Proposed cyclization/rearrangement pathways of TKW14C-2 expressing

Two truncated tricyclic products and the tetracyclic lanosterol were formed during the mutagenesis of Tyr99 in ERG7. In the view of the cyclization/rearrangement mechanism, the protonated epoxide of oxidosqualene initiates the formation of carbocation, following the rings annulations via cation-π interaction to a Markovnikov-favored 6-6-5 tricyclic C-14 cation without disruption at either monocyclic or bicyclic cationic location. Afterwards, the (13αH)-isomalabarica-trien-3β-ols form from elimination of proton at C-15 or C-26. Alternatively, C-ring has a process of ring expansion via a shift of the Markovnikov tertiary cation at the C-14 position to the anti-Markovnikov secondary cyclohexyl carbocation at the C-13 position, and then D-ring closure generates the protosteryl C-20 cation. Subsequently, a series of hydride and methyl group rearrangement generated the lanosteryl C-8/C-9 cation. Finally, a highly specific deprotonation abstracted the proton, which was either originally at C-9 or following the shift of the hydride from the C-9 to C-8 position, yields lanosterol. (Fig.

3.10)

To our surprise, none of truncated monocyclic, bicyclic intermediates, and (13αH)-isomalabarica-14(26),17,21-trien-3β-ol^[44,47,63] were produced in the ERG7^Y99X mutants. The formation of the truncated cyclization intermediates by the ERG7^Y99X mutants implies that the Tyr99 in ERG7 plays a crucial role in the generation of tricyclic intermediates but not in promoting the boat conformation for lanosterol B-ring formation.

To probe into the reactivity, Tyr99 may be involved in the stabilization of the Markovnikov tricyclic cation and/or the subsequently alternation of the deprotonation position with differential stereochemical control.

O

C-14 cation lanosteryl C8/C9 cation

lanosterol

Figure 3.10 proposed cyclization/rearrangement pathway occurred in the ERG7^Y99X site-saturated mutants.

3.1.4 Analysis of the ERG7

^Y99X

in the OSC homology modeling

The effects of amino acid mutations on enzymatic activity, the formation of aberrant products, and relative product proportions are complicated and are only partially understood. Together with product profiles, we applied homology modeling of the S.

cerevisiae ERG7, which was derived from the human OSC X-ray crystal structure, to

provide an insight into the relationships between mutant enzyme structure and product specificity.

(A) ERG7 w/ lanosteryl C-8 cation (B) ERG7 w/ protosteryl C-14 cation

Figure 3.11 ERG7 residues form a putative π-electron pocket. These ERG7 residues interact with the (A) lanosteryl C-8 cation and (B) protosteryl C-14 cation

The previously performed site-saturated mutagenesis experiments on the His234 and Phe445 residues of ERG7 showed that both of ERG7^F445X and ERG7^H234X produced truncated tricyclic and altered deprotonation products, indicating the catalytic role of the residues in cationic stabilization at the C-14 position for the tricyclic product, and/or the C-8/C-9 position for the final deprotonation product.^{[42-45 ]} In addition, a π–electron-rich pocket of aromatic residues around the Phe445 of ERG7 was suggested to be involved in substrate folding that affect either stabilizing the electron-deficient cationic intermediates

or availing against the equilibrium shift toward the lanosteryl C-8/C-9 cation during the rearrangement process for the formation of lanosterol. ^[33,45] Among these aromatic residues, His234 of ERG7 is hydrogen bonded to Tyr510 and located near the ceiling of the active site cavity, proximally to the B/C ring fusion and the C-8 and C-14 positions. Tyr99 is located at a distance of approximately 3.7 Å to interact with the Phe445 residue and close to the C/D ring of the substrate. (Fig. 3.11)

Scrutiny of the model suggests that Tyr99 residue is positioned in the middle side wall of the active site cavity by interacting with the C-ring, with the phenolic hydroxyl side chain pointed towards the substrate. The C-14 cation was found at a distance of approximately 4.4Å to the observed phenolic oxygen of Tyr99, a distance within the range needed to stabilize the dipole of Tyr99. (Fig. 3.11)

It is comprehensible that changes at Tyr99 strongly affect the orientation or electrostatic interaction of the phenolic oxygen of Tyr99 that is originally positioned to stabilize the C-14 cation for C-ring expansion and further D-ring closure. Perhaps some additional space for the free rotation of the hydrocarbon side chain moiety is offered through mutations at this position, resulting in abstracting protons from different position and/or orientations.

