reaction. They showed that the same principle of tighter transition-state binding (Fig. 1) is at play with the template-directed synthesis of crown ethers. This formalism explains why different metal cations accelerate the rate of reaction to different degrees, while some actually slow down the rate of reaction. The size of the metal dictates the specific conformation of the superstructure, an ordered complex which in turn determines the relative affinity to the transition state. Equation (2) is implicit in eq. (1), and can be seen readily by dividing both sides by k0 and replacing kc with kT. The ratio kT/k0 is often referred [49] to as the effective molarity, but is formally attributable [54] to either entropic or enthalpic lessening of the energy needed to reach the transition state brought about by the template.

Later on, the Ercolani group [55–61] demonstrated in multiple instances over the course of a decade that the same principle holds true in the template-directed synthesis of the π-electron-poor tetracationic cyclobis(paraquat-p-phenylene) (CBPQT4+) ring, first reported [62] by the Stoddart group, using a variety of π-electron-rich donors as templates. These donors bind more tightly to the transition state involving the final SN2 reaction that leads to the cyclized product than they do to the acyclic substrate, and so accelerate the rate of reaction.

In one study, Ercolani et al. [57] performed a kinetic selection experiment (Fig. 5) between two different templates T1 and T2 with the same acyclic precursor 33+ in the template-directed synthesis of the [2]catenanes C14+ and C24+. Each macrocyclic polyether T1 and T2 possesses different binding affinities KT1 and KT2 to the acyclic substrate as well as different template-accelerated rate constants kT1 and kT2. Assuming that both of the templates are in large excess such that their concentrations do not change significantly throughout the course of the reaction, then the product ratio of the two different [2]catenanes formed from the two competing templates is:

Fig. 5 
          Kinetic selection experiment involving two different templates T1 and T2 in the synthesis of the two catenanes C14+ and C24+ incorporating the CBPQT4+ ring shown in blue. The great affinity KT2 of the T2 template for the 33+ acyclic intermediate together with the larger rate constant kT2 lead to the faster, i.e., selective formation of C24+.
Fig. 5

Kinetic selection experiment involving two different templates T1 and T2 in the synthesis of the two catenanes C14+ and C24+ incorporating the CBPQT4+ ring shown in blue. The great affinity KT2 of the T2 template for the 33+ acyclic intermediate together with the larger rate constant kT2 lead to the faster, i.e., selective formation of C24+.

(3) [ C 1 4 + ] [ C 2 4 + ] = k T 1 K T 1 [ T 1 ] k T 2 K T 2 [ T 2 ]  (3)

This equation very elegantly shows that the relative amount of catenanes C14+ and C24+ formed, given equal amounts of the T1 and T2 templates are present in solution, depends not only on the relative binding constants of the acyclic precursor 33+ to the templates, but also on the relative rate enhancements produced by each template. By dividing the numerator and denominator by k0, one can see that the relative amount of C14+ and C24+ catenanes formed is proportional to the ratio of the respective transition-state binding constants.

Template-product dead ends

In the case of a synthetic kinetic template, the reacted product typically binds stronger than the unreacted substrates. This is likely a consequence of the preorganized [44] structure of the product that is generated from the templated synthesis. In a mechanistically similar fashion, an oligonucleotide template with multiple binding sites reacts until each site in the sequence is copied with its complementary nucleic acid. At the end of the reaction sequence, the template forms a very stable duplex with the preorganized complementary strand. These products of template-directed reactions are trapped [7] in a thermodynamic dead end, with no way to escape, unless an additional source of energy is applied to the system that disturbs the equilibrium.

To illustrate this dead-end effect, consider one of the early examples of oligonucleotide template-directed synthesis, in this case DNA, demonstrated (Fig. 6) by von Kiedrowski and coworkers [63, 64]. Von Kiedrowski employed a self-complementary hexamer DNA template HEX and reacted in its presence two trideoxynucleotides TRI1 and TRI2 using 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (CDI) as an activating agent. He was able to demonstrate a rate enhancement brought about by binding of the two trimers to the HEX template, however, he recognized that the binding constant of the newly formed HEX2 duplex ought to be much greater than that of the two unreacted trimers. This dead-end duplex formation is one of the great challenges that remain [7] towards achieving sustainable nonenzymatic template-directed oligonucleotide replication. Without an additional input of energy to separate the two strands, further replication comes to a standstill.

