A NMR method for relative stereochemical assignments of the tricyclic core of cephalosporolides, penisporolides and related synthetic analogues

Abstract

The misassignments of the relative configuration of natural products such as cephalosporolides H and I and penisporolide B were frequently detected by synthetic studies (total synthesis) with considerable effort and cost. Reported herein is the development of a NMR analysis method that can be applied to reliably discriminate the four possible diastereomers of the tricyclic core of cephalosporolides and penisporolides using only proton and/or carbon NMR data without relying on sophisticated 2D NMR spectral analysis. The effectiveness of this NMR method was examined with 28 synthetic compounds, leading to detection and/or correction of 11 misassignments.

---

Natural products have played highly significant and multifaceted roles in chemistry. They not only are the most productive source for the development of new drugs¹ but also greatly inspire chemists to develop new synthetic strategies and methodologies by challenging the state-of-the-art chemical synthesis.² Their molecular structures including the relative and absolute configurations are of paramount importance to these roles, in particular, to total synthesis and drug discovery. However, many natural products were reported with wrong structural assignments if suitable single crystals were unavailable for X-ray diffraction analysis.³ There are few general methods available for structural corrections and confirmation besides NMR analysis⁴ including computational NMR methods and total synthesis. Although unambiguous, total synthesis usually requires tremendous synthetic efforts with considerable cost and time.⁵ For example, recent synthetic studies⁶ including one from our group consistently revealed that the structures for cephalosporolides H and I⁷ and penisporolide B⁸ were incorrect and that the relative configuration of the tricyclic [5,5]-spiroacetal-cis-fused-γ-lactone (SAFL) should be revised (Fig. 1). In light of the potential biological activity of this family of compounds and new members of SAFL-containing natural products probably being isolated in the future, we are interested in developing a reliable NMR analysis method (similar to the Kishi NMR database^4e,f or Rychnovsky acetonide ¹³C-NMR analysis^4g) to determine the relative configuration of the SAFL core without synthetic studies. Prior observation of the spiro-differentiating spin–spin coupling (SSC) pattern first by Ramana⁹ and then by Britton¹⁰ and Brimble^6f further inspired us to undertake a comprehensive analysis of the NMR data of the SAFLs reported in the literature and our laboratory with the expectation of stereochemical discrimination of all four possible diastereomers using only the NMR analysis method. Note: since the γ-lactone of the SAFLs is always cis-fused to the spiroacetal with the same configuration (3R,4R or 3S,4S), there are only four diastereomers derived from the four chiral centers (C3, C4, C6 and C9), namely, SAFL-A with the (3S*,4S*,6S*,9R*) relative configuration, (3S*,4S*,6R*,9R*) SAFL-B, (3R*,4R*,6S*,9R*) SAFL-C, and (3R*,4R*,6R*,9R*) SAFL-D (Fig. 1).

Fig. 1 Structural revision of the tricyclic core of cephalosporolides H and I and penisporolides B and their four possible diastereomers.

Our studies started with a comprehensive analysis of the ¹H-NMR spectra derived from four synthetic diastereomers SAFL-A1, -B1, -C1, and -D1, whose relative configurations have been unambiguously determined by the combination of NMR spectroscopic (including 2D NMR) and X-ray diffraction analysis.^6e After the careful comparative analysis of the ¹H-NMR spectra¹¹ (Fig. 2), the spin–spin coupling pattern and chemical shifts of H₄, H₅ and H₈ were identified to be useful for distinguishing these four diastereomers. The first obvious difference was the SSC pattern of H4: triplet (t) at 5.08–5.10 ppm versus doublet of doublet of doublet (ddd) at 5.02 ppm, which is consistent with the observation in the related spiroketal compounds reported by Ramana,⁹ Britton¹⁰ and Brimble.^6f This distinct H4 SSC with slightly different chemical shifts differentiated SAFL-A1/D1 from SAFL-B1/C1, corresponding to the stereochemical discrimination of the spiroketal chiral centers (3S*,4S*,6S* vs. 3S*,4S*,6R*), irrespective of the C9 relative configuration. The different SSC patterns of these two pair diastereomers are explained by their slightly different conformations resulting from opposite stereochemistry at the spiroketal carbon, as illustrated by X-ray diffraction analysis of SAFL-B1 and -D1 [Fig. 3(a) and (b)].