Alternatively, it is possible that mutations at Tyr99 generate a new base that leads to the Markovnikov C-B 6-6-5 cation and subsequent disproportional formation of (13αH)-isomalabarica-14E, 17, 21-trien-3β-ol and (13αH)-isomalabarica-14Z, 17, 21-trien-3β-ol (compound 1 and 2). On the other hand, a deletion at Tyr99 position may lead to a local main chain adjustment in the mutant compared with the wild type, obstructing the substrate binding and subsequent catalysis. Therefore, no product could be obtained from the ERG7^Y99 deletion mutant.

Some mutants, such as Y99Gly, Y99Ala, Y99Ser, Y99Thr and Y99Pro, which led to produce more truncated tricyclic products than the wide type product lanosterol may cause

from that the shortened reactive distance or the lost of π-electrons is insufficient for destabilization of the Markovnikov tertiary cation created at C-14 during the C-ring formation. Either a distance variation between the functional side chain of the substitution amino acid and C-14 cation, or a electrostatic change of the replaced residues may be in proportion to the products yielding ratio.

3.1.5 The analysis of product energy profile

Quantum mechanical calculations were performed mainly with Gaussian 03 for modeling the folding pathway to get an optimal stabilization of the chemicals which was in the lowest energy states. Herein we diagramed the calculation results to show the relative energy profiles among substrate and products as well as various tricyclic intermediate conformers. (Fig. 3.12)

The results are consistent with the previous results of Matsuda that showed that progressive reaction energy release was primary observed during the A, B, and C-ring formation but much less energetic for D-ring closure in neglecting the role of the enzyme.^[70]

Figure 3.12 The product energy profile from quantum mechanical calculations. The relatively energetic states of compounds A [(13αH)-isomalabarica-14(26), 17E, 21-trien-3β-ol], 2 [(13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol], 1 [(13αH)-isomalabarica-14E, 17E, 21-trien-3β-ol], and lanosterol were measure under the minimum energy states and comprised with the substrate of the enzymatic cyclization, oxidosqualene. The resulting data provided by Cheng-Hsiang Chang.

Comparing energies among various tricyclic conformers showed energy levels down from compounds A to 1 to 2. Little influence of the enzymatic effect was observed when the tricyclic cation was converted into the tetracyclic intermediate. However, the substitution of amino acid residues at different spatial positions may alter a kinetically favored double-quaternary double bond deprotonation that produces the thermodynamically favored tertiary-quaternary double bond products. Notably, the differences in energies of various tricyclic intermediate conformers could be compensated by the amino acid residues in the stabilization of the Markovnikov tricyclic cation and/or the subsequently alteration of the deprotonation position with differential stereochemical control, with compound 2 in favor of compound 1.

3.2 Functional analysis of ERG7

^W443

within S. cerevisiae

A series of amino acids residues, sequences ⁴⁴¹GAWGFSTKTQGYT⁴⁵³ within S.

cerevisiae ERG7, were subjected to both alanine-scanning mutagenesis and plasmid

shuffle selection for the identification of possible residues involved in the complementation of cyclase-deficient yeast strain CBY57.^[62] One of the three inactive mutations is Trp443Ala mutation, which failed to complement the cyclase deficiency. In my thesis, I genetically selected Trp443 site-saturated mutants (W443X) and characterized each of mutants for determination of the functional role of W443 and to investigate the effects of substitutions of this residue on other proteinogenic amino acids in terms of catalysis and product specificity.

3.2.1 Generation of site-saturated mutants of ERG7

^W443X

Tryptophan 443 of the S. cerevisiae ERG7 gene was substituted with other 18 amino acids (W443Ala has been analyzed before.^[63]) by using QuickChange site-directed mutagenesis strategy with the respective mutagenic primers. A silent mutation was concomitantly introduced to easily screen the desired mutants, according to a restriction enzyme (Sty I/ Xho I) mapping confirm. The positive mutants were digested into five fragments including 5.5Kb, 0.89Kb, 0.59Kb, 0.2Kb and 0.021Kb (unapparent) comparing with the wild type plasmids pRS314OSC which was digested into four fragments 5.5Kb, 1.1Kb, 0.59Kb and 0.021Kb (unapparent). The DNA agarose gel electrophoresis of the mapping results were shown in Appendix 2. The presence of the mutations was verified by sequence determination.