Fig. 6 
          The template-directed ligation of a self-complementary sequence HEX from two trimeric precursors TRI1 and TRI2. The two trimers first bind to the HEX template. The presence of CDI as an activating agent facilitates the coupling between the 3′-phosphate of TRI1 and the 5′-hydroxyl of TRI2 at a template-enhanced rate constant kT. The resulting HEX2 duplex is more stable than the HEX⊂(TRI1, TRI2) complex, and so the reaction eventually comes to a standstill.
Fig. 6

The template-directed ligation of a self-complementary sequence HEX from two trimeric precursors TRI1 and TRI2. The two trimers first bind to the HEX template. The presence of CDI as an activating agent facilitates the coupling between the 3′-phosphate of TRI1 and the 5′-hydroxyl of TRI2 at a template-enhanced rate constant kT. The resulting HEX2 duplex is more stable than the HEX⊂(TRI1, TRI2) complex, and so the reaction eventually comes to a standstill.

A natural question arises, and that is, are these thermodynamic dead ends a necessity of the mechanisms of information copying employed during nonenzymatic template-directed oligonucleotide synthesis? Given the equation of mutual information [21] in terms of entropy discussed earlier, it is possible to make a mathematical statement regarding the energetic cost of information copying with regard to nucleic acid replication, which can provide some insight. Recall that the mutual information I(T, X) contained in the process of monomers aligning in a specific complementary sequence X along a template T is equal [21] to minus the change in entropy ΔS(T, X) between the free and bound states of the monomers. Since ΔS for monomer binding is very likely negative (although never actually directly measured experimentally), this noncovalent reaction will self-evidently never be entropically driven. In order for monomer binding to proceed spontaneously, it must be associated with an appropriately large negative change in enthalpy ΔH such that ΔG, equal to ΔHTΔS, is negative. As the mutual information and the change in entropy scales [21] with the number of nucleotide monomers, so will the free energy ΔG necessary to make the monomer binding spontaneous.

MIMs such as catenanes and rotaxanes prepared using kinetic template-directed protocols are perhaps the ultimate example of when the reacted product is more stably bound to the template than its substrate precursors. The energy required to separate [12] them is that of the weakest covalent bond in the structure. A case in point is the template-directed synthesis [33] of the homo[2]catenane HC8+ (Fig. 7) composed of two identical mechanically interlocked CBPQT4+ rings. The synthesis is templated by self-complementary π–radical–radical interactions taking place between its constituent bipyridinium units in their radical cation (BIPY·+) forms. The final ring-closing step of the acylic precursor 3·+(+) is templated by the diradical dication CBPQT2(·+) ring, producing an exact copy of itself, and encoding one bit of molecular information. In order for the CBPQT2(·+) ring to produce another copy of itself, the two rings must first be separated. This separation is easier said than done as the two rings are mechanically interlocked. Even after oxidation to its fully charged 8+ state, the [2]catenane HC8+ remains unperturbed.

Fig. 7 
          The template-directed synthesis of the homo[2]catenane HC8+. Radical-radical recognition between the CBPQT2(·+) ring and the acyclic precursor 3(·+)(+) results in a stable inclusion complex that quickly reacts to form the [2]catenane. Even after oxidation to its fully charged 8+ state, the two CBPQT4+ rings cannot be separated.
Fig. 7

The template-directed synthesis of the homo[2]catenane HC8+. Radical-radical recognition between the CBPQT2(·+) ring and the acyclic precursor 3(·+)(+) results in a stable inclusion complex that quickly reacts to form the [2]catenane. Even after oxidation to its fully charged 8+ state, the two CBPQT4+ rings cannot be separated.

Dissociation of the reacted substrates with the template

How can this challenge to replication brought on by dead-end duplex formation associated with template-directed nucleic acid synthesis be overcome? Let us turn our attention again to synthetic organic chemistry for some insights. One potential source of energy that can be used to separate [65] a template (Fig. 8) from its product comes by changing the pH. The synthesis of the CBPQT4+ ring is commonly templated by π-electron-rich donors, one of the most common being 1,5-bis[2-(2-hydroxyethoxy)ethoxy]naphthalene (BHEEN). The association constant [66] of the CBPQT4+ ring with BHEEN is relatively large (3.64×104 M–1, MeCN, 298 K). By first exchanging the template with an alternative 1,5-bis[2-(2-hydroxyethoxy)ethylamino]naphthalene (BHEAN) template using Le Chatelier’s principle, the excess BHEEN and BHEAN can be quickly removed by aqueous extraction with CH2Cl2 or CHCl3. Decreasing the pH in order to protonate the amines, the positively charged BHEAN·2H2+ very quickly is expelled [65] from the cavity of the CBPQT4+ ring on account of Coulombic repulsion. The free CBPQT4+ ring can then be precipitated by counterion exchange and collected by filtration.

