Total Synthesis of Cephalotaxus Alkaloids and Synthetic Studies Toward Bazzanin K

Home , Acyloin, Aromatization, Cyclophane, Methanesulfonyl chloride

AN ABSTRACT OF THE DISSERTATION OF

Xuan Ju for the degree of Doctor of Philosophy in Chemistry presented on May 22, 2019.

Title: Total Synthesis of Cephalotaxus Alkaloids and Synthetic Studies toward Bazzanin K

Abstract approved: ______Christopher M. Beaudry

Dearomatization reactions exist broadly in biosynthesis and chemical synthesis. The highly functionalized atoms of arenes can be masked by it aromaticity, and upon dearomatization, those reactive atoms can be readily applied to bond formation and further manipulation. The total synthesis of (–)-cephalotaxine and (–)- homoharringtonine is described via an oxidative furan opening-spontaneous transannular Mannich reaction. The first generation route involved the development of coupling reaction between 3-methoxyfuran-2-carboxylate and bromide to form diarylmethane. Due to the lack of electrophilicity in the ester functionality, the formation of a macrocycle was not observed. The second generation route utilized a Friedel-Crafts alkylation to avoid this problem, enabling the preparation of the macrocycle for the key transformation. Racemic cephalotaxinone was obtained by the oxidative furan opening-spontaneous transannular Mannnich reaction of the macrocycle. A Noyori asymmetric hydrogenation converted racemic cephalotaxinone to (–)-cephalotaxine in excellent yield and enantioselectivity (krel=278), (–)- cephalotaxine was advanced to (–)-homoharringtonine via a three-step sequence. The undesired enantiomer of cephalotaxinone could be recycled through an acid-mediated racemization.

Molecular chirality plays a critical role in chemistry, biology, and medicine. Identification of chirality in molecules without sp3-hybridized stereogenic carbon atoms is not straightforward. Bazzanin K is a macrocyclic bis(bibenzyls) with diastereotopic protons at its two methylene position, indicating the possibility of conformational chirality. We describe a synthetic approach toward bazzanin K by a double Suzuki coupling. Sequential Suzuki couplings were tested and the desired terphenyl was isolated as two disastereomers. Ring closing metathesis of terphenyls furnished the phenanthrene moiety of Bazzanin K. The one-pot three-component Suzuki coupling reaction is under investigation.

Total Synthesis of Cephalotaxus Alkaloids and Synthetic Studies toward Bazzanin K

by Xuan Ju

A DISSERTATION

submitted to

Oregon State University

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Presented May 22, 2019 Commencement June 2020

Doctor of Philosophy dissertation of Xuan Ju presented on May 22, 2019

APPROVED:

Major Professor, representing Chemistry

Head of the Department of Chemistry

Dean of the Graduate School

I understand that my dissertation will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my dissertation to any reader upon request.

Xuan Ju, Author

ACKNOWLEDGEMENTS

I would like to express my thanks to my parents for their love and unconditional support through my graduate career at Oregon State University. I would like to thank my advisor, Professor Chris Beaudry, for his guidance over the last six years. He taught me the knowledge of conducting scientific researches and methods of solving problems critically. His mentorship has greatly contributed to my success in this program. I would also like to thank both past and present members of the Beaudry research group and members of the organic chemistry division for their support, especially Yi Lu and Xiaojie Zhang for their encouragement, support and friendship of throughout my degree program.

CONTRIBUTION OF AUTHORS

Patrick Salvo assisted with data collection for chapter 5.

TABLE OF CONTENTS

Page

1 Introduction and Background……………………………………………………….1

1.1 Introduction of Dearomatization Reaction……………………………………...1

1.2 Dearomatization in the Syntheses of Strychnine……………………………….2

1.3 Dearomatization Researches in the Beaudry group…………….………………4

2 Introduction of Cephalotaxine and Overview of the Previous Total Syntheses...... 7

2.1 Isolation of Cephalotaxine and Homoharringtonine ………………………...... 7

2.2 Chronic Myeloid Leukemia (CML)………………………………………...... 7

2.3 Previous Total Syntheses of Cephalotaxine ………………………………...... 9

2.3.1 Sequential Construction - Construction of D/E ring, then C ring……...... 9

2.3.2 Sequential Construction - Construction of C/E ring, then D ring…...... 11

2.3.3 Sequential Construction - Construction of C/D ring, then E ring……...... 12

2.3.4 Construction of Two Rings in One Step……….….….……………...... 14

3 First Generation Route Toward (–)-Cephalotaxine and (–)-Homoharringtonine….18

3.1 Retrosynthetic Analysis of Cephalotaxine: First Generation Route…………..18

3.2 Preparation of Furan…………………………………………………………..18

3.3 Synthetic Efforts toward Macrocycle…………………………………………20

3.4 Supplemental Data………………………………………………………...... 23

4 Total Synthesis of (–)-Cephalotaxine and (–)-Homoharringtonine: Second Generation Route…………………………………………………………………….41

4.1 Retrosynthetic Analysis of Cephalotaxine: Second Generation Route………..41

4.2 Preparation of the Macrocycle………………………………………………...41

TABLE OF CONTENTS (Continued)

Page

4.3 Synthesis of Cephalotaxinone via Furan oxidation-Transannular Mannich Reaction ……………………………………………………………………………..44

4.4 Total Synthesis of (–)-Cephalotaxine and (–)-Homoharringtonine…………...47

4.5 Future Plan…………………………………………………………………….50

4.6 Supplemental Data ……………………………………………………………51

5 Introduction and Synthetic Studies toward Bazzanin K…………………………...95

5.1 Conformational Chirality……………………………………………………...95

5.2 Introduction of Macrocyclic Bisbibenzyls…………………………………….96

5.3 Introduction of Bazzanin K……………………………………………………97

5.4 Retrosynthetic Analysis……………………………………………………….98

5.5 Synthesis of Boronic Esters…………………………………………………...99

5.6 Sequential Suzuki Cross Couplings………………………………………….100

5.7 Double Suzuki Coupling……………………………………………………..103

5.8 Ring Closing Metathesis……………………………………………………..104

5.9 Supplemental Data…………………………………………………………...105

6 Summary and Furture Work……………………………………………………...139

LIST OF FIGURES

Figure Page

2.1 Cephalotaxus Alkaloids …………………………………………………………..7

2.2 Philadelphia Chromosome and Imatinib (Gleevec)……………………………….8

2.3 Mechanism of Action of HHT ……………………………………………………8

2.4 C/D/E Ring System of Cephalotaxine……………………………………….……9

4.1 Binding Site of Narcilasine, Lycorine and Homoharringtonine…………………50

5.1 Recent Conformational Chirality Researches in Beaudry Group………………..95

5.2 Three Types of MBBs……………………………………………………………96

5.3 Structure of Asterelin A, Cavicularin and Bazzanin K…………………………..97

5.4 Structure of Bazzanin K and Cavicularin…………………….………………….97

5.5 Stereochemical Analysis of Terphenyl via NOESY Correlation Spectroscopy..103

LIST OF TABLES

Table Page

4.1 Reaction Condition Screen for the Friedel–Crafts Alkylation…………………...43

4.2 Reaction Conditions Screened for Synthesizing Cephlotaxinone………………..45

4.3 Previous Total Syntheses of Cephalotaxine………………………………………49

5.1 Reaction Conditions Screen for First Suzuki Coupling………………………….101

5.2 Reaction Conditions Screen for Second Suzuki Coupling………………………102

LIST OF SCHEMES

Scheme Page

1.1 Examples of Dearomatization………………….………………………………….1

1.2 Dearomatization in Woodward’s Total Synthesis of Strychnine………………….3

1.3 Dearomatization in Vanderwal’s Total Synthesis of Strychnine……….…………3

1.4 Dearomatization in the Total Synthesis of Cavicularin …………………………..4

1.5 Dearomatization in the Total synthesis of Arundamine…………………………..5

2.1 Semmelhack’s Total Synthesis of (±)-Cephalotaxine……………………………10

2.2 Mori’s Total Synthesis of (–)-Cephalotaxine……………………………………10

2.3 Chandrasekhar’s Total Synthesis of (±)-Cephalotaxine…………………………11

2.4 Gin’s Total Synthesis of (–)-Cephalotaxine……………………………………..11

2.5 Fuchs’ Total Synthesis of (±)-Cephalotaxine……………………………………12

2.6 Weinreb’s Total Synthesis of (±)-Cephalotaxine………………………………..13

2.7 Hanaoko’s Total Synthesis of (±)-Cephalotaxine………………………………..13

2.8 Li’s Formal Synthesis of (–)-Cephalotaxine……………………………………..14

2.9 Mariano’s Total Synthesis of (±)-Cephalotaxine………………………………...14

2.10 Li’s Total Synthesis of (±)-Cephalotaxine……………………………………...15

2.11 Ishibashi’s Total Synthesis of (–)-Cephalotaxine………………………………15

3.1 Retrosynthetic Analysis of Cephalotaxine……………………………………….18

3.2 Wasserman-based Furan Synthesis………………………………………………19

3.3 Alternative Way of Preparing the Furan…………………………………………20

3.4 Efforts toward Reduction of the Ester…………………………………………...20

3.5 Conversion of Ester to Other Functionalities……………………………………21

LIST OF SCHEMES (Continued)