Fig. 2 Truncated ¹H-NMR spectra of SAFL-A1-D1. Note: full NMR spectra are available in the ESI of ref. 6e.

Fig. 3 (a) Thermal ellipsoid diagram of SAFL-B1 and Newman projection along C4–C5. (b) Thermal ellipsoid diagram of SAFL-D1 and Newman projection along C4–C5. (c) Chemical difference of H₅ in the SAFLs. (d) Chemical difference of H₈ in the SAFLs.

The U-shaped conformation of SAFL-D1 [Fig. 3(b)] has a ∼90° dihedral angle of H₄ and H_5a (H_5a is trans-oriented to H₄, 86.16° is the torsion angle of O2–C4–C5–C6 in the X-ray data^6e), resulting in no spin–spin coupling between H₄ and H_5a according to the Bothner-By¹² approximation of the Karplus correlation¹³ between ³J-coupling constants and dihedral angles and coincidentally identical ³J coupling constants of H₄–H₃ and H₄–H_5b. The V-shaped SAFL-B1 [Fig. 3(a)] reveals an unusual eclipsed conformation along C4–C5 with the <5° dihedral angle of H₄–C₄–C₅–H_5a (104.3° is the torsion angle of O2–C4–C5–C6 in the X-ray data^6e), which according to the Bothner-By equation explains the ddd SSC of H₄ with three different neighboring protons (H_5a, H_5b and H₃). This stereochemical discrimination of the spiroketal chiral center through the H4 SSC pattern could be further consolidated by the significant difference of the two H₅ chemical shifts: Δδ₅ (δ_5a–δ_5b) 0.32–0.33 ppm for SAFL-A1/D1 and 0.14–0.15 ppm for SAFL-B1/C1 (Fig. 2). This chemical shift difference may also arise from different conformations revealed by the X-ray diffraction analysis. The fully eclipsed conformation along C4–C5 (B ring) in SAFL-B1 (and probably SAFL-C1) might cause orbital distortions¹⁴ of H₅ protons or steric compression¹⁵ and therefore affect the chemical shifts [H_5a (dd) is shifted downfield to 2.5 ppm], while the conformation along C4–C5 in SAFL-D1 (probably and SAFL-A1) is between the staggered and eclipsed conformations, which explains the H_5b (dd) proton downfield at ∼2.1 ppm. An alternative rationalization of this difference might be that the diamagnetic anisotropic chemical shift effect of a CH₂ is consistently larger when flanking substituents are cis to each other¹⁶ [Fig. 3(c)]: cis-4,6-dioxy substitution giving rise to the large Δδ₅ (δ_5a–δ_5b = 0.32–0.33 ppm, SAFL-A1/D1); the trans-4,6-dioxy substitution resulting in the small Δδ₅ (δ_5b–δ_5a = 0.14–0.15 ppm, SAFL-B1/C1).¹⁷ A combination of the H₄ SSC pattern and H₅ chemical shift difference could provide a reliable way to determine the relative configuration of C6 and C3/4 and distinguish SAFL-A1/D1 from SAFL-B1/C1. To differentiate SAFL-A1 from SAFL-D1 (and SAFL-B1 from SAFL-C1) for determination of the relative configuration of C9 and C6, the chemical shifts of H₈ were found to be the only useful resonances with a considerable difference. Δδ₈ (δ_8a–δ_8b) 0.65 ppm was observed in SAFL-A1 and only Δδ₈ 0.33 ppm in SAFL-D1. Similar Δδ₈ difference was found between SAFL-B1 (Δδ₈ 0.24 ppm) and SAFL-C1 (Δδ₈ 0.63 ppm). However, the rationalization of this chemical shift difference (Δδ₈) is much more difficult than the case of Δδ₅ because the different C6 substitution is three bonds away from the H₈ nuclei. After reviewing the X-ray ORTEP diagrams of SAFL-B1/D1 using the commercial software Mercury, we suspect that the effect of steric compression¹⁸ on SAFL-B1/D1 [Fig. 3(d)] might shift downfield of the pseudo-axial H_8b, leading to the smaller Δδ₈ value (note: the axial protons are upfield of the equatorial ones in most substituted cyclohexanes with the Δδ (δ_e–δ_a) value of ∼0.5 ppm. At this stage, the four SAFL diastereomers could be discriminated by the proton NMR analysis and the relative configuration of an unknown tricyclic SAFL may be assigned correctly and readily.