Fig. 8 
          The template-directed synthesis of CBPQT4+ followed by template exchange and removal of the exchanged template. The formation of the CBPQT4+ ring is templated by BHEEN, which binds more tightly to the BHEEN⊂CBPQT4+ inclusion complex than the acyclic precursor. By adding (1) an excess of BHEAN, the BHEEN can be readily removed (2) by liquid-liquid extraction. Protonation of the amines result in rapid expulsion of BHEAN·2H2+ from the cavity of CBPQT4+, which can then be collected by precipitation.
Fig. 8

The template-directed synthesis of CBPQT4+ followed by template exchange and removal of the exchanged template. The formation of the CBPQT4+ ring is templated by BHEEN, which binds more tightly to the BHEEN⊂CBPQT4+ inclusion complex than the acyclic precursor. By adding (1) an excess of BHEAN, the BHEEN can be readily removed (2) by liquid-liquid extraction. Protonation of the amines result in rapid expulsion of BHEAN·2H2+ from the cavity of CBPQT4+, which can then be collected by precipitation.

Another example involves the use of redox energy to separate the template from its reacted substrates. The synthesis of the CBPQT4+ ring can be templated [67] by the redox-active π-electron-rich tetrathiafulvalene (TTF). The Stoddart group has shown that the oxidation of the TTF unit to its radical cation or dicationic form results in its expulsion from the cavity of the CBPQT4+ ring on account of Coulombic repulsion. In one study (Fig. 9), the Stoddart group synthesized the TD template containing [68] the TTF unit flanked on either side by oligoethylene glycol chains bearing bulky tert-butylphenoxy groups on their termini. These bulky groups dramatically slow down the rate of threading of the CBPQT4+ ring onto the central TTF unit. Upon oxidation of the TTF unit, the CBPQT4+ ring eventually dissociates from the TD·+ template slipping over the bulking stoppers. Re-reduction of the TTF back to its neutral form reinstates the initial state of the system with a free TD template available for another (potential) round of template-directed synthesis. If the rate of the template-directed synthesis of the CBPQT4+ ring could be made to outcompete the rate of threading, then perhaps this system could be made to undergo sustained replication, at least with respect to the ring component, by making use of a clever implementation of oxidation and reduction cycles.

Fig. 9 
          An oxidation-reduction cycle of the TD⊂CBPQT4+ complex leading to dissociation of the CBPQT4+ ring. Oxidation of the TTF unit results in repulsion of CBPQT4+ towards the end of the dumbbell, where it becomes momentarily trapped by the relatively bulky tert-butylphenoxy stopper. Eventually the ring will slip over the stopper and totally dissociate from TD·+. Reduction of TD·+ back to its initial state results in the formation of the free TD template. The rate of threading of the CBPQT4+ ring is slowed by the presence of the stopper, potentially making TD available for another round of template-directed synthesis of CBPQT4+.
Fig. 9

An oxidation-reduction cycle of the TD⊂CBPQT4+ complex leading to dissociation of the CBPQT4+ ring. Oxidation of the TTF unit results in repulsion of CBPQT4+ towards the end of the dumbbell, where it becomes momentarily trapped by the relatively bulky tert-butylphenoxy stopper. Eventually the ring will slip over the stopper and totally dissociate from TD·+. Reduction of TD·+ back to its initial state results in the formation of the free TD template. The rate of threading of the CBPQT4+ ring is slowed by the presence of the stopper, potentially making TD available for another round of template-directed synthesis of CBPQT4+.