Scheme Page

3.6 Coupling Conditions for Preparing Diarylmethanes……………………………..21

3.7 Attempted Diarylmethane Synthesis……………………………………………..22

4.1 Retrosynthetic Analysis of Cephalotaxine: Second Generation Route…………..41

4.2 Preparation of the Furan and Benzyl Chloride ………………………………….42

4.3 Conditions for Removal of TBS Group………………………………………….43

4.4 Base-mediated Cyclization to Form the Macrocycle…………………………….44

4.5 Proposed Mechanism to Enamine 123 and Amine 124………………………….46

4.6 Racemic Synthesis of Cephalotaxinone………………………………………….46

4.7 Attempts to Synthesize (–)-Cephalotaxinone with Chiral Acids………………...47

4.8 Kinetic Resolution of Cephalotaxinone………………………………………….48

4.9 Racemization of (+)-Cephalotaxinone…………………………………………...48

5.1 Retrosynthetic Analysis of Bazzanin K………………………………………….98

5.2 Preparation of Boronic Ester 141………………………………………………...99

5.3 Preparation of Boronic Ester 142……………………………………………….100

5.4 Atropisomerization of Terphenyls……………………………………………...103

5.5 Double Suzuki Coupling………………………………………………………..104

5.6 Ring Closing Metathesis of Terphenyls………………………………………...104

Chapter 1: Introduction and Background

1.1 Introduction of Dearomatization Reactions

Since the first time benzene was proposed as a cyclohexatriene structure by Kekule in 1865, it raised great interest in the physical properties and applications of aromatic compounds. Aromaticity is used to describe the enhanced chemical stability in aromatic compounds, compared to similar non-aromatic molecules. Despite the high resonance energy of aromatic compounds, dearomatization has been applied extensively in biosynthesis and chemical synthesis, 1 such as redox reactions, enzymatic dearomatizations, thermo/photochemical cyclizations, alkylative dearomatizations and transition metal-mediated dearomatizations.2

MeO MeO Li, NH3 EtOH a) 85% OMe OMe

O OH PhI(OAc)2 (CF3)2CHOH; b) Me2CO OH 51% O

A. eutrophus B9 c) CO2H CO H 79%, >95% ee 2 OH OH

MeO d) hν OMe + 1:1, 65% OMe H H

O OH CO Me e) 2 1) tBuOK, tBuOH Br 2) NaOMe, MeOH CO2Me

OMe Li OEt CHO N H 1) f) 2) MeI, CO, HMPA OEt

Cr(CO)3 3) NaOEt, MeI 53%, >95% ee O Scheme 1.1 Examples of Dearomatization

Some representative dearomatization reaction are shown in Scheme 1.1. Redox reaction is the most widely used and synthetically versatile dearomatization reactions. 2

Birch reduction (Scheme 1.1a) 3 and hydrogenation of aromatic compounds are classical reductive dearomatization reaction. Oxidative dearomatization is typically performed with mild and selective oxidant like hypervalent iodine reagents (Scheme 1.1b).4 Enzymatic dearomatizations are usually applied on simple arene substrates to prepare chiral building blocks. As shown in Scheme 1.1c, Myers and coworker developed a dihydroxylation of benzoic acid with Alcaligenes eutrophus strain B9 with high enatioselectivity. 5 The thermo and photochemical cyclization of aromatic compounds can deliver intriguing polycyclic molecular skeleton. Wender group published a [3+2] cycloaddition of arene to afford two regioisomeric cyclopropanes (Scheme 1.1d). 6 Alkylative dearomatization was applied in Corey’s synthesis to generate a spirocyle (Scheme 1.1e).7 Due the steric hindrance of metal species, the transition metal-mediated dearomatization can render nucleophilic addition stereoselective, which is observed in the nucleophilic addition/acylation/alkylation to form the aldehyde shown in Scheme 1.1f.8 It is well known that aromatic compounds are desirable building blocks. These ubiquitous compounds are easily functionalized and naturally protected by their aromaticity. Upon dearomatization, they represent 5 or 6 contiguous sp2 hybridized carbon (or N, O) atoms, which can readily participate in bond formation and further manipulation for constructing complicated scaffolds.

1.2 Dearomatization in the Syntheses of Strychnine

Dearomatization displays a crucial role in total synthesis of complex natural products. Strychnine serves as a useful case study. Strychnine (1) is a terpene indole alkaloid known for its congested structure. Five of its contiguous stereogenic carbon atoms are included in its E ring, and one is quartnary carbon center. There have been a lot of excellent total syntheses of this natural product, and two classic strategies developed by Woodward9 and Vanderwal10 utilized dearomatization as their key feature. The first total synthesis of strychnine was accomplished by Woodward in 1954. As shown in Scheme 1.1, a 3,4-dimethoxyphenol was used as a precursor for constructing G and F ring of strychnine. Ozonylsis opened the aromatic ring to give diester 3. 3

Exposure of diester 3 to HCl in methanol enabled deacylation, lactamization and isomerization to the production of pyridone 4, which was converted to strychnine after further manipulations. The arene dearomatization provided an abundant source of contiguous functionalized carbon atoms, perfectly positioned for constructing the framework of strychnine.

NTs NTs O3, AcOH H2O CO2Et CO2Et OMe N 29% N CO Me Ac Ac 2 CO Me OMe 2 2 3

N NTs C D E HCl steps A H CO2Et B MeOH N N 75% CO2Me GH H F O O O H 4 strychnine (1) Scheme 1.2 Dearomatization in Woodward’s Total Synthesis of Strychnine

Vanderwal and coworkers reported an efficient synthesis of strychnine in 2011. A dearomatization and cycloaddition was applied to rapidly construct the functionalized tetracyclic ring system. N-allyl tryptamine 5 attacked 2,4-dinitrobezenepyridium chloride to form aminal 6, which spontaneously underwent an electrocyclic ring opening to generate imine 7, producing Zincke aldehyde 8 after basic hydrolysis. Tetracyclic aldehyde 9 was prepared by treatment of diene 8 with tBuOK in THF through a Diels–Alder cycloaddition and olefin isomerization. The dearomatization enabled a conversion from a planar pyridium scaffold to the three-dimensional architecture of strychnine, especially the most congested E ring.

O2N N N NO2 N Cl DNP N HN N EtOH N H N H 5 H 6 7 NDNP

N C N N E D NaOH KOtBu steps A B H N 81% THF, 80℃ F N GH H N 64% H H H O O CHO CHO H 8 9 strychnine (1) Scheme 1.3 Dearomatization in Vanderwal’s Total Synthesis of Strychnine 4

1.3 Dearomatization Research in the Beaudry Group

Beaudry’s research group has been actively exploring total synthesis of natural products and methodology development featuring dearomatization. Our main investigations include pyridinium oxide cycloadditions and rearomatization of heteroarenes such as pyrones and furans. Two published total syntheses of natural products utilized Diels–Alder cycloaddition as the key transformation. Cavicularin (10) is a cyclophane natural product isolated by Asakawa and co-workers in 1996. Its A ring is distorted in a boat-shaped configuration, displaying a strained molecular architecture.11 In the total synthesis of cavicularin, a dearomatization- rearomatization from a pyrone to phenol is shown (Scheme 1.3).12 An intramolecular Diels–Alder reaction between the pyrone and vinyl sulfone moiety 11 followed by elimination of phenylsulfonic acid afforded unsaturated bicycle 13, which underwent rapid retro-Diels–Alder reaction to produce phenol 14. The strained A ring of the cavicularin was constructed through the transformation.