The recombinant plasmids were confirmed and transformed into TKW14C2 by the same strategies as previously described in section 3.1.1. The genetic selection of the TKW14C2[pERG7^W443X] mutants were shown in Table 3.5. The genetic selection results showed that only six mutants including W443Val, W443Leu, W443His, W443Cys,

W443Met, and W443Phe which allowed for ergosterol-independent growth. These results indicated that this position is decisive for the catalytic function of the OSC.

OSC^mut Enzyme

mapping

Acidic and amide group Hydroxyl-group W443Ser (S)

W443Thr (T)

V V

—

— Sulfur-containing W443Cys (C)

W443Met (M) W443 (wild type)

V

Table 3.5 The genetic selection results of S. cerevisiae TKW14C2 expressing the ERG7^W443X site-saturated mutagenesis.

3.2.2 Lipid extraction, column chromatography and product characterization

Each kind of recombinant yeast was incubated in 2.5L culture mediums and harvested by centrifugation. The collection and analytic protocols of NSL extract is similar to the analysis of TKW14C2[pERG7^Y99X] mutants (section 3.1.1). Four of six viable TKW14C2[pERG7^W443X] mutants including W443Val, W443His, W443Cys and W443Met yielded lanosterol as the only product with molecular mass of m/z=426. The other two viable TKW14C2[pERG7^W443X] mutants, W443Ala (according to previous analysis^[63]) and W443Lys revealed two monocyclic triterpenoid products with a molecule mass of m/z = 426: achilleol A and camelliol C comparing with the authentic sample. (Fig.3.13) The product profiles of each mutant are summarized in Table 3.6.

Figure 3.13 Electron-impact mass spectra of two monocyclic triterpenoid products Achilleol A Camelliol C

Achilleol A

Camelliol C

Products profile ratio (%) amino acids

substitution Ergosterol

supplement no products Lanosterol Achilleol A Camelliol C

Gly

Table 3.6 The products profile of S. cerevisiae TKW14C2 expressing the ERG7^W443X site-saturated mutagenesis.

3.2.3 Proposed cyclization/rearrangement pathways of TKW14C2 expressing ERG7

^W443X

The first report site-directed mutant that can produce both monocyclic achilleol A and camelliol C, the formation of various incomplete cyclization products was found in ERG7^Tyr510X. The position of Tyr510 was supposed to be involved in the stabilization of cationic intermediates during the exposide protonation and A-ring cyclization.^[46-47] The formation of achilleol A and camelliol C were identified as evidence for premature truncation of C-10 cationic intermediates formation following the proton abstraction from Me-25 or C-1 position. (Fig. 3.14)

C-10 cation lanosteryl C8/C9 cation

+ ⁸

Figure 3.14 S. cerevisiae OSC ERG7^W443X mutants convert oxidosqualene to

3.2.4 Analysis of the ERG7

^W443X

in the OSC homology modeling

The multiple sequence alignment analysis showed that the Trp443 was highly conserved in most cyclase, the Trp443 of S. cerevisiae ERG7 corresponds to F363 in A.

acidocaldarius SHC and to W470 in A.thaliana CAS. These three residues are all aromatic

amino acids; however their functional role analysis during catalytic cyclization mechanism has not been suggested and reported before.

According to the previous studies, the achilleol A and camelliol C were also produced from the other cyclase-inactive mutants, Lys448Ala which were previously identified from the region upstream of the putate active site in our laboratory by ergosterol complement experiment. Lys448 are located at the flexible loop region opposite to the position of the essential Asp456 and displayed interactions to hold the correct conformation in dimeric association with two amino acids, Phe426 and Asn332. Replacing Lys448 with Ala was supposed to disrupt the electrostatic interaction between subunits or held the cyclization/rearrangement cascade at the intermediate stage, thus forming only the initially cyclized A-ring.^[62-63]

In the previous homology model studies, the Trp443 was supposed to be positioned spatially opposite to the Asp456, below the molecular plain and close to the high-energy C-10 (lanosterol numbering) cationic intermediate. The Trp443 was suggested to be at the nearest neighbor to the active site residues and thereby stabilize the high-energy C-10 cation intermediate during the concerted process of epoxide opening and A-ring formation.