The entire fields of supramolecular and mechanically interlocked molecular switches [26, 69, 70] are a testament to the fact that there are numerous examples of how to reversibly dissociate templates from their reacted substrates. In the context of the mechanical bond, however, this reversible dissociation becomes relative intramolecular motion, and has often been exploited to harness the energy from these well-defined nanoscale movements, leading to applications such as molecular muscles [71], actuators [71], memory [72], unidirectional motion [73–75] and sequence-specific peptide synthesis [76]. For more details about mechanically interlocked molecular switches, see references [26, 69, 70]. In the context of oligonucleotide replication, the most commonly employed method to separate duplexed strands is by thermal denaturation. The rate of re-annealing of the two strands once cooled, however, is faster [7] by orders of magnitude than the rate of nonenzymatic synthesis. Clever ways are needed [7] to slow down the rate of annealing of the duplex and/or increasing the magnitude of the template effect if sustainable molecular replication is to be achieved [77].[1]

Conclusions and future perspectives

Nucleic acid replication and artificial template-directed syntheses share some fundamental mechanistic features. Each requires an initial binding step that organizes the substrate with respect to the template, and the resulting template-enhanced rate is a consequence of tighter noncovalent binding to the transition-state complex. Although the thermodynamic parameters, i.e., ΔH and ΔS, driving the binding of individual monomers [78] to a template in the absence of enzymes have never been directly measured for DNA or RNA, oliognucleotide duplex formation is nearly always an enthalpically driven process [79] mediated by an entropic penalty, that is, a negative ΔS term, which scales with the length of the duplex. It is highly likely that this negative entropy term is also true for individual mononomer binding, the mechanistic step of which determines sequence specificity and so is directly related to information transmission. Likewise, in the context of supramolecular and host-guest chemistries, binding of a host to a guest is often characterized [80] by a negative change in entropy. Although the thermodynamic parameters governing the binding of synthetic templates to their substrates prior to covalent reaction is less well studied, these binding events also likely proceed at an entropic cost, a large amount of which comes from restricted degrees of translational and conformational freedom of the substrate. It is a commonly held notion these entropic losses in degrees of freedom, leading in particular to conformationally less flexible substrates, is a prominent mechanistic feature which to a large extent is responsible [54, 81] for increasing the rate of the templated reaction. The resulting preorganized product, in which the entropically restricted conformation “lives on” [26], binds the template stronger than the initial substrate. Applying this reasoning to nonenzymatic replication of nucleic acid, one can make the hypothesis that the negative change in entropy upon monomer binding to the template, a process crucially connected to information copying, is also responsible, at least in part, for increasing the rate of the subsequent covalent phosphodiester bond formation. If polymerization along the template occurs at a enhanced rate, then the resulting oligonucleotide strand will be preorganized to bind with the template, and so dissociation will become even more unfavorable compared to that of the substrate. Because information, template-enhanced rates and the resulting preorganization of the final product are all coupled to the same initial changes in entropy after binding of the unreacted substrates, the resulting thermodynamically dead-end duplexes are an inherent feature of these nonenzymatic mechanisms of nucleic acid replication used to transmit information.

The development of artificial template-directed kinetic protocols beyond nucleic acids capable of realizing a succession of information-encoding, selective reactions along a template has yet to reach its potential. Mechanically interlocked molecules and supramolecular complexes have well-understood mechanisms of switching that allow for the harnessing of the free energy that results from their stimuli-induced relative motions – movements of which in the appropriate contexts can be interpreted as controllable removal and reassembly of the template. Because of this understanding, the realization of these artificial, successive information-encoding protocols, as well as the application of their switching mechanisms to nucleic acid replication, are foreseeable. Likewise, working to optimize the template effects in nonenzymatic nucleic acid replication in order to enhance selectivity as well as rates will provide valuable insights towards achieving artificial systems which can achieve the same information encoding properties. The ability to design, create and control molecular replication in the lab – whether prebiotically plausible or not – will help us understand the origins of life on Earth as well as obtain insights into the likelihood of it arising elsewhere in the universe [14] based on systems outside of the central dogma.


Article note

A collection of invited, peer-reviewed articles by the winners of the 014 IUPAC-SOLVAY International Award for Young Chemists.



Corresponding author: Albert C. Fahrenbach, Earth-Life Science Institute, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, 152-8551 Tokyo, Japan; and Howard Hughes Medical Institute, Department of Molecular Biology, and Center for Computational and Integrative Biology, Massachusetts General Hospital, Boston, MA 02114, USA, e-mail:

Acknowledgments

I would like to first of all thank IUPAC for providing me with the opportunity to write this review. A special thank you goes out to my PhD advisor Professor Fraser Stoddart for his constant support and encouragement that he continues to provide past my graduate career. I would like to acknowledge my current advisor, Professor Jack Szostak, for his very valuable input, advice and encouragement. Financial support during my PhD was provided by an NSF Graduate Research Fellowship. I would like to thank the Earth-Life Science Institute (ELSI) for the award of a postdoctoral fellowship and Aaron T. Larsen, Nicholas Guttenberg and Nathaniel Virgo for useful discussions.