O OH O O OMe O O MeO O MeO PhO2S O C O ligand, 3Å MS PhO2S O OH OH –HOSOPh EtOAc, 45℃ D

B OMe MeO MeO MeO MeO MeO 11 12 13

CF3 HO C MeO C S O O –CO 2 steps A D HN N CF3 A D H N HO OH OMe N B B MeO ligand HO HO 14 cavicularin (10) Scheme 1.4 Dearomatization in the Total Synthesis of (+)-Cavicularin

Bis(indole) alkaloids from Arundo donax,13 such as arundamine (15), were prepared by Beaudry group to measure their racemization half-lives, verifying their conformational chirality.14 An ynindole Diels–Alder reaction was implemented to form polycyclic compound 17, which fragmented and tautomerized to reveal phenol 19. It is speculated 5

that the electron pair on the nitrogen participates in the aromaticity of the indole instead of donating to the alkyne, leading to the occurrence of the ynindole Diels–Alder reaction.

BnMeN O BnMeN O BnMeN O

O PhMe O O 150℃ N N N 90% O O O NBoc N N Boc Boc 16 17 18

BnMeN O HMeN O steps N N HO NMe2 HO N N H Boc 19 arundamine (15) Scheme 1.5 Dearomatization in the Total synthesis of Arundamine

In summary, the utility of dearomatization offers a novel approach toward construction of architecturally complex natural products. The highly functionalized atoms of aromatic compounds could be used for bond formation and rearomatization. Considering our previous research is mainly related to arene-arene transformation, we are planning to take advantage of the reactive hydrocarbons in aromatic structures and apply them for constructing polycyclic architectures. One of our proposed strategies involves oxidative opening of furan and trannsannular Mannich reaction to synthesize Cephalotaxus alkaloids.

1 a) Thiele, B; Rieder, O.; Golding, B. T.; Müller, M.; Boll, M. J. Am. Chem. Soc. 2008, 130, 14050–14051; b) Kung, J. W.; Baumann, S.; Von Bergen, M.; Müller, M.; Hagedoorn, P.-L; Hagen, W. R.; Boll, M. J. Am. Chem. Soc. 2010, 132, 9850–9856. 2 a) Roche, S. P.; Porco Jr., J. A. Angew. Chem. Int. Ed. 2011, 50, 4068–4093; b) Mortier, Jacques. Arene chemistry: reaction mechanisms and methods for aromatic compounds. John Wiley & Sons, 2015. 3 Altman, R. A.; Nilsson, B. L.; Overman, L. E.; Read de Alaniz, J.; Rohde, J. M.; Taupin, V. J. Org. Chem. 2010, 75, 7519–7534. 4 Guerard, K. C.; Chapelle, C.; Giroux, M. A.; Sabot, C.; Beaulieu, M. A.; Achache, N.; Canesi, S. Org. Lett. 2009, 11, 4756–4759. 5 Myers, A. G.; Siegel, D. R.; Buzard, D. J.; Charest, M. G. Org. Lett. 2001, 3, 2923– 2926. 6 Wender, P. A.; Howbert, J. J. J. Am. Chem. Soc. 1981, 103, 688–690. 7 Corey, E. J.; Girotra, N. N.; Mathew, C. T. J. Am. Chem. Soc. 1969, 91, 1557–1559. 8 Kündig, E. P.; Cannas, R., Laxmisha, M., Ronggang, L.; Tchertchian, S. J. Am. Chem. Soc. 2003, 125, 5642–5643. 9 Woodward, R. B.; Cava, M. P.; Ollis, W. D.; Hunger, A.; Daeniker, H. U.; Schenker. K. J. Am. Chem. Soc. 1954, 76, 4749–4751. 10 Martin, D. B. C.; Vanderwal, C. D. Chem. Sci. 2011, 2, 649–651. 11 Toyota, M.; Yoshida, T.; Kan, Y.; Takaoka, S.; Asakawa, Y. Tetrahedron Lett. 1996, 37, 4745–4748. 12 Zhao, P.; Beaudry, C. M. Angew. Chem. Int. Ed. 2014, 53, 10500–10503. 13 Passalacqua, N. G.; Guarrera, P. M.; Fine, G. D. Fitoterapia 2007, 78, 52–68. 14 Chen, J.; Ferreira, A. J.; Beaudry, C. M. Angew. Chem. Int. Ed. 2014, 53, 11931– 11934.

Chapter 2 Introduction of Cephalotaxine and Overview of the Previous Total Syntheses

2.1 Isolation of Cephalotaxine and Homoharringtonine

Cephalotaxine (20), the most abundant alkaloid isolated from plum yew Cephalotaxus harringtonii, possesses l-azaspiro[4.4]nonane moiety fused with a benzazepine ring.15 Homoharringtonine (HHT, 21), the naturally occurring ester derivative of cephalotaxine, shows inhibitory activity against P-388 leukemia cells with IC50 value of 17 nM.16 As a protein translocation inhibitor, HHT was approved by the US FDA for the treatment of chronic myeloid leukemia in 2012.17 Despite huge efforts toward its synthesis, commercial HHT still involves the semi-synthesis from plant-derived cephalotaxine.

O O N N O O O H H O HO HO OH OMe OMe CO2Me cephalotaxine (20) homoharringtonine (21) Figure 2.1 Cephalotaxus Alkaloids

2.2 Chronic Myeloid Leukemia (CML)

Chronic myeloid leukemia, a cancer of white blood cells, is caused by a chromosomal translocation. A gene piece from chromosome 9 is fused with gene piece of chromosome 22, causing an abnormal chromosome 22 which is known as Philadelphia chromosome, and it contains a fusion gene called BCR-ABL. 18 The traditional treatment of CML is using tyrosine-kinase inhibitors (TKIs), such as imatinib (Gleevec, 22). Imatinib can bind to the ATP-binding site of the ABL kinase domain of BCR- ABL, decreasing the activity of this protein and inhibiting protein tyrosine phosphorylation.19 8

N O

N N N N H N imatinib (22)

Figure 2.2 Philadelphia Chromosome and Imatinib (Gleevec)

Drug resistance has emerged due to one particular mutation (T315I) in the fusion gene, rendering TKIs ineffective. Combination strategies involving imatinib and other approaches have been developed to overcome the problem, including HHT (Omacetaxine mepesuccinate). HHT can bind to the peptidyl tranferase center A site of the ribosome and prevent the correct positioning of the amino side chain of incoming tRNA, inhibiting peptide-bond formation.20

Figure 2.3 Mechanism of Action of HHT20

The structural complexity and therapeutic potential of their ester derivatives makes cephalotaxine and homoharringtonine attractive synthetic targets. We planned to 9

develop an efficient and innovative strategy to synthesize these natural products for future investments.

2.3 Previous Total Syntheses of Cephalotaxine

Since Weinreb reported the first total synthesis of cephalotaxine in 1972,21 there has been a substantial amount of synthetic studies. Strategically, these syntheses can be divided into two major types depending on the construction of CDE skeleton: 1) sequential construction of C/D/E ring system; 2) construction of two rings in one step. Some representative syntheses are shown below.

O A B C N O D H E HO OMe cephalotaxine (20) Figure 2.4 C/D/E Ring System of Cephalotaxine

2.3.1 Sequential Construction - Construction of D/E ring, then C ring

The most common way of synthesizing Cephalotaxus alkaloids is to build up the 1- azaspiro[4.4]nonane framework first and then close azepine ring to construct the polycyclic ring system. It is flexible to furnish 1-azaspiro[4.4]nonane after assembling with benzodioxole portion or prepare the 1-azaspiro[4.4]nonane building block separately. As shown in Scheme 2.1, Semmelhack and coworkers finished the total synthesis of cephalotaxine in 1972,22 the same year as Weinreb. Amino diester 23 went through a three step sequence involving acyloin-type ring closure, oxidation and methylation to afford functionalized 1-azaspiro[4.4]nonane 24. The alkylation between spirocycle 24 and iodosulfonate 25 proceed smoothly to give iodo ketone 26. Upon treatment with base in ammonia and irradiation under a mercury lamp, iodide 26 underwent a SRN1 reaction to build the benzazepine ring. Cephalotaxine (20) was obtained via DIBAL-H reduction of cephalotaxinone (28). 10

1) Na/K alloy, O O TMSCl, PhH HN ONs HN O I N 2) Br2, DCM 25 CO2Me O I 3) CH2N2, DCM iPrNEt, MeCN CO Me 2 45-55% (3 steps) O OMe 88% O 23 24 26 OMe

O O O N KOtBu, NH3 N DIBALH N O O O hv H 74% H 94% O O HO OMe OMe OMe 27 cephalotaxinone (28) cephalotaxine (20) Scheme 2.1 Semmelhack’s Total Synthesis of (±)-Cephalotaxine

The first asymmetric total synthesis of (–)-cephalotaxine was completed by Mori and 23 coworkers in 1995 as illustrated in Scheme 2.2. After LiAlH4 reduction of ester 29, the following Parikh-Doering oxidation delivered aldehyde 30. In the presence of

Me3SiSnBu3 and CsF, an intramolecular cyclization of the generated alkenyl anion of 31 on an aldehyde moiety furnished the 1-azaspiro[4.4]nonane skeleton. After subjection to polyphosphoric acid, an acid-catalyzed electrophilic cyclization was used to close the benzazepine ring, accomplishing the synthesis of cephalotaxine core.

