achilleol A camelliol C Lanosterol Parkeol
3.2.7 Site-saturated mutagenesis approach to investigate the functional importance of the critical Tyr-510 residuesimportance of the critical Tyr-510 residues
These saturated mutations of Tyr-510 in oxidosqualene-lanosterol cyclase were respectively created by using QuikChange Site-Directed Mutagenesis kit and confirmed by DNA sequencing using the ABI PRISM 3100 DNA sequencer.117-119 These recombinant plasmids were then electroporated into the yeast TKW14 strain for analyzing the functional activity, as previously described.117-119 The plasmids shuffle were also carried out with CBY57[pZS11] strain. From the functional complementary assay among these various S. cerevisiae ERG7Y510X mutants, the Tyr510Lys, Tyr510Arg, Tyr510Thr, Tyr510Pro and Tyr510Trp substitutions failed to complement the erg7 deficiency in the absence of exogenous ergosterol. This result suggested that the different substitution of Tyr-510 might influence the enzymatic activity and abolish the cyclization process of oxidosqualene.
The nonsaponifable lipid extracts from each mutant were subsequently characterized. The product profiles of S. cerevisiae ERG7Y510X mutants are listed in Table 3.3. Accordingly, the ERG7Y510X mutants produced diverse product profile ranging from monotonous to polycyclic compounds with molecular weight of m/z = 426. Based on the GC-MS spectrum as well as the NMR data, all of the compounds including achilleol A, camelliolC, (13αH)-isomalabarica-14(26),17,21-trien-3β-ol, lanosterol, and parkeol were respectively identified.115-117, 125
Table 3.3 The product profiles of S. cerevisiae TKW14 expressing the ERG7Y510X site-saturated mutants and two ERG7H234WY510X mutants.
amino acid
(13αH)-isomalabarica-substitution achilleol A camelliol C 14(26),17,21-trien-3β-ol lanosterol parkeol
Gly 15
▬17 27 41
Ala 27
▬ ▬39 34
Val 4
▬17 38 41
Leu
▬ ▬26 74
▬Ile 4
▬11 78 7
Asp 9
▬11 40 40
Asn 5
▬9 9 77
Glu 6
▬27 18 49
Gln 5
▬26 14 55
His 45
▬24 5 26
Lys 87 13
▬ ▬ ▬Arg
▬ ▬ ▬ ▬ ▬Ser
▬ ▬4 40 56
Thr
▬ ▬ ▬ ▬ ▬Cys
▬ ▬67 33
▬Met 1
▬15 43 41
Phe 4
▬10 50 36
Trp 94 6
▬ ▬ ▬Pro
▬ ▬ ▬ ▬ ▬Tyr
▬ ▬ ▬ 100 ▬product profiles of ERG7H234XY510X double mutants
H234W/Y510V
2
▬8 90
▬H234W/Y510W
99 1
▬ ▬ ▬H234Y/Y510A117 ▬ ▬ ▬ 100 ▬
Y510W
94 6
▬ ▬ ▬H234W117 ▬ ▬ ▬ ▬ 100
For the products analysis of Tyr-510 inactive mutants, replacement of tyrosine with arginine, proline and threoine failed to generate any product with molecular mass of m/z = 426, consistent with the genetic selection results. No product production might be caused by the incorrect protein folding of S. cerevisiae ERG7 from the individual amino acid substitution on the Tyr-510 position. The strongly basic arginine and the conformational restricted proline might directly disrupt the protein structure.
Interestingly, in contrast to the multiple products profile from other polar residue substitutions, the exact reason for no product extraction from Tyr510Thr mutant remained unclear. Moreover, another two nonviable mutants, ERG7Y510K and ERG7Y510W generated trace amount of monocyclic product, either achilleol A or camelliol C, as the previous section mentioned.125
For the viable mutants, similar product profiles with different ratio were observed. In most of the viable mutants, lanosterol or its alternatively deprotonated product, parkeol, distributed predominantly over the entire of product profiles (near or over 70%), except for the ERG7Y510H or ERG7Y510C mutants. This finding suggested that the cyclization/rearrangement cascade of oxidosqualene in the TKW14[pERG7Y510X] strains is almost unaffectedly complete until the final deprotonation step. Therefore, the position of Tyr-510 or its corresponding residue might be critical for the final deprotonation reaction as well as for the product specificity.106, 108 The nonpolar, aliphatic amino acid substitution, especially in the ERG7Y510L or ERG7Y510I mutants, might provide the selective pressure for the precisely deprotonated control to produce the abundant lanosterol. In contrast, the dominant production of parkeol in ERG7Y510S, ERG7Y510N, ERG7Y510E, or ERG7Y510Q mutants might be caused by their polar or negative charge characteristic for the alternative deprotonation reaction. Interestingly, the longer side chain substitution in either ERG7Y510E or ERG7Y510Q mutants further influenced the cyclization progression
and resulted in the generation of considerable quantities of premature (13αH)-isomalabarica-14(26),17,21-trien-3β-ol, which was previously observed from the S. cerevisiae ERG7Y510F or ERG7Y510H mutants by Mutsuda and his coworkers and also independently isolated from our previous S. cerevisiae ERG7H234X mutants.117,
126 The tricyclic byproduct provided direct mechanistic evidence for the formation of crucial chair-boat 6-6-5 tricyclic Markovnikov cationic intermediate in the ERG7-catalyzed oxidosqualene cyclization reaction. Interestingly, the production of (13αH)-isomalabarica-14(26),17,21-trien-3β-ol was observed from most of the S.