References

[1] G. F. Joyce. Nature338, 217 (1989).10.1038/338217a0Search in Google Scholar PubMed

[2] G. F. Joyce, L. E. Orgel. “Progress toward understanding the origin of the RNA world”, in The RNA World, R. F. Gesteland, J. F. Atkins, T. R. Cech, (Eds.), pp. 23–56, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (2006).Search in Google Scholar

[3] W. Gilbert. Nature319, 618 (1986).10.1038/319618a0Search in Google Scholar

[4] H. S. Zaher, P. J. Unrau. RNA13, 1017 (2007).10.1261/rna.548807Search in Google Scholar PubMed PubMed Central

[5] S. S. Mansy, J. P. Schrum, M. Krishnamurthy, S. Tobé, D. A. Treco, J. W. Szostak. Nature454, 122 (2008).10.1038/nature07018Search in Google Scholar PubMed PubMed Central

[6] T. F. Zhu, J. W. Szostak. J. Am. Chem. Soc.131, 5705 (2009).Search in Google Scholar

[7] J. W. Szostak. J. Syst. Chem.3:2, 1 (2012).10.1186/1759-2208-3-2Search in Google Scholar

[8] M. Krupkin, D. Matzov, H. Tang, M. Metz, R. Kalaora, M. J. Belousoff, E. Zimmerman, A. Bashan, A. Yonath. Phil. Trans. R. Soc. B366, 2972 (2011).10.1098/rstb.2011.0146Search in Google Scholar PubMed PubMed Central

[9] J. F. Stoddart. “Template-directed synthesis of new organic materials”, in Frontiers in Supramolecular Organic Chemistry and Photochemistry, H.-J. Schneider, H. Dürr, (Eds.), pp. 251–263, VCH, Weinheim (1991).Search in Google Scholar

[10] E. A. Wintner, J. Rebek Jr. Acta Chem. Scand.50, 469 (1996).Search in Google Scholar

[11] F. M. Raymo, J. F. Stoddart. “Templated synthesis of catenanes and rotaxanes”, in Templated Organic Synthesis, F. Diederich, P. J. Stang, (Eds.), pp. 75–104, Wiley-VCH, Weinheim (2000).10.1002/chin.200045272Search in Google Scholar

[12] J.-C. Olsen, K. E. Griffiths, J. F. Stoddart. “A short history of the mechanical bond”, in From Non-Covalent Assemblies to Molecular Machines, J.-P. Sauvage, P. Gaspard, (Eds.), pp. 65–139, Wiley-VCH, Weinheim (2011).10.1002/9783527632817.ch8Search in Google Scholar

[13] D. D. Sasselov. Nature451, 29 (2008).10.1038/451029aSearch in Google Scholar PubMed

[14] A. W. Howard, G. W. Marcy, S. T. Bryson, J. M. Jenkins, J. F. Rowe, N. M. Batalha, W. J. Borucki, D. G. Koch, E. W. Dunham, T. N. Gautier III, J. V. Cleve, W. D. Cochran, D. W. Latham, J. J. Lissauer, G. Torres, T. M. Brown, R. L. Gilliland, L. A. Buchhave, D. A. Caldwell, J. Christensen-Dalsgaard, D. Ciardi, F. Fressin, M. R. Haas, S. B. Howell, H. Kjeldsen, S. Seager, L. Rogers, D. D. Sasselov, J. H. Steffen, G. S. Basri, D. Charbonneau, J. Christiansen, B. Clarke, A. Dupree, D. C. Fabrycky, D. A. Fischer, E. B. Ford, J. J. Fortney, J. Tarter, F. R. Girouard, M. J. Holman, J. A. Johnson, T. C. Klaus, P. Machalek, A. V. Moorhead, R. C. Morehead, D. Ragozzine, P. Tenenbaum, J. D. Twicken, S. N. Quinn, H. Isaacson, A. Shporer, P. W. Lucas, L. M. Walkowicz, W. F. Welsh, A. Boss, E. Devore, A. Gould, J. C. Smith, R. L. Morris, A. Prsa, T. D. Morton, M. Still, S. E. Thompson, F. Mullally, M. Endl, P. J. MacQueen. Astrophys. J. Suppl. S.201, 1 (2012).Search in Google Scholar