MeO MeO N TMSSnBu3, 1) LiAlH4, THF N CsF, DMF MeO MeO C MeO O 2 2) SO3-pry., DMSO 85% 81% (2 steps) 29 30 H I I

O MeO MeO N PPA N N steps O HO H MeO 66% MeO H HO OMe 31 32 cephalotaxine (20) Scheme 2.2 Mori’s Total Synthesis of (–)-Cephalotaxine

The assembly of 1-azaspiro[4.4]nonane with benzodioxole portion was also used in of recent synthesis developed by Chandrasekhar and coworkers (Scheme 2.3). 24 An intramolecular Aldol reaction of 33 followed by ester hydrolysis gave spiro pyrrolidine 34. Treating with TFA removed the Boc-protection of 34 and the subsequent amidation occurred to synthesize the pentacyclic compound 35. 11

CO Et CO H 2 2 1) TFA, DCM O NaH, PhH O 2) EDC, HOBt amylalcohol BocN NBoc DCM, DIPEA O O O 56%(3 steps) O Me 33 34 O

O O O N N steps O O H HO OMe 35 O cephalotaxine (20) Scheme 2.3 Chandrasekhar’s Total Synthesis of (±)-Cephalotaxine

2.3.2 Sequential Construction-Construction of C/E ring, then D ring

Assembling cyclopentenone-fused benzazepine moiety of cephalotaxine before pyrrolidine is another kind of strategy. Due to the 5,7-membered fused ring system, this type of construction is rarely seen in synthesis of cephalotaxine. A strain-release rearrangement and a [4+2] cycloaddition are shown as representatives. Gin’s synthesis (Scheme 2.4) 25 involves a [3,3]-rearrangement of N-vinyl-2- arylaziridine 36 and following tautomerization to give benzazepine 37. Tertiary vinylogous amide 38 was synthesized from 37, using Me3SiCH2I to install the alkyl substituent. Acylation of carbonyl group and desilylation happened to form azomethine ylide 39, which spontaneously underwent the [3 + 2] cyclization with vinylsulfonate to generate the spiro-pyrrolidine 40.

O Cs CO O O N 2 3 NH TMSCH2I dioxane N O Cs2CO3, MeCN, O O TMS O 76% 75% O O O O O O O 36 37 38 O

O O O PivCl, AgOTf N N DCM N CH2 steps O O O then TBAT, SO2Ph H O O SO2Ph PivO HO HO 77% O O OMe 39 40 cephalotaxine (20) Scheme 2.4 Gin’s Total Synthesis of (–)-Cephalotaxine

In Fuchs’ synthesis, 26 the benzazepine and pyrrolidine precursor were introduced using a Diels–Alder reaction. Depicted in Scheme 2.5, treating hydroxamic acid 41 with tetra- n-butyl ammonium periodate delivered acylnitroso 42, which underwent an intramolecular Diels–Alder reaction to form two diastereomeric lactams 43a and 43b. Reductive cleavage of the N-O bond, mesylation and intramolecular nitrogen alkylation of the resultant mesylate generated the desired diastereomer 44, completing the construction of pentacyclic ring system.

O NHOH O N O O O O O N O nBu4NIO4 O O O DCM H +

O O O 41 O 42 O O 43a

O O O O O N O 1. Na(Hg), EtOH N N O 2. MsCl, TEA, DCM O steps O H H 3. NaH, THF H HO O 46% (3 steps) O O O OMe 43b 44 cephalotaxine (20) Scheme 2.5 Fuchs’ Total Synthesis of (±)-Cephalotaxine

2.3.3 Sequential Construction-Construction of C/D ring, then E ring

This strategy involves forging ring E ring after the pyrrolobenzazepine portion, which involves the assembly of a tertiary carbon center. Owing to the difficulty of making all three stereocenters in one ring, only one of the syntheses prepared cephalotaxine enantioselectively. As shown in Scheme 2.6, Weinreb and coworkers published the first total synthesis of cephalotaxine in 1972.21 Cyclization of aldehyde 45 followed by reduction gave enamine 46. Enamine 46 reacted with a mixed anhydride prepared from pyruvic acid and ethyl chloroformate to afford dicarbonyl 47 in 73% yield. In the presence of magnesium methoxide, an intramolecular Michael reaction proceed to give demethylcephalotaxinone 48, which contains the pentacyclic core of cephalotaxine. 13

O 1) BF3Et2O O O pyruvic acid, MeCN CHCl , 87% N 3 N ClCO2Et, NaHCO3 O O 2) LiAlH4, THF 73% O quant. 45 46

O O O N N N Mg(OMe)2 O MeOH O steps O H 58% O HO HO O OMe 47 O 48 cephalotaxine (20) Scheme 2.6 Weinreb’s Total Synthesis of (±)-Cephalotaxine

Another classic total synthesis of cephalotaxine was developed by Hanaoko group.27 Upon exposure to polyphosphoric acid, an intermolecular acylation of 49 formed the pyrrolobenzazepine 50. Treating 50 with 2,3-dichloropropene in the presence of sodium hydride in DMF generated enol ether 51. After heating at 150°C, a Claisen rearrangement occurred, and a subsequent reduction yielded alcohol 52. Treatment with sulfuric acid resulted in a cationic cyclization to give the tetracyclic ketone 53, which contains the cephalotaxine core.

O O O Cl MeO MeO MeO Cl N PPA N N MeO 74% MeO NaH, DMF MeO HO2C 91% O O 49 50 51 Cl O O O MeO MeO N 1) 150℃ N H2SO4 N steps O 2) NaBH MeO MeO H 4 69% H 97%(2 steps) HO HO OMe 52 Cl 53 O cephalotaxine (20) Scheme 2.7 Hanaoko’s Total Synthesis of (±)-Cephalotaxine

Li and coworkers reported a formal synthesis of cephalotaxine via enantioselective Tsuji allylation, shown in Scheme 2.8.28 Ketone 55 was introduced by formic acid promoted ring expansion of aldehyde 54, the following enolization of 55 generated carbonate 56. In the presence of Pd(dba)2 and a PHOX ligand, an intramolecular asymmetric allylic alkylation delivered 57 in 95% yield and 93% ee. Wacker oxidation of recrystallized olefin 57 followed by an aldol condensation afforded enone 59, which contains the pentacyclic skeleton of cephalotaxine. 14

O O O O O N HCO2H N LiHMDS, THF O N O O 79% O allyl chloroformate OHC O O 92% O 54 55 56 O

O O O Pd(dba) , O PdCl2, CuCl 2 N N ligand, THF O2, DMF/H2O MeOH, KOH O O 95% 76% 95% O O 57 58 O

O CH3 O O N N steps O O H O HO (p-CF -Ph) OMe 3 2P N O 59 cephalotaxine (20) PHOX ligand tBu Scheme 2.8 Li’s Formal Synthesis of (–)-Cephalotaxine

2.3.4 Construction of Two Rings in One Step

Some representatives about forging two rings of the C/D/E system in one step was shown. After introduction of all requisite carbon assemblage, transannular cyclization of large-sized rings and radical cyclization furnished the pentacyclic core of cephalotaxine. A quasi-biomimetic transannular cyclization was reported by Mariano and coworkers in 1994.29 Reductive amination of aldehyde 60 followed by base-mediated cyclization produced macrocyclic amino enone 61. α-Hydroxy enone 62 was thus obtained in three steps after debenzylation, Boc protection, and α-oxidation with assistance of (1R)-(−)- (10-Camphorsulfonyl)oxaziridine. Swern-Moffat oxidation generated endione 63. Deprotection of Boc and sequentical transannular cyclization of endione 63 constructed the pyrrolobenzazepin in one step. The resulted demethylcephalotaxinone 48 could be converted to cephalotaxine.

