FLOW CHEMISTRY APPLIED TO THE PREPARATION OF SMALL MOLECULES POTENTIALLY USEFUL AS THERAPEUTIC AGENTS

In recent years, despite the large amount of novel and clinically validated targets identified from the human genome project, the number of new drug launched on the market is decreasing and the overall costs for the development of a drug are rising significantly. Pharmaceutical and biotechnology companies are under a strong pressure to produce a steady stream of innovative, well-differentiated drugs with a reduced cost both for discovery and development. Currently it takes an estimated 10-14 years to develop and market a drug at a cost that exceeds 1 billion dollars. With the aim at increasing the productivity of original and highly pure molecules as potential modulators of therapeutic targets, different and novel technologies, related to synthesis, work-up and isolation, were developed. In particular the so called “Enabling Techniques” have emerged and were studied in a large extent in Academia. Among these new technologies continuous flow organic synthesis is now being investigated widely in fine chemistry and, with the advent of commercially available microreactors, also in pharmaceutical industry. In the framework of my PhD thesis exploring the application of the so called “Enabling Techniques” in a medicinal chemistry laboratory, my efforts were devoted to the evaluation of the benefits that continuous flow chemistry could provide in Drug Discovery programs and in the synthesis of natural products in comparison with traditional synthetic techniques. Flow technologies have recently received a great deal of attention and a fair number of scientific publications have demonstrated their potential for improving productivity in organic synthesis. Established continuous flow chemistry advantages include precise control of temperature, pressure, concentration, residence time and heat transfer. All these aspects significantly affect the reaction outcome improving yield and selectivity. Within my thesis, continuous flow chemistry was firstly applied to the synthesis of hydroxamic acids, a class of well known inhibitors of important biological targets such as metalloproteinases and histone deacetylases. As a part of a medicinal chemistry project, a simple conversion of ester into hydroxamic acids (Scheme 1) was envisaged as a suitable and convenient synthetic method for the preparation of a collection of compounds featuring such privileged substructure. The effects of flow rate, reactor volume and temperature were examined and the optimized reaction conditions were then successfully applied for the preparation of a small collection of ten hydroxamic acids featuring a range of functional groups. Good yields, purity and high reproducibility were observed using this simple protocol. R = Aryl, Alkyl, Heteroaryl, Aminoalkyl; R' = Me, Et Scheme 1. Synthesis of Hydroxamic Acids No racemisation occurred when the reaction was performed on protected amino acids. The yields were comparable and, in some cases, even better than what reported in literature where the same transformation was performed by MW irradiation. Even if the reaction time is relatively longer than with MW, no limitation in scale-up is present using flow chemistry. Based on the good results obtained in the development of the continuous flow synthesis of hydroxamic acids this new methodology was applied to the synthesis of SAHA (suberoylanilide hydroxamic acid). Our two-step sequence entails the conversion of the commercially available methyl suberoyl chloride into methyl suberanilate under Schotten-Baumann conditions, followed by the transformation of ester by aqueous hydroxylamine in presence of sodium methoxide (Scheme 2). Scheme 2. Synthesis of SAHA (suberoylanilide hydroxamic acid) To avoid a time consuming work-up procedure and extensive manual purification of the final compound, an integrated sequential flow synthetic pathway was set-up employing immobilized scavenger. The reaction stream was directly passed through a short packed column containing silica supported quaternary amine for the selective removal of the carboxylic acid by-product. The solution containing the product and traces of the starting material was collected and, after solvent evaporation, crystallization from MeOH afforded SAHA in 84% yield and 99% purity (80% yield over two steps). With the aim at studying the applicability of flow technique also to the synthesis of a natural compound, the continuous flow multi-step synthesis of Dumetorine was undertaken. (+)-Dumetorine, isolated in 1985 from the tubers of Dioscorea dumetorum Pax, shows a notable use in folk medicine and arrow poisons. Its total batch-synthesis was recently published by our group (Scheme 3). Scheme 3. Batch-Total synthesis of Dumetorine The planned synthetic route envisaged a flow process where the synthetic steps were combined into only one continuous sequence minimizing work up and purifications. The reaction crudes had to be processed through packed columns containing immobilized reagents, catalysts, scavengers or catch and release agents. In this way the improvements gained through the precise control of reaction conditions and the reduction in manual handling could be easily evaluated. To synthesize this natural alkaloid, new protocols had to be developed for performing classical reactions under continuous flow conditions. With this aim the flow addition of Grignard reagent to carbonyl compounds was the first reaction that we broadly investigated. In fact, despite many classes of reaction was successfully transferred to continuous flow approach, the addition of Grignard reagents to aldehydes and ketones under flow conditions