[15] B. Nagel, H. Dellweg, L. M. Gierasch. Pure Appl. Chem.64, 143 (1992).Search in Google Scholar

[16] IUPAC. Compendium of Chemical Terminology Gold Book. (2014), vol. 2.3.3.Search in Google Scholar

[17] R. M. Hazen, P. L. Griffin, J. M. Carothers, J. W. Szostak. Proc. Natl. Acad. Sci. USA104, 8574 (2007).10.1073/pnas.0701744104Search in Google Scholar PubMed PubMed Central

[18] D. H. Busch. J. Inclusion Phenom. Mol. Recognit. Chem.12, 389 (1992).10.1007/BF01053876Search in Google Scholar

[19] T. D. Schneider, G. D. Stormo, L. Gold, A. Ehrenfeucht. J. Mol. Biol.188, 415 (1986).Search in Google Scholar

[20] E. Smith. J. Theor. Biol.252, 185 (2008).10.1016/j.jtbi.2008.02.010Search in Google Scholar PubMed

[21] J. R. Arias-Gonzalez. PLoS One7, e42272 (2012).10.1371/journal.pone.0042272Search in Google Scholar PubMed PubMed Central

[22] G. F. Joyce. PLoS Biol.10, e1001323 (2012).10.1371/journal.pbio.1001323Search in Google Scholar PubMed PubMed Central

[23] M. C. Thompson, D. H. Busch. J. Am. Chem. Soc.86, 213 (1964).Search in Google Scholar

[24] S. Anderson, H. L. Anderson, J. K. M. Sanders. Acc. Chem. Res.26, 469 (1993).Search in Google Scholar

[25] P. T. Corbett, J. Leclaire, L. Vial, K. R. West, J.-L. Wietor, J. K. M. Sanders, S. Otto. Chem. Rev.106, 3652 (2006).Search in Google Scholar

[26] J. F. Stoddart. Chem. Soc. Rev.38, 1802 (2009).10.1039/b819333aSearch in Google Scholar PubMed

[27] C. A. Schalley, T. Weilandt, J. Brüggemann, F. Vögtle. “Hydrogen-bond-mediated template synthesis of rotaxanes, catenanes, and knotanes”, in Templates in Chemistry I, C. A. Schalley, F. Vögtle, K. H. Dötz, (Eds.), pp. 141–200, Springer-Verlag, Berlin (2004).10.1007/b99913Search in Google Scholar

[28] E. R. Kay, D. A. Leigh. “Hydrogen bond-assembled synthetic molecular motors and machines”, in Molecular Machines, T. R. Kelly, (Ed.), pp. 133–177, Springer-Verlag, Berlin (2005).10.1007/128_011Search in Google Scholar

[29] G. Barin, A. Coskun, M. M. G. Fouda, J. F. Stoddart. ChemPlusChem77, 159 (2012).10.1002/cplu.201100075Search in Google Scholar

[30] D. H. Busch, N. A. Stephenson. Coord. Chem. Rev.100, 119 (1990).Search in Google Scholar

[31] J. E. Beves, B. A. Blight, C. J. Campbell, D. A. Leigh, R. T. McBurney. Angew. Chem. Int. Ed.50, 9260 (2011).Search in Google Scholar

[32] J. M. Spruell. Pure Appl. Chem.82, 2281 (2010).10.1351/PAC-CON-10-08-03Search in Google Scholar

[33] J. C. Barnes, A. C. Fahrenbach, D. Cao, S. M. Dyar, M. Frasconi, M. A. Giesener, D. Benítez, E. Tkatchouk, O. Chernyashevskyy, W. H. Shin, H. Li, S. Sampath, C. L. Stern, A. A. Sarjeant, K. J. Hartlieb, Z. Liu, R. Carmieli, Y. Y. Botros, J. W. Choi, A. M. Slawin, J. B. Ketterson, M. R. Wasielewski, W. A. Goddard, III, J. F. Stoddart. Science339, 429 (2013).10.1126/science.1228429Search in Google Scholar PubMed

[34] G. Illuminati, L. Mandolini, B. Masci. J. Am. Chem. Soc.105, 555 (1983).Search in Google Scholar

[35] J. Dale, P. O. Kristiansen. Acta Chem. Scand.26, 1471 (1972).Search in Google Scholar

[36] R. N. Green. Tetrahedron Lett.13, 1793 (1972).10.1016/S0040-4039(01)85270-9Search in Google Scholar