To minimize the errors arising from different machines and sample preparation and generalize this NMR method for different SAFLs, the value of Δδ₅ and Δδ₈ is further averaged among six different SAFLs with the same relative configuration but different C9 substituents, which had been prepared in our laboratory (Table 1). In general, the results are highly consistent with those observed in the SAFLs with the C9 methyl group. For example, the average of Δδ₅ and Δδ₈ in the six SAFL-A series is 0.33 ppm and 0.62 ppm with standard deviation (SD) of 0.009 and 0.018, respectively, which suggests the homogeneity of the data and the least influence of the C9 substitution on Δδ₅ and Δδ₈. Therefore, the average value of Δδ₅ and Δδ₈ is suggested to be used in the subsequent stereochemical assignments of each type of SAFL as follows. SAFL-A: Δδ₅ (ppm) 0.33, Δδ₈ (ppm) 0.62; SAFL-B: Δδ₅ (ppm) 0.14, Δδ₈ (ppm) 0.22; SAFL-C: Δδ₅ (ppm) 0.15, Δδ₈ (ppm) 0.60; SAFL-D: Δδ₅ (ppm) 0.32, Δδ₈ (ppm) 0.32.

Table 1 Values of Δδ₅ and Δδ₈ from 24 SAFLs

H#	SAFL-A (1–6)	SAFL-B (1–6)	SAFL-C (1–6)	SAFL-D (1–6)
Note: Avg = average; SD = standard deviation.
H5 (Δδ5)	0.33, 0.34	0.14, 0.15	0.15, 0.16	0.32, 0.32
	0.34, 0.33	0.15, 0.14	0.15, 0.15	0.33, 0.32
	0.32, 0.32	0.15, 0.13	0.15, 0.13	0.32, 0.31
Avg	0.33	0.14	0.15	0.32
SD	0.009	0.008	0.010	0.006
H8 (Δδ8)	0.65, 0.63	0.24, 0.22	0.63, 0.62	0.33, 0.34
	0.60, 0.63	0.21, 0.21	0.59, 0.58	0.32, 0.31
	0.61, 0.62	0.22, 0.18	0.57, 0.58	0.31, 0.30
Avg	0.62	0.22	0.60	0.32
SD	0.018	0.020	0.024	0.015