Substitution of Trp with Ala might disrupt the steric or cation- electronic effect between substrate and enzyme; elongation of the cyclization cascade would thus be inhibited and the reaction be held at monocyclic triterpenes.^[62-63] However, in my homology modeling analysis, the Trp443 is positioned spatially above to the Asp456 and the molecular plain whereas it seems to be far from substrate (10.28Å between oxygen of Trp and C-10 of lanosterol; 10.1Å between oxygen of Trp and C-2 of lanosterol). (Fig. 3.15)

Figure 3.15 Local views of the homology modeled Asp456, Trp443, Lay448, and Phe445 positions in S. cerevisiae ERG7 structure based on the X-ray structure of lanosterol-complexed human OSC and determined by using the Insight II Homology program.

Obviously, Trp443 is not located in the putative π–electron pocket; however the site-saturated mutagenesis results showed that only W443Val, W443Leu (aliphatic residues), W443Cys, W443Met (sulfur-containing residues), W443Phe (aromatic residue) and His (basic residue) were able to complement to the OSC deficient strain and yield lanosterol. This observation revealed that W443 position is indispensable; changes of the side chain at W443 position may influence the interactive distances for the proper substrate binding and subsequent catalysis. On the other hand, two inactive mutants, W443A (aliphatic residue) and W443K (basic residue) produced achilleol A as a major product and camelliol C as a minor product. The formation of the truncated monocyclic intermediated suggested that Trp residue may also play an crucial role both in influencing the substrate

prefolding and stabilizing the epoxide protonation and inducing A-ring formation via the generation of the C-10 cation.

Moreover, the exact reason for the higher accumulation of achilleol A over that of camelliol C, and the production of achilleol A whenever camelliol C is produced, remain unclear. Furthermore, whether the Trp443 interact with the substrate directly or via the other residues in the active site pocket, needs more mutagenesis at neighboring residues and homology modeling of the W443X mutants, in order to clarify the functional roles of Trp443.

Chapter4 Conclusions

Site-directed mutagenesis is a molecular biology technique in which a mutation is created at a defined site in a DNA molecule, and site-saturated mutagenesis means the substitution of specific sites with other 19 proteinogenic amino acids. This technique was applied to obtain a detailed understanding of structure-function relationships for the putative active sites in the enzyme.

In our studies, site-saturated mutagenesis coupled with product isolation and characterization of the mutations at Tyr99 and Trp443 position of OSC ERG7 within S.

cerevisias revealed their catalytic function in affecting the cyclization/rearrangement

mechanism. Both of these two residues were suggested play crucial roles in enzyme catalytic cyclization/rearrangement. Herein we summarize several important conclusions of our studies:

4.1 The functional analysis of TKW14C2[pERG7

^Tyr99X

]

(1) The TKW14C2[pERG7^Tyr99X] expressed ERG7^Tyr99X as its sole oxidosqualene cyclase.

The genetic selection results showed that several Tyr99X mutants could complement to ergosterol-deficient growth except the deletion of Tyr99 as well as the mutation of Y99N, Y99H in ERG7.

(2) Several mutants including Y99A, Y99G, Y99I, Y99D, Y99S, Y99T, Y99F, and Y99P produced two novel products with a molecular mass of m/z = 426 except lanosterol, which are both truncated tricyclic triterpenoid products (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol and (13αH)-isomalabarica-14E, 17E, 21-trien-3β-ol, identified by ¹H and

13C NMR for the first time. The product profile of ERG7^Y99X demonstrates the truncation of the cyclization/rearrangement cascade at chair-boat 6-6-5 tricyclic

C-15 position with different stereochemical preferences.