OMs 1) BnNH ·HCl, NaOAc 1) H2, Pd/C, iPrOH, 83% O 2 O Bn O Boc NaCNBH3, THF N 2) Boc2O, DCM,100% N O O O 2) iPr2NEt, MeCN 3) LDA, CSA-oxaziridine O 64%(2 steps) 78% O O O 60 61 62 OH

O Boc O O TFAA, NEt 3 N N N DMSO TMSOTf, DCM steps O O O 50%(2 steps) H O HO HO O 63 48 O OMe cephalotaxine (20) Scheme 2.9 Mariano’s Total Synthesis of (±)-Cephalotaxine

As illustrated in Scheme 2.10, the Li group 30 also used a transannular cyclization for constructing C/D ring system in one step. DIBAL-H reduction of dioxolanone 64 led to a tethered 1,2-oxidopentadienyl cation 65, which underwent an oxy-Nazarov cyclization to yield hydroxyl cyclopentenone 66. Acetylation followed by deprotection of Troc group led to a spontaneous transannular cyclization to form the pentacyclic compound 67.

Troc Troc O O N N Troc O N O O DIBALH O O 55% H HO O O O O O 64 65 66

O O N N 1) Ac2O, pyr., 85% steps O O H 2) Zn, NaH2PO4 H THF-H2O, 52% HO AcO OMe O 67 cephalotaxine (20) Scheme 2.10 Li’s Total Synthesis of (±)-Cephalotaxine

A cascade radical cyclization was performed by Ishibashi and coworkers to furnish the pyrrolobenzazepin portion of cephalotaxine (Scheme 2.11). 31 Condensation of amine 68 and cyclopentanone 69 followed by acylation yielded enamide 70. After treatment with Bu3SnH and 1,1′-azobiscyclohexanecarbonitrile, a 7-endo aryl radical cyclization of enamide 70 proceed to form amidoyl radical intermediate 71. The intermediate underwent a 5-endo-trig cyclization to generate the pentacyclic compound 72. 16

O O O Bu SnH, 1) Ti(OiPr) , THF N 3 O 4 O I ACN, PhCl + 2) CH =CH–COCl NH2 2 27% O I Et2NPh, DMAP TBDPSO OTBDPS 50% TBDPSO 68 69 70 OTBDPS

O O O O O N N N O O steps O H H H TBDPSO TBDPSO HO OTBDPS OTBDPS OMe 71 72 cephalotaxine (20) Scheme 2.11 Ishibashi’s Total Synthesis of (–)-Cephalotaxine

To date, most of the previous syntheses built up the C/D/E ring system of cephalotaxine sequentially, and few of them constructed two rings in one step. Unlike the previous syntheses, our plan is to develop a furan oxidation-transannular Mannich cyclization to forge the three rings at once, increasing the yield and reducing the number of steps.

15 Powell, R. G.; Weisleder, D.; Smith, C. R.; Rohwedder, W. K. Tetrahedron Lett. 1970, 11, 815–818. 16 Takano, I.; Yasuda, I.; Nishijima, M. J. Nat. Prod. 1996, 59, 1192–1195. 17 Alvandi, F.; Kwitkowski, V. E.; Ko, C.-W.; Rothmann, M. D.; Ricci, S.; Saber, H.; Ghosh, D.; Brown, J.; Pfeiler, E.; Chikhale, E.; Grillo, J.; Bullock, J.; Kane, R.; Kaminskas, E.; Farrell, A. T.; Pazdur, R. Oncologist 2014, 19, 94–99. 18 Voncken, J. W.; Kaartinen, V.; Pattengale, P. K.; Germeraad, W. T. V.; Groffen, J.; Heisterkamp, N. Blood 1995, 86, 4603–4611. 19 Jabbour, E.; Cortes, J. E.; Giles, F. J.; O’Brien, S.; Kantarjian, H. M. Cancer 2007, 109, 2171–2181. 20 Gandhi, V.; Plunkett, W.; Cortes, J. E. Clin. Cancer Res. 2014, 20, 1735–1740. 21 (a) Auerbach, J.; Weinreb, S. M. J. Am. Chem. Soc. 1972, 94, 7172–7173. (b) Weinreb, S. M.; Auerbach, J. J. Am. Chem. Soc. 1975, 97, 2503–2506. 22 a) Semmelhack, M. F.; Chong, B. P.; Jones, L. D. J. Am. Chem. Soc. 1972, 94, 8629- 8630; b) Semmelhack, M. F.; Chong, B. P.; Stauffer, R. D.; Rogerson, T. D.; Chong, A.; Jones, L. D. J. Am. Chem. Soc. 1975, 97, 2507–2516. 23 Isono, N.; Mori, M. J. Org. Chem. 1995, 60, 115–119. 24 Gouthami, P.; Chegondi, R.; Chandrasekhar, S. Org. Lett. 2016, 18, 2044−2046. 25 a) Eckelbarger, J. D.; Wilmot, J. T.; Gin, D. Y. J. Am. Chem. Soc. 2006, 128, 10370– 10371; b) Eckelbarger, J. D.; Wilmot, J. T.; Epperson, M. T.; Thakur, C. S.; Shum, D.; Antczak, C.; Tarassishin, L.; Djaballah, H.; Gin D. Y. Chem. Eur. J. 2008, 14, 4293– 4306. 26 (a) Burkholder, T. P.; Fuchs, P. L. J. Am. Chem. Soc. 1988, 110, 2341–2342; (b) Burkholder, T. P.; Fuchs, P. L. J. Am. Chem. Soc. 1990, 112, 9601–9613. 27 Yasuda, S.; Yamada, T.; Hanaoka, M. Tetrahedron Lett. 1986, 27, 2023–2026. 28 Zhang, Z.-W.; Wang, C.-C.; Xue, H.; Dong, Y.; Yang, J.-H.; Liu, S.; Liu, W.-Q.; Li, W.-D. Z. Org. Lett. 2018, 20, 1050−1053. 29 Lin, X.; Kavash, R. W.; Mariano, P. S. J. Am. Chem. Soc. 1994, 116, 9791–9762. 30 Li, W.-D.; Duo, W.-G.; Zhuang, C.-H. Org. Lett. 2011, 13, 3538–3541. 31 Taniguchi, T.; Ishibashi, H. Org. Lett. 2008, 10, 4129–4131. 18

Chapter 3 First Generation Route Toward (–)-Cephalotaxine and (–)-Homoharringtonine

3.1 Retrosynthetic Analysis of Cephalotaxine: First Generation Route

Our retrosynthetic analysis is illustrated in Scheme 3.1. Cephalotaxine is known to be synthesized from cephalotaxinone (20) via sodium borohydride reduction. Owing to the β-amino-carbonyl functionality of cephalotaxinone, we envisioned that it could arise from an iminium ion intermediate 73 via Mannich reaction. Intermediate 73 originated from unsaturated 1,4-diketone 74, which could be generated by oxidative opening of the corresponding macrocyclic furan 75.32 Strategically, aromaticity of the furan could mask the reactive diketone prior to the key Mannich cyclization, the ring strain and transannular interactions of the 12-membered macrocyle could facilitate the occurrence of dearomatization and subsequent Mannich reaction. The C8–N bond of macrocycle 75 could be furnished by either a substitution reaction or a reductive amination of an amine and an oxygenated carbon. For the diarylmethylene group in macrocycle 75, our initial plan was cleaving the C4–C13 bond, leading to a well-known phenethylamine 7733 and a furan 78 that could be prepared based on Wasserman’s strategy.34

O O O O N N N HN O O O O H H O O HO O O OMe OMe OMe MeO cephalotaxine (20) cephalotaxinone (28) 73 74