---

One challenge in this ¹H-NMR analysis is to correctly assign the resonances to H₅ and H₈, which are usually in the upfield region (1.0–2.5 ppm) and frequently overlapping with other alkyl proton signals. This problem could be solved by an extensive 2D NMR analysis. For example, ¹H–¹H COSY can correlate the adjacent three-bond protons (e.g., H₃–H₄–H₅, H₇–H₈–H₉) and HMQC (or HSQC) can be used to identify the methylene resonances at C5 and C8. However, the chemical shifts of H5 and H8 might be erroneously reported or swapped with other signals without the authentic 2D NMR spectra, which would prevent the use of this proton NMR analysis method as an effective tool to stereochemically discriminate the SAFLs. In order to address this potential problem, we performed an extensive analysis of the ¹³C-NMR data derived from the 24 SAFLs that were used in the proton NMR analysis in Table 1. First of all, the carbon resonances that shift most over different types of SAFLs but least within the same type of SAFL should be identified. The value of standard deviation (SD) for each carbon resonance was compared within that of the same type and among the different types of SAFLs for identification of such stereochemistry-responsive carbon(s). The average chemical shifts and the SD values for each carbon within the same type of SAFL are tabulated in Table 2.¹⁹ Apparently, the chemical shifts of C3, C7 and C9 vary significantly over different types of SAFLs with the large SD values of 0.996, 1.118 and 0.760, respectively. This suggests that the chemical shifts derived from these carbons may be useful to distinguish the four types (relative configuration) of SAFLs. Secondly, the small SD values within the same type of SAFL were observed in C1 (0.06–0.25), C2 (0.13–0.21), C3 (0.04–0.09), C5 (0.06–0.18), C6 (0.06–0.12), and C7 (0.13–0.22), which suggests that the substituent at C9 has less influence on the chemical shift of these carbons. Taking all these observations together, we concluded the resonances of C3 and C7 were characteristic for each type of SAFL and could be used for stereochemical assignment of an unknown SAFL.²⁰ In order to quantify the assessment of each carbon chemical shift for the stereochemical discrimination in this analysis process, we proposed and defined the σ value for each carbon (Table 2). The σ value of each carbon can be obtained by the division of the SD value of different types of SAFLs by the average of all SD values that are obtained within the same type of SAFL. For example, σ(C3) = 0.996/[(0.04 + 0.09 + 0.04 + 0.04)/4] = 0.996/0.0525 = 19.0. The carbon with larger σ value is better for the stereochemical discrimination. As shown in Table 2, the σ value of C3 (19.0) and C7 (6.30) is larger than that of others, which indicates that C3 and C7 are better representatives for the stereochemical discrimination of SAFLs. Therefore, the relative configuration of SAFL-A could be confirmed by δ_C3 (ppm) 86.7 and δ_C7 (ppm) 34.4. Analogously, SAFL-B could be established by δ_C3 (ppm) 85.1 and δ_C7 (ppm) 37.0; SAFL-C by δ_C3 (ppm) 85.2 and δ_C7 (ppm) 35.6; SAFL-D by δ_C3 (ppm) 87.1 and δ_C7 (ppm) 36.3. It is recognized that a single carbon, C3 or C7, could not unambiguously distinguish the four diastereomers and the combination of these two carbon chemical shifts are essential. However, if the spiroketal chiral center is involved, the difference of either C3 or C7 is significant.

Table 2 The average of chemical shifts for each carbon derived from the same type of SAFL and the SD value of each carbon from different types of SAFLs

C#	SAFL-A (SD)	SAFL-B (SD)	SAFL-C (SD)	SAFL-D (SD)	SD	σ
1	180.93 (0.25)	180.94 (0.08)	180.87 (0.06)	180.77 (0.19)	0.078	0.54
2	44.38 (0.21)	44.50 (0.16)	44.50 (0.15)	44.61 (0.13)	0.094	0.58
3	86.67 (0.04)	85.09 (0.09)	85.24 (0.04)	87.06 (0.04)	0.996	19.0
4	79.97 (0.14)	80.43 (0.20)	80.49 (0.07)	80.92 (0.83)	0.389	1.25
5	41.77 (0.06)	41.83 (0.07)	42.29 (0.18)	41.62 (0.10)	0.289	2.82
6	115.03 (0.12)	115.35 (0.11)	115.55 (0.06)	115.17 (0.08)	0.225	2.43
7	34.40 (0.16)	37.04 (0.22)	35.62 (0.13)	36.27 (0.20)	1.118	6.30
8	30.22 (0.59)	30.90 (0.74)	30.29 (0.80)	31.12 (0.62)	0.446	0.65
9	78.39(1.55)	78.03 (4.94)	77.73 (1.48)	79.47 (0.87)	0.760	0.34