(3) The functional role of ERG7^Y99 is suggested to affect both chair-boat 6-6-5 tricyclic Markovnikov cation stabilization and the stereochemistry of the protons at the C-15 position for subsequent deprotonation, but not to enforce the boat conformation for lanosterol B-ring formation.

(4) In homology modeling analysis, the phenolic oxygen of Tyr99 residue is at a distance of approximately 4.4Å from the C-14 cation, and its location is differently in space from that to His234 and Phe445 to the common C-14 cation which affects the orientation or electrostatic interaction between the enzyme and its cationic intermediate, and results in the abstraction of a proton form a different position or orientation.

Therefore, changes at Tyr99 or a deletion of this position may strongly impact the structure and lead to an adjustment of the active site and result in obstruction of substrate binding and catalysis.

(5) The product energy profile from quantum mechanical calculations suggested that the energetics of stereochemical control during the tricyclic Markovnikov cation deprotonation step could be affected by the inclusion of these enzymatic effects. It may be the reason why the (13αH)-isomalabarica-14Z, 17E, 21-trien-3β-ol was produced as a major product in Y99Gly, Y99Ala, Y99Ser, Y99Thr and Y99Pro mutants.

4.2 The functional analysis of TKW14C2[pERG7

^Trp443X

]

(1) In the previous alanine-scanning mutagenesis and plasmid shuffle selection of the

441GAWGFSTKTQGYT⁴⁵³ within this of S. cerevisiae ERG7, Trp443Ala was one of the inactive mutants. The following ergosterol complementation experiment uncovered that ERG7^Trp443Ala produced two monocyclic triterpenen products concomitantly, achilleol A and camelliol C.

(2) The genetic selection demonstrated that only six mutants including Trp443Val, Trp443Leu, Trp443His, Trp443Cys, Trp443Met, and Trp443Phe which allowed for ergosterol-independent growth and yielded lanosterol as an only product with molecular mass of m/z=426. Whereas one of inactive mutant, Trp443Lys, revealed two monocyclic triterpenoid products with a molecule mass of m/z = 426: achilleol A and camelliol C, as well as the observation in the ERG7^W443A mutant.

(3) The formation of achilleol A and camelliol C were identified as evidence for premature truncation of C-10 cationic intermediates following the proton abstraction from Me-25 or C-1 position. This finding suggested that Trp residue may also play a crucial role both in influencing the substrate prefolding and stabilizing the epoxide protonation and A-ring formation to generate C-10 cation. However, the exact reason for the higher accumulation of achilleol A over that of camelliol C, and the production of achilleol A whenever camelliol C is produced, remain unclear.

(4) Although Trp443 is not located in the putative π–electron pocket and positioned spatially far from A-ring of lanosterol, it might provide an interaction with the neighboring residues to stabilize the carbocationic intermediates produced during protonation of epoxide and subsequent A-ring formation.

Chapter 5 Future Works

For the Tyr99 functional analysis, our results showed that how the structure-function relationships of the OSC via the expression of ERG7^Y99X site-saturated mutants in S.

cerevisiae. However, our substantiation of Tyr99 functional role contradicted the

supposition, which Tyr98 of human OSC is spatially positioned to enforce the energetically unfavorable boat conformation of OS for lanosterol B-ring formation via pushing the methyl group at C-8 (lanosterol numbering) below the molecular plane. The expression of site-directed mutants of Tyr98X^hOSC will be carried out to identify the function of this conserved residue.

Furthermore, the HEM1 ERG7 ERG1 triple knockout mutant which is the yeast strain with the deletion of both oxidosqualene cyclase (ERG7) and squalene epoxidase (ERG1). This triple knockout strain will be developed for the in vitro analysis of the mutated oxidosqualene cyclase via the addition of the substrates. This oxidosqualene free strain will prevent the interference due to the downstream enzymes and consequently ensure the more detailed understanding for the catalytic function of the putative active-sites.