cerevisiae ERG7
Y510X mutants. The different product ratio compared with that of previously established ERG7Y510F or ERG7Y510H mutants, might be attributed to the different yeast analysis system used, or the physiological metabolism of lanosterol.The accurate quantitation of product ratio from S. cerevisiae ERG7Y510X mutations without any background interference should be carefully carried out by using the in
vitro assay system or by using the over-expression system with the purified cyclase.
105 However, the result described herein could provide qualitative analysis for the functional role of Tyr-510 position in the S. cerevisiae ERG7. Moreover, the metabolized lanosterol in TKW14[pERG7Y510X] strain could facilitate the chemical characterization of the trace of novel triterpene alcohol that might be obscure due to the abundant lanosterol in the yeast host strain. Distinctively, the relatively dominant proportion of (13αH)-isomalabarica-14(26),17,21-trien-3β-ol in the product profiles of ERG7Y510C, might come from the powerful nucleophilic property of cysteine that disrupted the transient dipole interaction between carbocationic intermediate and the enzymatic active site, which retarded the tetracyclic ring formation and resulted in the early termination. Without other tricyclic alternatively deprotonated products might be due to steric position of the β facing C-14 methyl group that is close to H-9β protonAdditionally, the monocyclic achilleol A and camelliol C which is different in the alternative proton abstraction were also indisputably observed from most of the ERG7Y510X mutants. The higher accumulation of achilleol A over that of camelliol C or the production of achilleol A whenever camelliol C is produced might be due to the kinetically favored double-quaternary double bond deprotonation or the stereochemical control from the enzymatic active site. The bulky indole ring of tryptophan, the imidazole group of histidine, and the positive charged lysyl group of lysine might have dramatically influenced the progression of cyclization and resulted in the production of large amount of monocyclic products. The cyclization process might be retarded either from the enzymatic perturbation or the substrate misfolding.
Interestingly, the ERG7Y510H mutation which produces abundant monocyclic products but also with lanosterol for supporting the yeast survivability might be caused by the relatively small electrostatic change in the active site cavity or the slighter influence on the cyclization process, compared to that of the lysine or the tryptophan substitutions.
In consideration of the generated product profiles for various amino acid substitutions of the Tyr-510 position, the functionally important role of this residue on the oxidosqualene cyclization/rearrangement cascade could be elucidated. Scheme 3.4 shows the proposed oxidosqualene cyclization/rearrangement cascade occurred in the
S. cerevisiae ERG7
Y510X site-saturated mutants. The epoxide ring was first opened via a proper catalytic acid activation and the first monocyclic C-10 cation was immediately generated. The progression of cyclization was partially blocked due to the steric influence from the substitution of an original tyrosine residue for the bulky group and resulted in the production of achilleol A and camelliol C.Scheme 3.4 Proposed cyclization/rearrangement pathway of oxidosqualene in TKW14 expressing ERG7Y510X site-saturated mutants.
During the further B- and C-ring cyclization, oxidosqualene adopted a
“chair-boat” conformation and a Markovnikov-favored 6-6-5 ring closure to produce a tricyclic C-14 cation. The C-26 methyl proton from this cationic intermediates was directly abstracted to yield (13αH)-isomalabarica-14(26),17,21-trien-3β-ol in most of
S. cerevisiae ERG7
Y510X site-saturated mutants. After successful C-ring expansion coupled with D-ring annulation, a tetracyclic protosteryl C-20 cation with the natural C-20 R-form and C-17 β-side chain configuration was generated. Interestingly, none of the ERG7Y510X site-saturated mutants was found to affect the following backbone rearrangement stage of two hydrogen groups and two additional methyl group shifts until the lanosteryl C-9 cation generation. The skeletal rearrangement byproducts have been isolated from other functionally important residues, but not in our ERG7Y510XHO H HO
C-10 cation C-14 cation lanosteryl C-9cation
achilleol A camelliol C
site-saturated mutations.117, 118 Without generation of any 9β-lanosta-7,24-dien-3β-ol further demonstrated that the cationic rearrangement really arrives at the lanosteryl C-9 position.109, 119 Finally, the deprotonation reactions alternatively occurred at either the C-8 or C-11 position to form the lanosterol or parkeol in the S. cerevisiae ERG7Y510X site-saturated mutants, respectively.
