ARTICLESTotal synthesis of marine natural productswithout using protecting groupsPhil S. Baran1, Thomas J. Maimone1& Jeremy M. Richter1The field of organic synthesis has made phenomenal advances in the past fifty years, yet chemists still struggle to designsynthetic routes that will enable them to obtain sufficient quantities of complex molecules for biological and medical studies.Total synthesis is therefore increasingly focused on preparing natural products in the most efficient manner possible. Herewe des cribe the preparative-scale, enantioselective, total syntheses of members of the hapalindole, fischerindole,welwitindolinone and ambiguine families, each constructed without the need for protecting groups—the use of such groupsadds considerably to the cost and complex ity of synthese s. As a consequence, molecules that have previously requiredtwenty or more steps to synthesize racemically in milligram amounts can now be obtained as single enantiomers insignificant quantities in ten steps or less. Through the extension of the general principles demonstrated here, it should bepossible to access other complex molecular architectures without using protecting groups.Although the field of total synthesis1,2has made great advances since1828 (ref. 3), it is still far from being a mature or applied science4,5.For example, precise control over the individual reactivity of func-tional groups within a complex molecular architecture (chemoselec-tivity) still remains a largely unanswered challenge. Historically, theuse of protecting groups has been the standard solution to this prob-lem because they allow functional groups to be dealt with on anindividual basis. Indeed, these functionality masks have permeatedorganic chemistry to the extent that textbooks state that avoidingthem is impossible6,7. Their use has become routine even on mole-cules of low complexity8. Ideally, protecting groups are easilyappended, allow one to smoothly perform the initially intendedtransformation, and then gracefully depart without incident. In prac-tice, however, these artificial devices add at least two steps each to asynthetic sequence and sometimes dramatically lower the efficiencyof a synthesis owing to unforeseen difficulties encountered duringtheir removal or unintended side reactions initiated by their pres-ence9. Ironically, their presence can lead to an additional layer ofchemoselectivity considerations that often take centre stage withina complex total synthesis endeavour8. The multitude of complica-tions imparted by protecting-group manipulations contributes to theperception that natural products, despite their overwhelming utilityin medicine, are too complex to be synthesized efficiently in a drugdiscovery setting10,11.Figure 1 summarizes three different approaches to chemical syn-thesis using the complex natural product ambiguine H (1) as anexample. In a biological setting, where the goal of synthesis is to createfunction rather than a specific target molecule, simple feedstockchemicals are woven together without protecting groups by usingexquisitely selective enzymes12. For instance, it has been proposedthat the key C–C bonds of the ambiguines are forged from an enzym-atic enantioselective cation-olefin cyclization of a simple hydrocar-bon with a 3-substituted indole13. Indeed, emulating nature (bio-mimetic synthesis) can sometimes lead to extremely efficient syn-thetic routes1,2,14–16. In contrast, a standard approach to synthesis usesstrategic disconnections that are often made in order to shield per-ceived functional group incompatibilities en route to a specific target.Here we describe syntheses, the inspiration for which comes partlyfrom studying the biosynthetic pathway, strictly avoid the use ofprotecting groups, and harness the natural reactivity of specific func-tional groups within a complex setting. This approach has led tosolutions that would not have been apparent had the natural tend-encies of the reactive centres been masked.1Department of Chemistry, The Scripps Research Institute, 10550 N. Torrey Pines Road, La Jolla, California 92037, USA.++––NaturalsynthesisStandarddisconnectionsNHNMeMe Me• Function-oriented• No protecting groups• Enzymes needed to promote/control reactivity(PP = pyrophosphate) NMeOMeNH• Target-oriented• No protecting groups• No enzymes• Natural reactivity of functional groups is used constructively• Target-oriented• Protecting groups (PG) needed• Reactivity is ‘caged’ until appropriate timeMeMe XOHMeONRExample:Proposed origin:MeMeOPPThisworkNCNHMeMeMeHHHMeMeAmbiguine H (1)PGPGPGCFigure 1|Approaches to chemical synthesis. Here we show ambiguine H(1) as an example.Vol 446|22 March 2007|doi:10.1038/nature05569404Nature ©2007PublishingGroupThe Stigonemataceae family of cyanobacteria has produced a classof over 60 biogenetically related, architecturally complex, topologic-ally unique, and functionally rich indole natural products that formthe basis of the hapalindole17,18, fischerindole13, welwitindolinone13,and ambiguine19,20alkaloids (Fig. 2). They exhibit a broad range ofbioactivities including antifungal, antibacterial, antimycotic andanticancer properties, with some members having potencies com-parable to clinical agents (streptomycin, puramycin and amphoter-icin)13,17–20. Further study of these potential medicinal agents ishampered by the fact that the cyanobacteria produce complex mix-tures of these natural products in low yield. For instance, smallquantities (about 5 mg) of 1, 2, 4 and 5 have been isolated in yieldsranging from 0.00671% (for 2 ) to 0.0213% (for 5) following tediouspurification and HPLC separation.Total synthesis of hapalindole U (2) and ambiguine H (1)Although there have been no published synthetic routes to the ambi-guines, racemic hapalindole U (2, Fig. 2) has been constructed in 20steps by a non-stereocontrolled sequence with multiple protectinggroups21. Fig. 3 outlines a simple, enantioselective entry to the ambi-guine alkaloid family, by way of 2, facilitated by newly developedmethodology for C–C bond formation and a deliberate effort toeliminate the use of protecting groups.The synthesis commenced with readily available terpene 7, whichis synthesized in four steps by a route that closely parallels the strategyof ref. 22 (see Supplementary Information). The indole and terpenesubunits were then merged without protecting groups using a directindole coupling, a reaction that was
View Full Document
Unlocking...