---

At this stage, the NMR analysis method for stereochemical assignments of the four diastereomeric SAFLs was established and therefore the decision tree²¹ is summarized in Fig. 4. Although the ¹³C-NMR method is independent of the ¹H-NMR analysis, the combination of ¹H- and ¹³C-NMR analysis usually leads to a quick and conclusive determination of the relative configuration of an unknown SAFL. Practically, the following steps are suggested for an unknown SAFL: (i) SSC of H4 (downfield, t vs. ddd); (ii) δ_C3 (85–87 ppm); (iii) δ_C7 (34–37 ppm); (iv) Δδ₅ (0.14 vs. 0.33 ppm); (v) Δδ₈ (0.22, 0.32, or ∼0.60 ppm). This order of steps does not correspond to the order of priority or importance, but it facilitates a quick determination of the relative configuration of an unknown SAFL because the information of the first three steps can be easily extracted from the spectra without ambiguity.

Fig. 4 Decision tree used for stereochemical assignments of SAFLs by NMR.

To showcase the application of this NMR method, the reported NMR data of penisporolide A⁸ was analyzed (Fig. 5). The H₄ SSC of penisporolide A was reported as a multiplet at 5.01 ppm. It was more likely to be ddd since in some cases the ddd pattern was not well resolved for clear interpretation. The Δδ₅ (0.17 ppm) and the chemical shifts of C3 (85.2 ppm) and C7 (35.5 ppm) were well consistent with that of SAFL-C, although the Δδ₈ (0.05 ppm) value was not matching at all. After re-examination of original assignments, it was thought that the resonance (1.55 ppm) assigned to H₁₂ might be derived from H_8b, which would result in the matching Δδ₈ (0.60 ppm) value as SAFL-C. If the structural revision was true, this NMR analysis could serve as an effective tool to detect possible misassignments of NMR resonances.

Fig. 5 Possible stereochemical revision of the SAFL core of penisporolide A by the NMR analysis method.

To further demonstrate the predictive ability of this NMR analysis method, the assignments of the relative configuration of 19 SAFL-containing compounds reported in the literature were re-examined and the results are presented in Table 3. Most NMR resonances of these compounds had been correctly assigned, which supports the effectiveness of our NMR method. However, several wrong assignments of carbon and proton resonances (carbon 3: B8, B9, C8, and D8; carbon 7: D12; and proton 8: A8, A9, C8 and D8) could be detected and corrected by our NMR analysis method, in support of the predictive ability of our method irrespective of the substitution at C9. The detection and predictive ability of this NMR method would be of high importance to further/future research on this family of natural/non-natural products.

Table 3 Examination of 19 SALF-containing compounds reported in the literature

	A7	A8	A9	A12	B7	B8	B9	B12	C7	C8	C10	C11	C12	C13	C14	D7	D8	D11	D12	D13	D14
Complete NMR data are available in the ESI of ref. 6a–f.a Wrong assignment of C3 or C7.b Wrong assignment of H8. ✗ Δδ8 (0.2–0.4 ppm) does not match.
SSC	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓
δ C3	✓	✓	✓	✓	✓	✓^a	✓^a	✓	✓	✓^a	✓	✓	✓	✓	✓	✓	✓^a	✓	✓	✓	✓
δ C7	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓^a	✓	✓
Δδ5	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓
Δδ8	✓	✗	✗	✓	✓	✓	✓	✓	✓	✗	✓	✓	✓	✓	✓	✓	✓^b	✓	✓	✓	✓
Ref.	6e	6f	6f	6a	6e	6f	6f	6a	6e	6f	6f	6e	6d,e	6c,d	6d	6e	6f	6d	6a–d	6c,d	6d

---

It is recognized that some natural products contain a similar tricyclic core structure with different substitutions at C2. It would be interesting to attest this NMR method with these types of SAFL-containing compounds (Table 4). Remarkably, our ¹H-NMR analysis method was applicable to these SAFL natural products such as cephalosporolides E (1) and F (2),²²ent-ascospiroketal B (7),²³ and some synthetic compounds (3–6 and 8),^6c,9 although the ¹³C-NMR analysis fails in most cases due to the considerable substituent effects on C3 and/or C7.²⁴ In addition, the wrong assignments of the relative configuration of spiroketal centers of 3 and 4 were easily detected. It is expected that similar NMR analysis methodology will be developed for stereochemical examinations/assignments of other types of compounds.