[37] C. J. Pedersen. Org. Synth.52, 66 (1972).10.15227/orgsyn.052.0066Search in Google Scholar

[38] K. Madan, D. J. Cram. J. Chem. Soc., Chem. Commun. 427 (1975).10.1039/c39750000427Search in Google Scholar

[39] R. T. Gray, D. N. Reinhoudt, C. J. Smit, I. Veenstra. Recl. Trav. Chim. Pays-Bas95, 258 (1976).10.1002/recl.19760951103Search in Google Scholar

[40] D. N. Reinhoudt, R. T. Gray, C. J. Smit, I. Veenstra. Tetrahedron32, 1161 (1976).10.1016/0040-4020(76)85040-5Search in Google Scholar

[41] C. J. Pedersen. J. Am. Chem. Soc.89, 7017 (1967).10.1021/ja01002a035Search in Google Scholar

[42] L. Mandolini, B. Masci. J. Am. Chem. Soc.99, 7709 (1977).Search in Google Scholar

[43] R. Cacciapaglia, L. Mandolini. Chem. Soc. Rev.22, 221 (1993).Search in Google Scholar

[44] D. J. Cram. Angew. Chem. Int. Ed. Engl.25, 1039 (1986).10.1002/anie.198610393Search in Google Scholar

[45] L. E. Orgel. Crit. Rev. Biochem. Mol. Biol.39, 99 (2004).Search in Google Scholar

[46] A. E. Engelhart, M. W. Powner, J. W. Szostak. Nat. Chem.5, 390 (2013).Search in Google Scholar

[47] I. A. Kozlov, L. E. Orgel. Mol. Biol.34, 781 (2000).Search in Google Scholar

[48] M. P. Robertson, G. F. Joyce. Cold Spring Harb. Perspect. Biol.4, a003608 (2012).Search in Google Scholar

[49] Y. Brudno, D. R. Liu. Chem. Biol.16, 265 (2009).Search in Google Scholar

[50] J. Berger, M. Oberhuber. Chem. Biodivers.7, 2581 (2010).Search in Google Scholar

[51] R. Naylor, P. T. Gilham. Biochem.5, 2722 (1966).Search in Google Scholar

[52] A. Kanavarioti, E. E. Baird, T. B. Hurley, J. A. Carruthers, S. Gangopadhyay. J. Org. Chem.64, 8323 (1999).Search in Google Scholar

[53] L. Pauling. Nature161, 707 (1948).10.1038/161707a0Search in Google Scholar PubMed

[54] M. I. Page. Chem. Soc. Rev.2, 295 (1973).10.1039/cs9730200295Search in Google Scholar

[55] S. Capobianchi, G. Doddi, G. Ercolani, J. W. Keyes, P. Mencarelli. J. Org. Chem.62, 7015 (1997).Search in Google Scholar

[56] S. Capobianchi, G. Doddi, G. Ercolani, P. Mencarelli. J. Org. Chem.63, 8088 (1998).Search in Google Scholar

[57] C. D’Acerno, G. Doddi, G. Ercolani, P. Mencarelli. Chem. Eur. J.6, 3540 (2000).Search in Google Scholar

[58] G. Doddi, G. Ercolani, S. Franconeri, P. Mencarelli. J. Org. Chem.66, 4950 (2001).Search in Google Scholar

[59] C. D’Acerno, G. Doddi, G. Ercolani, S. Franconeri, P. Mencarelli, A. Piermattei. J. Org. Chem.69, 1393 (2004).Search in Google Scholar

[60] G. Doddi, G. Ercolani, P. Mencarelli, A. Piermattei. J. Org. Chem.70, 3761 (2005).Search in Google Scholar

[61] G. Doddi, G. Ercolani, P. Mencarelli, G. Papa. J. Org. Chem.72, 1503 (2007).Search in Google Scholar

[62] B. Odell, M. V. Reddington, A. M. Z. Slawin, N. Spencer, J. F. Stoddart, D. J. Williams. Angew. Chem. Int. Ed. Engl.27, 1547 (1988).Search in Google Scholar

[63] G. von Kiedrowski. Angew. Chem. Int. Ed. Engl.25, 932 (1986).10.1002/anie.198609322Search in Google Scholar

[64] G. von Kiedrowski, B. Wlotzka, J. Helbing. Angew. Chem. Int. Ed. Engl.28, 1235 (1989).Search in Google Scholar

