3.1 Functional analysis of ERG7 Tyr99 within S. cerevisiae
3.1.2 The characterization and identification of novel products
In order to collect products for identification, 58L of mutant yeast were grown and the NSL was extracted, and the products with m/z = 426 were further acetylated for increasing the polarity differences between each other and spotted on the AgNO3-impregenated TLC.(Fig.3.2)
Figure 3.2 Separation of lanosterol fraction on AgNO3-impregenated TLC. Lane 1 is acetylated lanosterol standard; lane 2 is the acetylation crude of lanosterol fraction.
Each single compound was isolated by AgNO3-impregenated silica gel column chromatography by using 3~15% diethyl ether in hexane and subsequently applied for deacetylation as previously described (sections 2.2.7-2.2.8).
Lanosterol-acetate
Compound 1-acetate
Compound 2-acetate
1 2
Non C30H50O triterpenen
GC and GC-MS analysis
Further GC-MS-based product analysis of the lanosterol-positioned product, revealed three triterpenoid products with a molecule mass of m/z = 426: lanosterol and two novel products, compound 1 and 2. Compound 1 migrated on the GC column with a retention time of 0.2 min relative to compound 2. (Fig. 3.3 and Fig. 3.4)
Figure 3.3 GC analysis of the NSL extracts derived from ERG7Y99Thr. Peak 1 indicates compound 1; peak 2 is compound 2; LA means lanosterol. The other peaks which are not marked are not triterpenen products with m/z = 426.
The EI mass spectrum showed that both of novel products had the same molecular ion at m/z = 426 and exhibited similar fragment peaks at 357, 339, and 247, corresponding to the molecular formula C30H50O ([M]+), [M-C5H9]+, [M-C5H9-H2O]+, and [M-C13H21-H2]+, respectively. Therefore, we suggested both of them might be mass the incompleted cyclization products. (Fig. 3.4)
1 LA
2
Figure 3.4 Electron-impact mass spectra oftwo novel products and acetylated forms Compound 1
Compound 2
Compound 1-acetate
Compound 2-acetate Lanosterol
Lanosterol-aceate
The product profiles of each mutant are summarized in Table 3.4. No products with molecular masses of m/z = 426 were observed for inactive mutants, ERG7Y99N, ERG7Y99H and ERG7ΔY99, consistent with the genetic selection results. On the other hand, the viable mutants produced either lanosterol alone (Y99V, Y99L, Y99Q, Y99K, Y99R, Y99C, Y99M, and Y99W) or lanosterol and two novel products (the substitutions of G, A, E, S, T, P).
Additionally, the compound 2 was a major product among those ERG7Y99X mutants that produced three compounds (lanosterol, compound1 and 2).
Products profile ratio (%) amino acids
substitution no products conpound 1 compound 2 lanosterol Gly
Table 3.2 The products profile of S. cerevisiae TKW14C2 expressing the ERG7Y99X site-saturated mutagenesis.
NMR spectral analysis
Firstly, compound 2 was characterized to be tricyclic triterpene preliminarily through 1H NMR and DEPT (Fig. 3.5); these correlations showing in the figure 3.5 suggested the involvement of a tricyclic ring skeleton. Then the structures were subsequently identified following the analysis by nuclear magnetic resonance (1H,
13CNMR, DEPT, 1H-1H COSY, HMQC, HMBC, and NOE) and comparison with those of authentic sample. All of the related NMR spectra were shown in Appendix 3.
Figure 3.5 The major features of compound 2 in 1H NMR and DEPT spectrum. 1H NMR shows distinct chemical shifts with four vinylic methyl signals (δ 1.657, 1.600, 1.584 and 1.577), four methyl singlets (δ 1.072, 0.964, 0.918 and 0.763), and three sets of triplet alkene protons (δ 5.186, 5.076 and 5.043). DEPT revealed the presence of three tertiary-quaternary substituted double bonds (δ = 124.34-131.32, deep yellow arrows;
123.30-134.77, light blue arrows and 126.77-137.80, magenta arrows), which are characteristics of double bonds at the exocyclic hydrocarbon side chains.