Table 4 NMR analysis of other C2 substituted SAFL compounds

	1	2	3	4	5	6	7	8
a Wrong assignment of C6 stereochemistry.
SSC	✓	✓	✓	✓	✓	✓	✓	✓
δ C3	✗	✗	✗	✗	✗	✗	✗	✓
δ C7	✓	✗	✓	✗	✗	✓	✓	✓
Δδ5	✓	✓	✓	✓	✓	✓	✓	✓
Δδ8	✓	✓	✓	✓	✓	✓	✓	✓
Ref.	9 and 22a	9 and 22a	6c	6c	9	9	22	22

---

Conclusions

In summary, a reliable NMR method was established to discriminate the relative configuration of the four possible diastereomers of SAFLs, a tricyclic core structure found in a small family of natural products. The effectiveness of this NMR analysis method was demonstrated by 28 SAFL compounds reported in the literature. This NMR method can also serve as a detection tool for stereochemistry and resonance misassignments as exemplified by the possible stereochemical revision of penisporolide A and compounds 3 and 4 and by the wrong assignments of C3 and/or C7 of 8 synthetic SAFLs (SAFL-B8, -B9, -C8, -D8, and -D12). It is expected that this NMR analysis will provide guidance for stereochemical assignments of new SAFL compounds isolated or synthesized in the future and inspire the development of similar NMR methods for determination of the relative configuration of other types of compounds.

Acknowledgements

This research was financially supported by HKUST and the Research Grant Council of Hong Kong (ECS 605912, GRF 605113, and GRF 16305314) and partially supported by the NSFC (Project No. 21472160). We are very grateful to Xin Miao (HKUST) for some preliminary NMR analysis.

