Article pubs.acs.org/jnp
Using HMBC and ADEQUATE NMR Data To Define and Differentiate Long-Range Coupling Pathways: Is the Crews Rule Obsolete? Mary M. Senior,† R. Thomas Williamson,‡ and Gary E. Martin*,§ †
Discovery and Preclinical Sciences, Process and Analytical Chemistry, Structural Elucidation Group, Merck Research Laboratories, Kenilworth, New Jersey 07033, United States ‡ Discovery and Preclinical Sciences, Process and Analytical Chemistry, Structural Elucidation Group, Merck Research Laboratories, Rahway, New Jersey 07065, United States § Discovery and Preclinical Sciences, Process and Analytical Chemistry, Structural Elucidation Group, Merck Research Laboratories, Summit, New Jersey 07901, United States S Supporting Information *
ABSTRACT: It is well known that as molecules become progressively more proton-deficient, structure elucidation becomes correspondingly more challenging. When the ratio of 1H to 13C and the sum of other heavy atoms falls below 2, an axiom that has been dubbed the “Crews rule” comes into play. The general premise of the Crews rule is that highly proton-deficient molecules may have structures that are difficult, and in some cases impossible, to elucidate using conventional suites of NMR experiments that include proton and carbon reference spectra, COSY, multiplicity-edited HSQC, and HMBC (both 1H−13C and 1H−15N). However, with access to modern cryogenic probes and microcyroprobes, experiments that have been less commonly utilized in the past and new experiments such as inverted 1JCC 1,n-ADEQUATE are feasible with modest sized samples. In this light, it may well be time to consider revising the Crews rule. The complex, highly proton-deficient alkaloid staurosporine (1) is used as a model proton-deficient compound for this investigation to highlight the combination of inverted 1JCC 1,n-ADEQUATE with 1.7 mm cryoprobe technology.
T
coupling pathways.9 Hence, structure elucidation strategies using a combination of HMBC and 1,1-ADEQUATE spectra effectively solve the problem of differentiating 2JCH from 3JCH correlations.10−12 Proton-deficient molecules frequently involve structures with carbons that require a 4JCH correlation to observe and position them in the skeletal framework being elucidated. Unless one happens to fortuitously observe the necessary correlation(s) in an HMBC spectrum optimized for a small coupling constant, e.g., 3 Hz, assignment strategies combining HMBC and 1,1ADEQUATE spectra are still insufficient. In contrast, although correspondingly less sensitive than 1,1-ADEQUATE in the sense that HMBC is less sensitive than HSQC, 1,nADEQUATE data provide predominantly 3JCC long-range correlations that are analogous to 4JCH HMBC correlations.10,11 The caveat associated with the 1,n-ADEQUATE experiment is that 1JCC correlations unavoidably “leak” into the spectrum and must be differentiated from 3JCC correlations in the same sense that 2JCH and 3JCH correlations must be differentiated in an HMBC spectrum.9,10,13−15 Nevertheless, 1,n-ADEQUATE data do provide a potential means to structurally characterize proton-deficient molecules whose structures may defy elucidation by more conventional means. A new variant of the 1,n-ADEQUATE experiment, dual-optimization inverted 1 JCC 1,n-ADEQUATE, provides the means of specifically
wo-dimensional (2D) experiments are the foundation of most contemporary structure determinations.1 NMR has the capability of defining atom-to-atom connectivities that cannot be established with other spectroscopic techniques. One of the inherent challenges associated with the utilization of typical natural product structure elucidation strategies is the difficulty of differentiating 2JCH from 3JCH correlations in HMBC spectra.2 Routine experimental methods also generally fail to provide what can be crucial 4JCH correlation information. To differentiate 2JCH from 3JCH correlations, a number of 2D NMR experiments have been devised that surmount this challenge with varying degrees of success. The oldest such experiment was the 13C-detected XCORFE experiment reported in 1987 by Reynolds and co-workers.3 The first of the proton-detected experiments focused on this task was the 2 3 J, J-HMBC4 experiment, which was followed by H2BC and variants.5−7 These experiments represent a partial solution to the problem in that they can identify adjacent protonated carbons (via 2JCH), but they cannot identify adjacent quaternary carbon resonances. In effect, the additional information provided by these experiments is no better than a standard COSY that may generally be acquired in minutes on any modern spectrometer. There is also the occasional problem of differentiating an unusually strong 4JCH correlation from 3JCH correlations in an HMBC spectrum, although this is much less common.8 In contrast, although considerably less sensitive, the 1,1-ADEQUATE experiment first reported in 1996 can identify both protonated and nonprotonated adjacent carbons via 1JCC © XXXX American Chemical Society and American Society of Pharmacognosy
Received: July 10, 2013
A
dx.doi.org/10.1021/np400562u | J. Nat. Prod. XXXX, XXX, XXX−XXX
Journal of Natural Products
Article
identifying 1JCC correlations in the spectrum and affords higher sensitivity than a conventional 1,n-ADEQUATE experiment.15 We now wish to report the application of these advanced 2D NMR techniques in conjunction with 1.7 mm MicroCryoProbe technology using the proton-deficient staurosporine molecule as an example.
establish the 1H−13C long-range heteronuclear couplings (Figure S2) and to assign the 15N resonances (Figures S3 and S4, respectively) of staurosporine for the first time. The sequencing and assignment of the proton and carbon resonances for the H1−H4 four-spin system using the IDRHSQC-TOCSY data are shown in Figure 1 (see also Supplemental Figure S5). The overlapped resonances of the other four-spin system were assigned in an identical manner. From the ensemble of correlation information extracted from the series of 2D NMR experiments normally used for natural product structure elucidation, there were several quaternary carbon resonances in the central core of the molecule more than three bonds removed from the nearest proton resonance and hence “silent” in the HMBC spectrum. Resonances of this type can create difficulties in structure elucidation studies and form the basis for the so-called Crews rule.17 The Crews rule, as it has been dubbed,17 is based on an observation18 that when the ratio of the number of protons in a molecule to the sum of the heavy atoms (C, N, O, S, etc.) is