Compound 2 was identified by NMR spectroscopy as (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol, a chair-boat (C-B) 6-6-5 tricyclic product with trans-syn-trans stereochemistry and Δ14Z, 17E, 21 double bonds. (Fig.3.4, 3.5, 3.6 and Tab. 3.3)
Figure 3.6 The structure of compound 2.
No. 13C 1H No. 13C 1H No. 13C 1H 3.20, (dd, J=11.59, 5.06Hz) -
Table 3.3 NMR assignments for (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol for dilute CDCl3 solution. Spectra were referenced to tetramethylsilane (TMS) at 0 ppm (1H) or CDCl3 at 77.0 ppm (13C).
(13αH)-Isomalabarica-14Z, 17E, 21-trien-3β-ol
1
Figure 3.7 Bond connectivity and stereochemistry of (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-olestablished by HMBC/HSQC (▃) and NOEs (←→) spectrums.
The presence of NOEs among H-3/Me-29, Me-30/Me-28, Me-29/H-5, H-3/H-5, Me-28/H-9, H-15/Me-26, and Me-27/H-13 while the absence of NOEs among Me-27/Me-28, Me-27/H-9, H-9/H-13, Me-28/H-5, and H-13/H-5, were uniquely consistent with the stereochemistry of the C-B 6-6-5 tricyclic nucleus and the Z conformation for a double bond between C-14 and C-15. (Fig. 3.7)
It is more difficult to purify and isolate the other novel product, compound 1, because of its less quantity in the ERG7Y99X. Fortunately, we found that both of compound 1 and 2 were also produced in the mutant of ERG7F699Met (carried out by Hao-Yu Wen).
Compound 1 exhibited similar parent peak and fragment peak patterns to the previously identified (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol in EI mass. We suggested that it might be an analogous nucleus skeleton of incomplete cyclization. Furthermore, the
1H-NMR spectra also showed four distinct vinylic methyl signals (δ 1.662, 1.605, 1.582 and 1.511), four methyl singlets (δ 1.044, 0.956, 0.912 and 0.756), and three alkene protons (δ 4.997, 5.081 and 5.094), supposing a tricyclic ring skeleton. (Appendix 3)
H O
H H
H H
H
3 5
30 29 28
9
27 26
13 10
15
Via correlating with 13C-NMR, HMQC, HMBC, 1H-1H COSY and NOE spectra, compound 1 was determined to be (13αH)-isomalabarica-14E, 17, 21-trien-3β-ol, a tricyclic product with trans-syn-trans stereochemistry and Δ14E, 17E,21 double bonds. It is structurally similar to compound 2 with differences only in the stereochemistry of the carbon double bond located at the C-14/C-15 position (Fig 3.8 and 3.9, Tab. 3.4).
Figure 3.8 The structure of compound 1.
No. 13C 1H No. 13C 1H No. 13C 1H
Table 3.4 NMR assignments for (13αH)-isomalabarica-14E, 17E, 21-trien-3β-ol for dilute CDCl3 solution. Spectra were referenced to tetramethylsilane at 0 ppm (1H) or CDCl3 at 77.0 ppm (13C).
(13αH)-isomalabarica-14E, 17E, 21-trien-3β-ol
Figure 3.9 Bond connectivity and stereochemistry of (13αH)-isomalabarica-14E, 17, 21-trien-3β-ol established by HMBC/HSQC (▃) and NOEs (←→) spectrums.
The presence of NOEs among Me-29/H-3, Me-29/H-5, Me-30/Me-28, Me-28/H-9, Me-27/H-13, and H-13/H-15, as well as the absence of NOEs between Me-30/H-3, Me-28/Me27, H-13/H-9, H-13/Me-26, and H-15/Me-26, indicate the α–orientation for H-13 and E-conformation for a double bond between C-14 and C-15.
These two compounds exhibited very similar structures with the previously isolated (13αH)-isomalabarica-14(26), 17E, 21-trien-3β-ol, which abstracts the C-26 proton after generating a tricyclic C-14 cation. We suggested that it resulted from a direct trapping of the common C-B 6-6-5 tricyclic Markovnikov C-14 cation but with a different deprotonation position and stereochemistry.[44,47,63]
H O
H H
H H
H
3 5
30 29
28 9
27 13 15
26