Notes and references

1. For selected reviews, see: (a) A. L. Harvey, Drug Discovery Today, 2008, 13, 894–901; (b) E. Patridge, P. Gareiss, M. S. Kinch and D. Hoyer, Drug Discovery Today, 2016, 21, 204–207; (c) D. J. Newman and G. M. Cragg, J. Nat. Prod., 2012, 75, 311–335; (d) G. M. Cragg, P. G. Grothaus and D. J. Newman, Chem. Rev., 2009, 109, 3012–3043.
2. For selected reviews on representative examples, see: (a) M. Büschleb, S. Dorich, S. Hanessian, D. Tao, K. B. Schenthal and L. E. Overman, Angew. Chem., Int. Ed., 2016, 55, 4156–4186; (b) K. C. Nicolaou, D. Vourloumis, N. Winssinger and P. S. Baran, Angew. Chem., Int. Ed., 2000, 39, 44–122; (c) A. M. Armaly, Y. C. DePorre, E. J. Groso, P. S. Riehl and C. S. Schindler, Chem. Rev., 2015, 115, 9232–9276; (d) C. Cordier, D. Morton, S. Murrison, A. Nelson and C. O'Leary-Steele, Nat. Prod. Rep., 2008, 25, 719–737; (e) J. T. Mohr, M. R. Krout and B. M. Stoltz, Nature, 2008, 455, 323–332.
3. (a) K. C. Nicolaou and S. A. Snyder, Angew. Chem., Int. Ed., 2005, 44, 1012–1044; (b) H.-D. Yoo, S.-J. Nam, Y.-W. Chin and M.-S. Kim, Arch. Pharmacal Res., 2016, 39, 143–153.
4. NMR analysis is the most studied method. For excellent reviews, see: (a) J. M. Seco, E. Quiñoá and R. Riguera, Chem. Rev., 2004, 104, 17–118; (b) G. Bifulco, P. Dambruoso, L. Gomez-Paloma and R. Riccio, Chem. Rev., 2007, 107, 3744–3779; (c) M. Elyashberg, A. J. Williams and K. Blinov, Nat. Prod. Rep., 2010, 27, 1296–1328; (d) M. W. Lodewyk, M. R. Siebert and D. J. Tantillo, Chem. Rev., 2011, 112, 1839–1862. For selected examples, see: (e) J. Lee, Y. Kobayashi, K. Tezuka and Y. Kishi, Org. Lett., 1999, 1, 2181–2184; (f) S. Higashibayashi, W. Czechtizky, Y. Kobayashi and Y. Kishi, J. Am. Chem. Soc., 2003, 125, 14379–14393; (g) S. D. Rychnovsky, B. Rogers and G. Yang, J. Org. Chem., 1993, 58, 3511–3515; (h) S. D. Rychnovsky, Org. Lett., 2006, 8, 2895–2898; (i) W. R. Roush, T. D. Bannister, M. D. Wendt, M. S. VanNieuwenhze, D. J. Gustin, G. J. Dilley, G. C. Lane, K. A. Scheidt and W. J. Smith, J. Org. Chem., 2002, 67, 4284–4289; (j) S. G. Smith and J. M. Goodman, J. Org. Chem., 2009, 74, 4597–4607; (k) F. López-Vallejo, M. Fragoso-Serrano, G. A. Suárez-Ortiz, A. C. Hernández-Rojas, C. M. Cerda-García-Rojas and R. Pereda-Miranda, J. Org. Chem., 2011, 76, 6057–6066; (l) P. H. Willoughby, M. J. Jansma and T. R. Hoye, Nat. Protocols, 2014, 9, 643–660 , and references therein.
5. For reviews covering incorrectly assigned structures of natural products and their structural revisions by total synthesis, see: (a) M. E. Maier, Nat. Prod. Rep., 2009, 26, 1105–1124; (b) T. L. Suyama, W. H. Gerwick and K. L. McPhail, Bioorg. Med. Chem., 2011, 19, 6675–6701. For recent examples of structural revisions by total synthesis from our group, see: (c) L. Song, K.-H. Lee, Z. Lin and R. Tong, J. Org. Chem., 2014, 79, 1493–1497; (d) L. Zhu, Y. Liu, R. Ma and R. Tong, Angew. Chem., Int. Ed., 2015, 54, 627–632; (e) L. Zhu and R. Tong, Synlett, 2015, 1643–1648; (f) L. Zhu and R. Tong, Org. Lett., 2015, 17, 1966–1969.
6. (a) S. F. Tlais and G. B. Dudley, Org. Lett., 2010, 12, 4698–4701; (b) S. F. Tlais and G. B. Dudley, Beilstein J. Org. Chem., 2012, 8, 1287–1292; (c) R. A. Fernandes and M. B. Halle, Asian J. Org. Chem., 2013, 2, 593–599; (d) J. Li, C. Zhao, J. Liu and Y. Du, Tetrahedron, 2015, 71, 3885–3889; (e) J. Wang and R. Tong, J. Org. Chem., 2016, 81, 4325–4339. Brimble achieved the synthesis of the tricyclic SAFL core of CesH, see: (f) O. C. Finch, D. P. Furkert and M. A. Brimble, Tetrahedron, 2014, 70, 590–596.