Chapter6 References

1. I. Abe, M. Rohmer, G. D. Prestwich Chem. Rev. 93, 2189(1993)

2. K. U. Wendt, G. E. Schulz, E. J. Corey, D. R. Liu Angew. Chem. Int. Ed. 39, 2812(2000).

3. R. Xu, G. C. Fazio, S. P. T. Matsuda Phytochemistry 65, 261(2004).

4. I. Abe, Y. F. Zheng, G. D. Prestwich Biochemistry 37, 5779(1998).

5. T. Dang, I. Abe, Y. F. Zheng, G. D Prestwich Chem. & Bio. 6, 333(1999).

6. M. J. R. Segura, B. E. Jackson, S P. T. Matsuda Nat. Prod. Rep. 20, 304(2003).

7. T. Schulz–Gasch, M. Stahl J. Comput. Chem. 24, 741(2003).

8. E. J. Corey, S. C. Virgil J. Am. Chem. Soc. 112, 6429(1990).

9. E. J. Corey, S. C. Virgil J. Am. Chem. Soc. 113, 4025(1991).

10. E. J. Corey, S. C. Virgil, S. Sarshar J. Am. Chem. Soc. 113, 8171(1991).

11. E. J. Corey, S. C. Virgil, R. Liu, S. Sarshar J. Am. Chem. Soc. 114, 1524(1992).

12. E. E. van Tamelen J. Am. Chem. Soc. 104, 6480(1982).

13. D. Rittenberg, R. Schoenheimer J. Bio. Chem. 235(1937).

14. K. Bloch, D. Rittenberg J. Bio. Chem. 625(1942).

15. R. B. Woodward, K. Bloch J. Am. Chem. Soc. 75, 2024(1953).

16. R. K. Maudgal, T. T. Tchen, K. Bloch J. Am. Chem. Soc. 80, 2589(1958).

17. J. W. Cornforth, R. H. Cornforth, C. Donninger, G. Popják, Y. Shimizu, S. Ichii, E.

Forchielli, E. Caspi J. Am. Chem. Soc. 87, 3224(1965).

18. E. J. Corey, W. E. Russey, P. R. Ortiz de Montellano J. Am. Chem. Soc. 88, 4750(1966).

19. E. E. van Tamelen, J. D. Willett, R. B. Clayton, K. E. Lord J. Am. Chem. Soc. 88, 4752(1966).

Chem. 242, 4182(1967).

21. D. H. R. Barton, T. R. Jarman, K. C. Watson, D. A. Widdowson, R. B. Boar, K.

Damps Perkin Trans. 1, 1134(1975).

22. G. Ourisson, M. Rohmer, K. Poralla Annu. Rev. Microbiol. 41, 301(1987).

23. G. Ourisson, M. Rohmer, R. Anton Recent. Adv. Phytochem. 13, 131(1979).

24. M. Rohmer, P. Bouvier, G. Ourisson Proc. Nadl. Acad. Sci. 76, 847(1979) 25. W. S. Johnson, S. J. Telfer, S. Cheng, U. Schubert J. Am. Chem. Soc. 109,

2517(1987).

26. W. S. Johnson, S. D. Lindell, J. Steele J. Am. Chem. Soc. 109, 5852(1987).

27. C. J. Buntel, J. H. Griffin J. Am. Chem. Soc. 114, 9771(1992).

28. Z. Shi, C. J. Buntel, J. H. Griffin Proc. Nadl. Acad. Sci. 91, 7370(1994).

29. R. Kelly, S. M. Miller, M. H. Lai, D. R. Kirsch Gene 87, 177(1990).

30. K.U. Wendt Angew. Chem. Int. Ed. 44, 3966(2005).

31. K.U. Wendt, K. Poralla, G.E. Schulz, Science 277, 1811(1997).

32. K.U. Wendt, A. Lenhart, G.E. Schulz, J. Mol. Biol. 286, 175 (1999).

33. R. Thoma, T. Schulz-Gasch, B. D’Acry, J. Benz, J. Aebi, H. Dehmlow, M. Hennig, M.

Stihle, A. Ruf Nature 432, 118(2004).

34. E. J. Corey, H. Cheng, C. H. Baker, S. P. T. Matsuda, D. Li, X. Song J. Am. Chem.

34. E. J. Corey, H. Cheng, C. H. Baker, S. P. T. Matsuda, D. Li, X. Song J. Am. Chem.

在文檔中 利用飽和定點突變方法研究酵母菌氧化鯊烯環化酵素內的假設活性區胺基酸對於催化環化/重組反應的影響 (頁 75-0)