How the various single amino acid substitutions of ERG7Y510X generated diverse products with different ratio are complicated and poorly understood. According to the crystal structure of human OSC, Tyr-503 (the corresponding residue for yeast Tyr-510) is hydrogen bonded to His-232 which was considered as the only closest basic residue for the final deprotonation step.56, 105, 127 In addition, our previous study on the site-saturated mutagenesis of His-234 (the corresponding residue for human His-232) showed that the substitutions of the His-234 position might influence the stabilization of the C-13 and C-20 positions as well as the C-14 Markovnikov tertiary cation during the rings formation.115, 117 Different substitutions of ERG7H234X affected the steric and/or electrostatic interactions between the cationic intermediate and the side chain of active site residues, and resulted in the distinct products ratio. For example, the substitution of His-234 with small nonpolar hydrophobic residues facilitated the production of polycyclic products such as parkeol, protosta-12,24-dien-3β-ol, and protosta-20,24-dien-3β-ol, but interfered with the monocyclic achilleol A formation.115 Moreover, substitution for the tyrosine or phenylalanine residue on His-234 induced the steric hindrance or electrostatic repulsion to the Tyr-510 residue, relocated the possible proton acceptor, and resulted in the production of achilleol A and other altered products. In contrast, lanosterol was the sole product when TKW14 expressed the ERG7H234Y/Y510A mutant, which might be due to the released electrostatic repulsion and the returned active site environment in this double substitution mutation.115
In parallel, the fascinating and compatible result was also observed in the products profiles of ERG7Y510X described herein. Accordingly, the substitutions of larger bulky or basic amino acids on Tyr-510 tended to produce monocyclic products, whereas the small or acidic amino acids substitutions tended to produce polycyclic products. Thus, the impact on the hypothetical hydrogen-bonding basic dyad between His-234 and Tyr-510 could be imagined from the observation of individual site-directed mutation. For example, the substitution of Tyr-510 with small residues might slightly influence the position of His-234 and generate the alternative polycyclic deprotonation products. This transient disturbance increased when Tyr-510 was changed into the polar or acidic amino acids including Asp, Asn, Glu, and Gln.
The electronic density or the polar group influenced the orientation of His-234 for stabilizing the C-14 cation and resulted in the isolation of tricyclic byproduct.
Additionally, the substitution of Tyr-510 with basic group or the large amino acid such as His, Lys, Arg and Trp might cause the steric or electronic repulsion to His-234 residue. These mutants thus increased the production of monocyclic compounds. To our surprise, the ERG7Y510W mutant produced almost monocyclic but no tricyclic or tetracyclic products. In contrast, the previous ERG7H234W mutant generated almost one hundred percentage of parkeol without any other cyclization products.117 These findings suggested that the spatial influence between these two important residues might disrupt the proper orientation of substrate or the coordinated interaction among other functional residues in the enzyme active site cavity. In order to investigate the coordinative action in these two hypothetical hydrogen-bonding dyads, the ERG7H234W/Y510V and ERG7H234W/Y510W double mutants were further created and analyzed (Table 3.3). The nearly opposite result of the product profiles ratio revealed that the side chain substitution of either Tyr-510 or His-234 position, especially in the
reaction. The abundant lanosterol was observed in the ERG7H234W/Y510V mutant. The similar tetracyclic scaffold but with the alternatively deprotonated site, comparing with the previous parkeol production in the ERG7H234W mutation, supported that the smaller valine group substitution might release the steric pressure and adjust the orientation of the substrate for the proper deprotonation reaction. On the other hand, the tryptophan residue introduced in the histidine position of ERG7H234W/Y510W mutant produced almost one hundred percent of monocyclic product as ERG7Y510W mutant.
The functional importance of Tyr-510 for the stabilization of C-10 cation, His-234 for the stabilization of C-13 or C-20 cations, and their coordinative action for the deprotonation reaction could be elucidated carefully from the analysis of the product profiles described above. However, the altered hydrogen-binding networks, the newly generated catalytic base for the deprotonation reaction, even a new cyclization pathway, or novel enzymatic activity caused by mutagenesis cannot be excluded.
In conclusion, a site-saturated mutation on the Tyr-510 position as well as on the simultaneous His-234/Tyr-510 double positions provides critical evidence for the functional role of Tyr-510 involved in the complicated ERG7 catalyzed oxidosqualene cyclization. The diverse products, including truncated monocyclic, tricyclic and the altered deprotonation products, suggested the catalytic role of this residue in affecting the cationic intermediate stabilization for the cyclization stage and for final deprotonation step. Without any truncated rearrangement products also supported that the function of Tyr-510 is not crucial for the hydride/methyl groups’ migration. In addition, the isolation of (13αH)-isomalabarica-14(26),17,21-trien- 3β-ol implied the destabilization of C-14 cationic intermediate caused by the spatially influenced His-234 residue. The different combination of ERG7H234X/Y510X double mutations further supported the coordinative action of this intrinsic His-234:Tyr-510 hydrogen-bonding network.