O O O 8 HN OH HN NH2 O O 13 4 Br O O Br O HO 77 MeO2C O

MeO MeO 75 76 MeO 78 Scheme 3.1 Retrosynthetic Analysis of Cephalotaxine

3.2 Preparation of Furan

As shown in Scheme 3.2, our initial synthetic route commenced with a three-step preparation of 3-methoxyfuran 78, which is an extension of Wasserman’s furan synthesis. An aldol reaction of phosphorane 79 followed by ozonolysis gave dihydrofuranone 85. Although presented as quantative yield in the crude NMR, the dihydrofuranone 85 decomposed slowly on silica gel chromatography, leading to relatively low isolated yield. Upon treatment with HCl in methanol, dihydrofuranone 85 underwent TBS deprotection, acid catalyzed dehydration and methylation to afford furan 78. To our dismay, furan 78 was isolated as a mixture with an unidentified compound. Various solvent systems and separation methods were evaluated; however, none of them gave an efficient separation. In an effort to mitigate the difficulty of separation, we decided to proceed with Dess–Martin oxidation of the mixture. Gratifyingly, the polarity of the compounds differs after oxidation, enabling the purification of aldehyde 86 and identification of the byproduct 87. It is worth noting that the triphenylphosphine oxide formed from ozonlysis complicates the chromatograph purification and byproducts generated from dehydration step could be removed in vacuo. We tried conducting the ozonlysis, dehydration and oxidation steps without purification, delivers aldehyde 86 in 31% yield. The result is comparable with performing the three step sequence with purified compounds.

Wasserman: O O O O O OH RCHO O TsOH OMe 3 OMe CO2Me PhH CO2Me LDA, THF PPh3 O O PPh3 R OH R OH R 79 80 R = Ph, alkyl 81 82

Our work:

OTBS O O O O O O 83 TBSO O3 TBSO OMe OMe LDA, THF DCM CO2Me PPh3 PPh3 67% OH O OH 79 84 85

OMe OMe MeO HCl HO DMP O O OMe + CO2Me MeOH CO2Me DCM CO2Me O O O OMe 78, 42% (NMR yield, CH2Br2 86, 31% (3 steps) 87 as internal standard) isolated spot, mixture no seperation by FCC, HPLC Scheme 3.2 Wasserman-based Furan Synthesis

An alternative way has also been investigated to avoid decompostion when purifying 35 dihydrofuranone 85 by chromatography (Scheme 3.3). TiCl4 promoted aldol reaction of aldehyde 83 and diazoacetoacete 88 afforded α-diazo-β-ketoester 89. Upon exposure to dimethyldioxirane (DMDO), the generated dihydrofuranone 85 could be simply purified by evaporation of the solvent.

O O O O O OTBS TBSO O TBSO DMDO OMe OMe acetone CO2Me TiCl4, Et3N N2 81%(NMR yield) O OH N2 CH2Cl2, 58% OH 88 89 85 Scheme 3.3 Alternative Way of Preparing the Furan

3.3 Synthetic Efforts toward Macrocycle

The first attempt to synthesize macrocycle was shown in Scheme 3.4. Reductive amination of aldehyde 86 and prepared phenethylamine 77 occurred to deliver secondary amine 90. However, conditions to give alcohol 76 could not be identified. The ester functionality of furan 90 was resistant to any type of hydride reduction or nucleophilic additions. We envisioned that the decreased reactivity originated from the lack of electrophilicity of the ester functionality, due to the lone pair electrons from the methoxy group and furanyl oxygen donating to the ester carbonyl.

O NH2 O O OMe O 77 HN HN O Br AcOH, MeOH; O [H] O Br Br O CO2Me MeO2C O HO O NaBH3CN 52% MeO 86 MeO 90 76 LiAlH4, LiBH4, BH3, RedAl, DIBAL, etc. Scheme 3.4 Efforts toward Reduction of the Ester

Similarly, furan 78 was resistant to reduction or nucleophilic addition of reagents derived from bromide 77 and we sought to convert the ester to another functionality. To our delight, when subjecting ester 78 to N,O-dimethylhydroxylamine hydrochloride and iPrMgCl, Weinreb amide 92 was formed in 74% yield. A DIBAL-H reduction gave furfural 93 in quantative yield, which could be further reduced to furfuryl alcohol 94 in 96% yield. Having several furans in hand, we became interested in exploring coupling 21

reactions between furans and bromide with its derivatives.

OMe OMe OMe HO imidazole TBSO CH3ONHCH3·HCl TBSO TBSCl, DMF iPrMgCl, DCM O CO2Me CO2Me O 77% O 74% O N OMe 78 91 92

OMe OMe DIBALH TBSO TBSO O LiAlH4, THF OH THF -20℃ to 0℃ quant. O H 96% O 93 94 Scheme 3.5 Conversion of Ester to Other Functionalities

A survey of the literature revealed that diarylmethanes (and congeners) can be constructed with various kinds of couplings. Some representative examples are shown in Scheme 3.6. Diarylmethanes have been prepared by a Nickel-catalyzed cross- electrophile coupling of benzylic pivalates and aryl halides, as shown by Jarvo and coworkers.36 Furyl-substituted pivalate 95 provided the coupled product 97 in high yield. MacMillan also reported a Csp3−Csp2 cross-coupling of oxalates and aryl 37 halides. In the presence of NiBr2·dtbbpy, Ir photocatalyst and CsHCO3, oxalate 98 was exposed to bromobenzoate 99 under blue LED, yielding diarylmethane 100 in 95%. Another method to generate diarylmethanes bearing furan was published by Barluenga and coworkers. 38 They developed a metal-free reductive coupling between tosylhydrazones and boronic acids.

Jarvo:

O OPiv I 15mol% NiBr2·glyme, 3eq Zn, O 15mol% dppf, DMA, rt + CH3 77% CH3 95 96 97 MacMillan: O Br 5mol% NiBr2·glyme, 5mol% dtbbpy, OH 1mol% Ir[dFppy] (dtbbpy)PF O + 2 6 O MeO2C CsHCO3, dioxane, 25℃, 14h CO2Me 98 99 blue LED, 95% 100

Barluenga:

B(OH)2 Et O NNHTs K2CO3, dioxane,110℃ + H Et 98% O 101 102 103 Scheme 3.6 Coupling Conditions for Preparing Diarylmethanes

Being aware of these modern conditions enable cross coupling reactions with alcohol 22

derivatives, we were able to perform the reactions by using the reagents in the literature. However, when applying the conditions to our own system, none of the conditions gave the desired result (Scheme 3.7). Either the decomposition of furan occurred or the the coupled product was not observed. Similarly, nucleophilic addition with carbonyl group at C2 of furan was also attempted. Grignard type addition is a typical way of making a C-C bond, but converting bromide 107 to Gringnard reagent proved to be difficult. Fortunately, the Weinreb ketone synthesis gave the desired coupled product. After deprotonation of Boc protected amine with KH and lithium-halogen exchange, the intermediate of bromide 107 was treated with Weinreb amide 92 to furnish the C- C bond, delivering ketone 108 in very low yield. Despite efforts to optimize the reaction conditions, the product yield could not be improved. Again, the obstacle is the modest electrophilicity of the carbonyl group.

O NHBoc O NHBoc FG2 MeO cross-coupling R O O O FG1 + OTBS MeO 104 105 O OTBS FG1 + FG2 = Br + CHO (Grignard-type Add'n) 106 = I + CH2OPiv (Jarvo’s condition) = B(OH)2 + CH=NHNTs (Barluenga’s condition) = Br + CH2-oxylate (MacMillan’s condition)

OMe O NHBoc N O KH, nBuLi NHBoc Me O THF O O MeO + O (very low yield O Br MeO OTBS ~5%) O OTBS 107 92 108 Scheme 3.7 Attempted Diarylmethane Synthesis

In conclusion, efforts to construct macrocycle via the connection between C13 and C4 was unsuccessful. We turned our attention toward a second generation synthesis of cephalotaxine using the Friedel–Crafts alkylation method to build the biarylmethylene bond.