[65] C.-H. Sue, S. Basu, A. C. Fahrenbach, A. K. Shveyd, S. K. Dey, Y. Y. Botros, J. F. Stoddart. Chem. Sci.1, 119 (2010).Search in Google Scholar

[66] J. W. Choi, A. H. Flood, D. W. Steuerman, S. Nygaard, A. B. Braunschweig, N. N. Moonen, B. W. Laursen, Y. Luo, E. DeIonno, A. J. Peters, J. O. Jeppesen, K. Xu, J. F. Stoddart, J. R. Heath. Chem. Eur. J.12, 261 (2006).Search in Google Scholar

[67] M. Asakawa, P. R. Ashton, V. Balzani, A. Credi, C. Hamers, G. Mattersteig, M. Montalti, A. N. Shipway, N. Spencer, J. F. Stoddart, M. S. Tolley, M. Venturi, A. J. P. White, D. J. Williams. Angew. Chem. Int. Ed.37, 333 (1998).Search in Google Scholar

[68] V. Balzani, A. Credi, G. Mattersteig, O. A. Matthews, F. M. Raymo, J. F. Stoddart, M. Venturi, A. J. P. White, D. J. Williams. J. Org. Chem.65, 1924 (2000).Search in Google Scholar

[69] E. R. Kay, D. A. Leigh, F. Zerbetto. Angew. Chem. Int. Ed.46, 72 (2007).Search in Google Scholar

[70] A. C. Fahrenbach, S. C. Warren, J. T. Incorvati, A.-J. Avestro, J. C. Barnes, J. F. Stoddart, B. A. Grzybowski. Adv. Mater.25, 331 (2013).Search in Google Scholar

[71] A. Coskun, M. Banaszak, R. D. Astumian, J. F. Stoddart, B. A. Grzybowski. Chem. Soc. Rev.41, 19 (2012).Search in Google Scholar

[72] A. Coskun, J. M. Spruell, G. Barin, W. R. Dichtel, A. H. Flood, Y. Y. Botros, J. F. Stoddart. Chem. Soc. Rev.41, 4827 (2012).Search in Google Scholar

[73] D. A. Leigh, J. K. Y. Wong, F. Dehez, F. Zerbetto. Nature424, 174 (2003).10.1038/nature01758Search in Google Scholar PubMed

[74] A. Arduini, R. Bussolati, A. Credi, S. Monaco, A. Secchi, S. Silvi, M. Venturi. Chem. Eur. J.18, 16203 (2012).Search in Google Scholar

[75] H. Li, C. Cheng, P. R. McGonigal, A. C. Fahrenbach, M. Frasconi, W.-G. Liu, Z. Zhu, Y. Zhao, C. Ke, J. Lei, R. M. Young, S. M. Dyar, D. T. Co, Y.-W. Yang, Y. Y. Botros, W. A. Goddard, III, M. R. Wasielewski, R. D. Astumian, J. F. Stoddart. J. Am. Chem. Soc.135, 18609 (2013).Search in Google Scholar

[76] B. Lewandowski, G. De Bo, J. W. Ward, M. Papmeyer, S. Kuschel, M. J. Aldegunde, P. M. E. Gramlich, D. Heckmann, S. M. Goldup, D. M. D’Souza, A. E. Fernandes, D. A. Leigh. Science339, 189 (2013).10.1126/science.1229753Search in Google Scholar PubMed

[77] D. Y. Zhang, G. Seelig. Nat. Chem.3, 103 (2011).Search in Google Scholar

[78] E. Kervio, B. Claasen, U. E. Steiner, C. Richert. Nucleic Acids Res.42, 7409 (2014).Search in Google Scholar

[79] E. T. Kool. Chem. Rev.97, 1473 (1997).10.1021/cr9603791Search in Google Scholar PubMed PubMed Central

[80] H.-J. Schneider. Angew. Chem. Int. Ed. Engl.30, 1417 (1991).10.1002/anie.199114171Search in Google Scholar

[81] G. W. Gokel, H. D. Durst. Synthesis-Stuttgart, 168 (1976).10.1055/s-1976-23980Search in Google Scholar

Published Online: 2015-02-04
Published in Print: 2015-02-01

©2015 IUPAC & De Gruyter

Downloaded on 29.3.2024 from https://www.degruyter.com/document/doi/10.1515/pac-2014-1004/html
Scroll to top button