7. X. Li, Y. H. Yao, Y. A. Zheng, I. Sattler and W. H. Lin, Arch. Pharmacal Res., 2007, 30, 812–815.
8. X. Li, I. Sattler and W. H. Lin, J. Antibiot., 2007, 60, 191–195.
9. C. V. Ramana, S. B. Suryawanshi and R. G. Gonnade, J. Org. Chem., 2009, 74, 2842–2845.
10. S. Chang and R. Britton, Org. Lett., 2012, 14, 5844–5847.
11. All NMR data used in this article were collected in CDCl₃.
12. A. A. Bothner-By, in Advances in Magnetic Resonance, ed. J. S. Waugh, Academic Press, New York, 1965, vol. 1, pp. 195–613.
13. M. Karplus, J. Am. Chem. Soc., 1963, 85, 2870–2871.
14. For related cases with geometry and orbital distortions that cause chemical shifts, see: (a) F. M. Beringer, L. L. Chang, A. N. Fenster and R. R. Rossi, Tetrahedron, 1969, 25, 4339–4345; (b) D. J. Sardella and E. Boger, Magn. Reson. Chem., 1989, 27, 13–20.
15. M. A. Cooper and S. L. Manatt, J. Am. Chem. Soc., 1970, 92, 4646–4652.
16. For similar effects in tetrahydrofurans, see: (a) D. R. Williams, Y. Harigaya, J. L. Moore and A. D'sa, J. Am. Chem. Soc., 1984, 106, 2641–2644; (b) E. D. Mihelich and G. A. Hite, J. Am. Chem. Soc., 1992, 114, 7318–7319. For similar effects in γ-lactones, see: (c) S. A. M. T. Hussain, W. D. Ollis, C. Smith and J. F. Stoddart, J. Chem. Soc., Perkin Trans. 1, 1975, 1480–1492.
17. The upfield shift of cis substituents compared to trans is often observed in a series of succinic anhydrides, see: L. E. Erickson, J. Am. Chem. Soc., 1965, 87, 1867–1875.
18. For the effect of steric compression on chemical shifts in cage-like molecules, see: S. Winstein, P. Carter, F. A. L. Anet and A. J. R. Bourn, J. Am. Chem. Soc., 1965, 87, 5247–5249 . Also see ref. 15.
19. See the ESI.†.
20. The characteristic deviation of the chemical shifts of the C7 in the ¹³C-NMR might be due to the shielding γ/δ-effect caused by hydrogen–hydrogen gauche interactions, especially in these SAFLs where C7 is oriented inside the concave face of the cis-fused-γ-lactone core, see: D. M. Grant and B. V. Cheney, J. Am. Chem. Soc., 1967, 89, 5315–5318.
21. For the use of decision trees in NMR, see: F. Qiu, A. Imai, J. B. McAlpine, D. C. Lankin, I. Burton, T. Karakach, N. R. Farnsworth, S.-N. Chen and G. F. Pauli, J. Nat. Prod., 2012, 75, 432–443.
22. For isolation of cephalosporolides E and F: (a) M. J. Ackland, J. R. Hanson, P. B. Hitchcock and A. H. Ratcliffe, J. Chem. Soc., Perkin Trans. 1, 1985, 843–847; (b) A. Farooq, J. Gordon, J. R. Hanson and J. A. Takahashi, Phytochemistry, 1995, 38, 557–558; (c) V. Rukachaisirikul, S. Pramjit, C. Pakawatchai, M. Isaka and S. Supothina, J. Nat. Prod., 2004, 67, 1953–1955; (d) J. L. Oller-López, M. Iranzo, S. Mormeneo, E. Oliver, J. M. Cuerva and J. E. Oltra, Org. Biomol. Chem., 2005, 3, 1172–1173; (e) H. M. T. B. Herath, M. Jacob, A. D. Wilson, H. K. Abbas and N. P. D. Nanayakkara, Nat. Prod. Res., 2013, 27, 1562–1568.
23. J. Wang and R. Tong, Org. Lett., 2016, 18, 1936–1939.
24. For substituent effects on carbon NMR chemical shifts, see: (a) C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165–195. It was reported that carbon NMR chemical shifts data are less discriminating than those of proton ones, see: (b) M. G. Chini, R. Riccio and G. Bifulco, Eur. J. Org. Chem., 2015, 1320–1324.

---

Footnote

† Electronic supplementary information (ESI) available: Ten tables showing NMR data of 24 compounds. See DOI: 10.1039/c6qo00556j

---