Nat. Chem.: a universally programmable chemical synthesis machine

 Organic synthesis is not only highly specialized but also "labor-intensive." Many workers in the field of organic synthesis laugh at themselves that their daily work is "moving bricks." In contrast, small molecule synthesis automation will greatly liberate productivity. However, due to the lack of general standards, only individual reactions can be automated, such as the automatic synthesis of oligopeptides, oligonucleotides, and oligosaccharides, as well as the MIDA-boronate-based Suzuki-Miyaura coupling reaction of Martin Burke's group. All of these methods are based on successive iterations of a small number of strong responses, and as such, they generally cannot be programmed, and the unit operations cannot be reused.

 Medicilon’s Synthetic Chemistry Services

Medicilon can undertake the synthesis of special reagents, intermediates and molecular fragments, preparation of standard products, synthesis design and preparation of impurities or metabolites, synthesis of stable isotope internal standards and synthesis of tritiated compounds.

At present, chemical synthesis automation platforms based on different strategies have been established, such as the iterative coupling-based synthesis machine from UIUC's Martin Burke research group (Science, 2015, 347, 1221), Pfizer's flow chemistry-based system (Science, 2018, 359, 429), and the Computer synthesis robot based on the traditional chemical experiment process of Leroy Cronin's research group (Science, 2019, 363, eaav2211). But most automated platforms require custom hardware for each reaction, so it isn't easy to automate the synthesis of target molecules by cascading multiple responses on a single machine. Recently, the Leroy Cronin research group from the University of Glasgow in the UK published a research paper on Nature Chemistry, demonstrating that Computers can be programmed to perform many different reactions in a unified system, including solid-phase peptide synthesis (SPPS) ), iterative cross-coupling (ICC) and synthesis of unstable diazoxide. The focus here is on developing general-purpose and modular hardware that can be automated using a single software system. In experiments, the authors' approach performed about 8,500 operations, reusing only 22 different steps across ten modules, and the code could support 17 different reactions.

Figure 1. Integrating disparate synthesis methods into a unified programmable platform. D, deprotection; C, coupling; P, purification. Image credit: Nat. Chem.

Previously, Chemputer had demonstrated that various molecules could be synthesized automatically on the same hardware. However, integrating the different existing automation strategies remains a significant challenge. The authors believe there is a need to develop a general approach that makes the system programmable and modular, unify the various synthetic steps, and thus enable the synthesis of almost any molecule that can be artificially synthesized. As shown in Figure 1, the authors demonstrate how the Computer system was developed such that a single system can automate MIDA-boronate-based iterative cross-coupling reactions, solid-phase peptide synthesis, and the synthesis of photo transducers with essential applications in biology and materials science. Joint agent NHS-diazirine.

Related Articles:

Medicilon Intermediate Synthesis Service

3D Medicine's First Independently Developed New Drug Approved For Clinical Use

Analysis of the Process and Problems of API Process Amplification

The first principle of the author's strategy is to simulate the workflow of artificial organic synthesis. Therefore, the first step is to automate in a unified manner utilizing the commonly used and standardized synthetic chemistry unit operations such as separation, filtration, evaporation, heating, cooling, stirring, etc. To achieve this, it is critical to define a set of unit operations that can be described in an unambiguous chemical language that can be interpreted and executed by standard automated chemistry hardware. As shown in Figure 2, the standard hardware required for the Suzuki–Miyaura coupling reaction is clear. The unit operations (feeding, stirring, heating, rotary evaporation) performed by this hardware can be performed independently and combined conveniently. This highly modular approach makes the architecture easily extensible, making it easy to change one module without affecting other system parts. This feature also means that it is easy to control the operation of each unit through code, making the system programmable.

Figure 2. Mapping of specific reactions to generic automated laboratory hardware assemblies. Image credit: Nat. Chem.

As shown in Figure 3, the first step in an automated synthesis is to analyze the basic unit operations required for the reaction. Once these unit operations are determined, each unit operation can be connected to the corresponding hardware module according to the given graphics file through the software interface. Once the preparation is complete, Chemputer can automatically perform the appropriate actions for each reaction. The Suzuki–Miyaura coupling reaction consists of three steps: deprotection, coupling reaction, and purification. The coupling reaction requires a strict anhydrous and oxygen-free environment. The combination of the unit modules designed by the author can meet the requirements of anhydrous and anaerobic conditions. The author finally synthesized 0.33 g with a yield of 52%, close to the reported 61%. Synthesis of the photocrosslinking NHS-diazirine requires an anhydrous environment and temperature- and light-sensitive reagents, which is a big challenge for automatic synthesis. To this end, the authors developed a reagent module that allows cooling, stirring, and maintaining an inert atmosphere, thus ensuring that sensitive reagents can be stored at low temperatures until use. Chemputer synthesized 4.2 g of the product with a yield of 21%, which is close to the 27% yield reported in the article. Solid-phase peptide synthesis is a repeated process of adding reagents, filtering, washing, and drying. The assembled peptide (9) is separated from the resin, and the peptide product is obtained as a solution in the TFA mixture. To isolate pure peptides from the mixture, precipitation with diethyl ether at low temperatures is required, followed by separation by filtration. Chemputer automatically synthesized 10a, 10b, and 10c in yields of 10%, 48%, and 50%, respectively. This result is comparable to the work and purity obtained with highly specialized SPPS systems. It should be pointed out that the commercial SPPS system cannot automatically perform lysis, deprotection, and precipitation operations, and these steps must be performed manually.

Figure 3. Automated synthesis reactions are involved in this paper. (a) Iterative MIDA borate cross-coupling reaction. (b) Synthesis of photo crosslinker NHS-diazirine. (c) After SPPS, the peptide was reacted with NHS-diazirine. Image credit: Nat. Chem.

 

Figure 4. Available hardware modules. Image credit: Nat. Chem.

Figure 5. The realization and structure of the automatic synthesis platform. Synthesize the hardware module (a) of NHS-diazirine and the diagram (b) on the corresponding software. Image credit: Nat. Chem.

Table 1. Three synthesis results. Image credit: Nat. Chem.

 

This work demonstrates that a single platform architecture can be used to unify the automated synthesis of different classes of molecules through a modular approach. The other chemical reactions required by these molecules can be accommodated with only minor modifications to the actual hardware instance of the system. For example, the NHS-diazirine (7) obtained in case 2 can be used in the peptide synthesis of case 3. Moreover, the cost of automatic synthesis with a Chemputer is not high. NHS-diazirine costs £221/50 mg, while a regular automated peptide synthesizer costs around £50,000. Chemputer costs £20,000 each, and synthetically obtained NHS-diazirine (4.2 g) costs around £15,000. As long as the synthesis of NHS-diazirine is repeated twice, the market value of production has exceeded the cost of Computer hardware.

Comment

Compared with the previous work on automatic synthesis machines, the biggest highlight is that it can continuously perform multiple steps of different reactions on the same platform to obtain the final product. It is expected to establish a unified experimental operation standard to solve the problem of experimental repeatability. However, for organic chemists, exploring reaction conditions and passing columns is undoubtedly the most time-consuming. However, the synthesis machine in this article is only suitable for reactions with complete synthesis conditions, and then the responses are standardized and streamlined. But it looks powerless for the more time-consuming and critical process of exploring requirements.