3.4 Supplemental Data

General Experimental Details: All reactions were carried out under an inert Ar atmosphere in oven-dried glassware. Flash column chromatography (FCC) was carried out with SiliaFlash P60, 60 Å silica gel. Reactions and column chromatography were monitored with EMD silica gel 60 F254 plates and visualized with potassium permanganate, ceric ammonium molybdate, molybdate, ninhydrin, or iodine stains. Tetrahydrofuran (THF), methylene chloride

(CH2Cl2), and methanol (MeOH) were dried by passage through activated columns. Diisopropylamine was dried and distilled over calcium hydride. All other reagents and solvents were used without further purification from commercial sources. Unless otherwise noted, melting points were obtained from material that solidified after chromatography. Instrumentation: FT-IR spectra were obtained on NaCl plates with a PerkinElmer Spectrum Vision spectrometer. Proton and carbon NMR spectra (1H NMR and 13C

NMR) were recorded in deuterated chloroform (CDCl3) unless otherwise noted on a Bruker 700 MHz Avance III Spectrometer with carbon-optimized cryoprobe and Bruker 400 MHz DPX-400 spectrometer. Multiplicities are abbreviated as follows: s = singlet, d = doublet, t = triplet, q = quartet, sept = septet, br = broad, m = multiplet. Melting points were determined with a Cole-Parmer instrument and are uncorrected.

Methyl 8-((tert-butyldimethylsilyl)oxy)-5-hydroxy-3-oxo-2-(triphenyl-λ5- phosphaneylidene)octanoate (84). To a solution of phosphorane 79 (157.8 mg, 0.419 mmol, 1.0 eq) in THF (4.2 mL, 0.1 M) at -78 °C was added 1.0 M LDA in THF (0.46 mL, 1.1 eq), the solution was warmed up to -5 °C and stirred for 1h. The reaction mixture was cooled back to –78 °C, aldehyde 83 (100 mg, 0.482 mmol, 1.15 eq) was added, and the mixture was slowly warmed to –20 °C over 2.5 h. The mixture was quenched with saturated NH4Cl solution (5 mL) and extracted with EtOAc (5 mL×3).

The combined organic phases were washed with brine, dried over Na2SO4, filtered and 24

concentrated. Purification by FCC (1:3 hexanes:EtOAc) yielded 86 as a colorless oil (161.1 mg, 67%).

Data for 84: mp 111–113 °C; Rf 0.63 (1:3 hexanes:EtOAc); IR (thin film) 2949, 2932, -1 1 2858, 1674, 1436, 1105, 1090 cm ; H NMR (700 MHz, CDCl3) δ 7.66 – 7.63 (m, 6 H), 7.55–7.53 (m, 3 H), 7.48–7.45 (m, 6 H), 4.24 (br, 1 H), 3.95 (m, 1 H), 3.64 (m, 2 H), 3.23 (dd, J = 16.1, 2.8 Hz, 1 H), 3.17 (d, J = 4.2 Hz, 3 H), 2.87 (dd, J = 16.1, 9.1 Hz, 1 H), 1.70–1.66 (m, 1 H), 1.63–1.60 (m, 1 H), 1.55–1.51(m, 2 H), 0.88 (s, 9 H), 13 0.04 (s, 6H); C NMR (176 MHz, CDCl3) δ 133.2, 133.1, 131.9, 131.8, 128.7, 128.6, 126.4, 125.9, 69.0, 63.4, 49.7, 46.1, 46.1, 33.4, 29.1, 26.0, 18.4; HRMS (TOF MS ES+) cald for C33H44O5SiP [M+H]: 579.2696, found 579.2700.

Methyl 3-methoxy-5-(3-oxopropyl)furan-2-carboxylate (86). To a stirred solution of phosphorane (95.0 mg, 0.164 mmol, 1.0 eq) in DCM (4.10 mL, 0.04 M) at -78 °C was bubbled O3 until the solution became blue. The reaction mixture was quenched by

Me2S (0.1 mL) and then warmed up to rt. The solution was concentrated and used directly without further purification. To a solution of crude residue from above in MeOH (1.5 mL, 0.11 M) was added 4 M HCl in dioxane (0.041 mL, 0.164 mmol, 1.0 eq). The solution was heated at reflux for 16 h. After cooling to rt, the reaction mixture was concentrated and used directly without further purification. To a solution of crude mixture from above in DCM (3.51 mL, 0.05 M) at 0c was added Dess–Martin periodinane (148.5 mg, 0.350 mmol, 2.1 eq). The mixture was warmed to rt and kept stirring for 14 h, upon which time saturated Na2S2O3 solution (2 mL) and

NaHCO3 solution (2 mL) was added to quench the reaction. Once the mixture became clear, it was extracted with DCM (3 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (12:1 DCM:EtOAc) yielded 86 as a colorless oil (10.8 mg, 31%).

Data for 86: mp 89–90 °C; Rf 0.66 (1:1 DCM:EtOAc); IR (thin film) 2954, 1705, 1611, -1 1 1476, 1193, 1098 cm ; H NMR (700 MHz, CDCl3) δ 9.81 (s, 1H), 6.11 (s, 1 H), 3.89 (s, 3H), 3.87 (s, 3 H), 2.97 (t, J = 7 Hz, 2 H), 2.88 (t, J = 7 Hz, 2 H); 13C NMR (176 25

MHz, CDCl3) δ 199.9, 159.2, 157.7, 155.8, 127.0, 99.2, 58.9, 51.5, 41.2, 21.4; HRMS

(ESI) calcd for C10H13O5 [M+H]: 213.0763, found 213.0765.

Methyl 8-((tert-butyldimethylsilyl)oxy)-2-diazo-5-hydroxy-3-oxooctanoate (89). To a solution of methyl diazoacetate 88 (213.2 mg, 1.50 mmol, 1.0 eq) in DCM (10.0 mL, 0.15 M) at -78 °C was added TiCl4 (0.181 mL, 1.65 mmol, 1.1 eq) dropwise and

NEt3 (0.231 mL, 1.65 mmol, 1.1 eq), the solution was stirred for 1 h. Then a solution of aldehyde 83 (311.1 mg, 1.50 mmol, 1.0 eq) in DCM (1.0 mL, 1.5 M) was added slowly, and the mixture maintained at -78 °C for 4 h. The mixture was quenched with saturated NH4Cl solution (8 mL) and extracted with DCM (10 mL×3). The combined organic phases were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (3:1 hexanes:EtOAc) yielded 89 as a yellow oil (216.2 mg, 58%).

Data for 89: Rf 0.39 (2:1 hexanes:EtOAc); IR (thin film) 2955, 2930, 2858, 2137, 1726, -1 1 1438, 1310, 1098, 835 cm ; H NMR (700 MHz, CDCl3) δ 4.12 (m, 1 H), 3.84 (s, 3 H), 3.65 (m, 2 H), 3.30 (br, 1 H), 3.05 (dd, J = 17.0, 3.0 Hz, 1 H), 2.97 (dd, J = 17.0, 13 8.9 Hz, 1 H), 1.63 (m, 4 H), 0.89 (s, 9 H), 0.05 (s, 6 H); C NMR (176 MHz, CDCl3) δ 192.7, 161.7, 67.9, 63.2, 52.3, 46.9, 33.6, 28.9, 26.0, 25.9, 18.4, -5.3; HRMS (TOF

MS ES+) cald for C15H28N2O5SiNa [M+Na]: 367.1665, found 367.1648.

Methyl 5-(3-((2-(6-bromobenzo[d][1,3]dioxol-5-yl)ethyl)amino)propyl)-3- methoxyfuran-2-carboxylate (90). A solution of aldehyde 86 (50.0 mg, 0.236 mmol, 1.0 eq) and amine 77 (115.0 mg, 0.471 mmol, 2.0 eq) in MeOH (2.40 mL, 0.1 M) at 0C was added acetic acid (28.3mg, 0.471 mmol, 2.0 eq), and then stirred for 5 min. NaBH3CN was added to the solution and it was maintained at 0 °C for 16 h. The mixture was then quenched with 1 M NaOH solution and extracted with EtOAc (3 mL×3). The combined organic phases were dried over Na2SO4, filtered and concentrated. Purification by FCC (20:1 CH2Cl2:MeOH with 10% NH4OH) yielded 90 as a colorless oil (54.2 mg, 52%). 26

Data for 90: Rf 0.54 (10:1 CH2Cl2:MeOH with 10% NH4OH); IR (thin film) 2933, -1 1 1705, 1611, 1477, 1233, 1102 cm ; H NMR (700 MHz, CDCl3) δ 6.98 (s, 1 H), 6.75 (s, 1 H), 6.12 (s, 1 H), 5.95 (s, 2 H), 3.89 (s, 3 H), 3.86 (s, 3 H), 2.90 (m, 4 H), 2.76 (t, J = 7.1 Hz, 2 H), 2.71 (t, J = 7.5 Hz, 2 H), 1.95 (sept, J = 7.3 Hz, 2 H); 13C NMR (176

MHz, CDCl3) δ 159.3, 156.0, 147.5, 147.1, 131.3, 126.7, 114.5, 112.8, 110.4, 101.8, 98.7, 59.0, 51.4, 49.1, 48.4, 35.6, 26.9, 26.5. HRMS (TOF MS ES+) cald for

C19H23NO6Br [M+]: 440.0709, found 440.0707.