is up-to-date poorly exemplified in the literature. The optimization of the experimental parameters was investigated by varying the temperature of the stored solution, the residence time and the number of Grignard equivalents. A short column containing polymer supported benzaldehyde was used for the scavenging of the excess of Grignard reagent. The optimized conditions (room teperature; 1.2 Grignard eq; residence time 33 minutes) were successfully applied on different aldehydes and ketones (arylic, heteroarylic, alkylic etc.) for the preparation of a small collection of alcohols (Scheme 5). Good yields (ranging from 88% to 96%), purity and high reproducibility were observed. Scheme4. Flow Grignard addition to carbonyl compounds The protocol was applied to the synthesis of Tramadol, a well known centrally active analgesic used for treating moderate to severe pain (Yield 96%). The developed conditions also allowed the selective addition of Grignard reagents to aldehydes and ketones in the presence of a nitrile function (Scheme 5). Scheme5. Flow Grignard addition to carbonyl compounds in presence of nitriles. In the light of the interesting results concerning the flow addition of Grignard reagent to carbonyl compounds, we started to perform the five-step continuous flow synthesis of (+) Dumetorine reported below (Scheme 6). Scheme 6. Flow-Total synthesis of (+)Dumetorine This synthesis entailed IBX oxidation of primary alcohol, Grignard addition on carbonyl compounds, acylation, ring closing metathesis and Eschweiler-Clarke reductive amination. In the second step the flow addition of suitable Grignard reagent to aldehyde (2) was performed under the previously optimized conditions. The yield of the isolated compound was 90% with a remarkable improvement respect to the batch process. This result is a direct consequence of the efficient mixing and heat dispersion due to the high surface area-to-volume ratios in the PTFE tubing that keeps the temperature constant minimizing the occurrence of side reactions. Compound (5) presents the structural features for ring-closing metathesis and this type of reaction was performed in batch using 2nd generation Grubbs catalyst in good yield. The same result was obtained flowing for short time (less than 20 min) the starting material in presence of dissolved catalyst (both 1st and 2nd generation catalysts were tested) but, in this way, the problem of the final purification by flash chromatography was still present. After a few failing attempts using not commercially available supported Grubbs catalysts of 1st (a) and 2nd (b) generation (Figure 1) , we evaluated the application to flow chemistry of a homogeneous PEG supported Grubbs catalyst in order to increase the performance in RCM maintaining the possibility of a simple catalyst recovery. Figure 1. Supported Grubbs Catalysts prepared A newly synthesized PEG- Supported Hoveyda catalyst (8) was prepared (Scheme 5) in collaboration with Prof. M. Benaglia and Dr. A. Caselli (Università degli Studi di Milano). Using this PEG-supported catalyst, the flow RCM was successfully and we observed the total conversion of (5) in (6). The pure compound (6) was simply obtained by evaporation of the solvent after the precipitation of the catalyst in presence of Et2O. The catalyst easily recovered can be recycled for RCM reactions. Scheme 5. Synthesis of PEG-supported Hoveyda catalyst The Dumetorine batch synthesis was affected by a low yielding acid catalyzed cleavage of Boc protecting group due to a Michael side reaction of the secondary amine on the α-β unsaturated lactone ring. As a consequence, the reductive amination resulted not an easy task to be managed. In order to overcome these problems, we performed the unprecedented flow Eschweiler-Clarke reaction, a particular amination reduction. Under optimized continuous flow conditions, we assessed the concomitant BOC deprotection and N-methylation. The solutions of starting materials and reagents were pumped in the PTFE tubing reactor and then through a column containing SCX cartridge (catch and release purification) obtaining (+)-Dumetorine in high purity and good yield (Overall yield: 29% (65% diastereoisomeric mixture); obtained (+)-Dumetorine amount: 227 mg). So, a flow-based synthesis of (+)-Dumetorine was accomplished; remarkable results were obtained in the Grignard reaction, performed in an efficient and safe manner at room temperature avoiding cryogenic temperature; in the RCM reaction, carried out with a newly synthesised PEG-supported Hoveyda catalyst and finally in the unprecedented flow Eschweiler-Clarke reaction with concomitant BOC deprotection and N-methylation in high yield. This flow total synthesis represents a significant improvement over the existing protocol characterized by lower yield and more steps and the synthetic route was also tested for the preparation of additional analogue derivatives. In fact on the basis of the results that we obtained in the synthesis of (+)-Dumetorine, we applied flow technology to the preparation of its simplified natural congeners (+)-Sedridine (9) and (-)-Sedamine (10). The synthesis of the two natural alkaloids was assessed with good results using protocols (Grignard addition, Eschweiler Clarke reaction) optimized for the preparation of (+)-Dumetorine. Figure 2. (+)-Sedridine (9) and (-)-Sedamine (10) All the results that were assessed in this PhD thesis clearly demonstrate how flow chemistry shows great potentiality in the Medicinal Chemistry field and how that this technique is of great advantage in the assembly of challenging molecules, as natural products, in terms of overall yield reaction time and limitation of handling and purification.