5-(3-((tert-butyldimethylsilyl)oxy)propyl)-N,3-dimethoxy-N-methylfuran-2- carboxamide (92). A solution of ester (400.0 mg, 1.22 mmol, 1.0 eq) and N,O- Dimethylhydroxylamine hydrochloride (237.6 mg, 2.44 mmol, 2.0 eq) in THF (17.4 mL, 0.07 M) was stirred at -20 °C for 5 min. Then 1.0 M iPrMgCl in THF (5.48 mL, 5.48 mmol, 4.5 eq) was added dropwise over 30 min, and the solution was kept stirring for additional 3 h. The reaction mixture was quenched with saturated NaHCO3 solution (10 mL) and extracted with EtOAc (15 mL×3).. The combined organic phases were dried over Na2SO4, filtered and concentrated. Purification by FCC (1:1 hexanes:EtOAc) yielded 92 as a colorless oil (216.2 mg, 58%).

Data for 92: Rf 0.32 (1:1 hexanes:EtOAc); IR (thin film) 2954, 1705, 1612, 1476, 1098 -1 1 cm ; H NMR (400 MHz, CDCl3) δ 6.06 (s, 1 H), 3.87 (s, 3 H), 3.79 (s, 3 H), 3.65 (t, J = 6.1 Hz, 2 H), 3.26 (s, 3 H), 2.71 (t, J = 7.6 Hz, 2 H), 1.87 (m, 2 H), 0.90 (s, 9 H), 13 0.05 (s, 6 H); C NMR (176 MHz, CDCl3) δ 161.1, 157.7, 154.8, 128.2, 98.3, 61.9,

61.7, 58.7, 30.5, 25.9, 25.3, -5.3; HRMS (TOF MS ES+) cald for C17H32NO5Si [M+H]: 358.2050, found 358.2057.

5-(3-((tert-butyldimethylsilyl)oxy)propyl)-3-methoxyfuran-2-carbaldehyde (93). To a solution of Weinreb amide 92 (200.0 mg, 0.559 mmol, 1.0 eq) in THF (11.2 mL, 0.05 M) at -78 °C was added 1.0 M DIBAL-H in THF (1.68 mL, 1.68 mmol, 3.0 eq), the solution was stirred for 3 h and kept stirring until clear after quenching with saturated Rochelle’s salt solution. The mixture was extracted with EtOAc and the 27

combined organic phases were washed with brine, dried over Na2SO4 and filtered. The solution was concentrated without further purification, yielding 93 as a colorless oil (167.9 mg, quant.).

Data for 93: Rf 0.62 (2:1 hexanes:EtOAc); IR (thin film) 2954, 2930, 2857, 1655, 1598, -1 1 1479, 1102 cm ; H NMR (700 MHz, CDCl3) δ 9.49 (s, 1 H), 6.07 (s, 1 H), 3.92 (s, 3 H), 3.65 (t, J = 6.1 Hz, 2 H), 2.74 (t, J = 7.5 Hz, 2 H), 1.89 (m, 2 H ) 0.90 (s, 9 H), 0.05 13 (s, 6 H); C NMR (176 MHz, CDCl3) δ 172.4, 164.1, 136.9, 98.4, 61.8, 58.9, 30.2, 25.9, 25.7, 18.3, -5.4.

28 29 30 31 32 33 34 35 36 37 38 39 40

32 (a) Piancatelli, G.; Scettri, A.; D'Auria, M. Tetrahedron 1980, 36, 661–663; (b) Kobayashi, Y.; Katsuno, H.; Sato, F. Chem. Lett. 1983, 1771–1774; (c) Lepage, L.; Lepage, Y. Synthesis 1983, 12, 1018–1019; (d) Gollnick, K.; Griesbeck, A. Tetrahedron 1985, 41, 2057–2068; (e) Dominguez, C.; Csaky, A. G.; Plumet, J. Tetrahedron Lett. 1990, 31, 7669–7670; (f) Adger, B. J.; Barrett, C.; Brennan, J.; McGuigan, P.; McKervey, M. A.; Tarbit, B. J. Chem. Soc., Chem. Commun. 1993, 15, 1220–1222; (g) Wahlen, J.; Moens, B.; De Vos, D. E.; Alsters, P. L.; Jacobs, P. A. Adv. Synth. Catal. 2004, 346, 333– 338; (h) Sayama, S. Heterocycles 2005, 65, 1347–1358; (i) Sayama, S. Synth. Commun. 2007, 37, 3067–3075. 33 Ito, K.; Tanaka, H.; Kayama, M. Chem. Pharm. Bull.1977, 25, 1249–1255. 34 Wasserman, H. H.; Lee, G. M. Tetrahedron Lett. 1994, 35, 9783–9786. 35 Abed, H. B.; Mammoliti, O.; Bande, O.; Lommen, G. V.; Herdewijn, P. J. Org. Chem., 2013, 78, 7845–7858. 36 Konev, M. O.; Hanna, L. E.; Jarvo, E. R. Angew. Chem. Int. Ed. 2016, 55, 6730– 6733. 37 Zhang, X.; MacMillan, D. W. C. J. Am. Chem. Soc. 2016, 138, 13862−13865. 38 Barluenga, J.; Tomás-Gamasa, M.; Aznar, F.; Valdés, C. Nat. Chem. 2009, 1, 494– 499. 41

Chapter 4 Total Synthesis of (–)-Cephalotaxine and (–)- Homoharringtonine: Second Generation Route

4.1 Retrosynthetic Analysis of Cephalotaxine: Second Generation Route

As shown in our retrosynthetic analysis (Scheme 4.1), application of a Friedel–Crafts alkylation was the focus in the second generation route toward cephalotaxine. Macrocycle 75 could be obtained by substitution reaction of mesylate 109. The formation of C3-C4 bond could take place via a Friedel–Crafts alkylation between benzyl chloride 110 and furan 111. This strategy would present C4 atom as an electrophilic benzyl chloride, a well-known electrophile. Moreover, the strategy would take advantage of the resonance polarization of the furan as a nucleophile thereby increasing the propensity of the Friedel–Crafts alkylation to occur.

O O OH HN NHTFA O OMs NHTFA O O 4 O O O O 3 Cl MeO MeO MeO 75 109 110 111 Scheme 4.1 Retrosynthetic Analysis of Cephalotaxine: Second Generation Route

4.2 Preparation of the Macrocycle

Our synthesis started with the preparation of furan 111 and benzyl chloride 110. As depicted in Scheme 4.2, a Claisen-type condensation of TBS-protected hydroxypentanone 112 gave γ-chloro β-diketone 113. The enolate formed from the methyl ketone could be quenched by the acidic proton of γ-chloro β-diketone before nucleophilic addition, resulting in the low yield in the first step. Upon exposure to DBU, diketone 113 underwent a base-mediated cyclization to generate furanone 114 in 57% yield over two steps.39 With the assistance of HMPA, methylation of furanone 114 delivered 3-methoxyfuran 111, which could be simply purified by filtration through a silica gel/Celite pad with hexanes. 42

Benzyl chloride 110 was prepared based on known procedure.40 Henry reaction of piperonal 115 followed by reduction and acidified with HCl to yield amine salt 116. After TFA protection, the acetimide was treated with dry HCl gas in formalin to afford benzyl chloride 110. The TFA protection is the only step that required chromatography, and the other compounds were purified by recrystallization.

O OTBS LDA, THF O O DBU, THF TBSO Me Cl O 57% (2 steps) Cl OEt 112 113 OTBS OTBS O NaH , HMPA O Me2SO4, THF 80% O MeO 114 111

O 1) NH4OAc, CH3NO2 1) NEt3, DMAP O NH2·HCl O NHTFA O H AcOH, 80% TFAA, DCM, 94% 2) formalin Cl O 2) LiAlH4, THF; O O HCl, 68% HCl gas, 69% 115 116 110 Scheme 4.2 Preparation of the Furan and Benzyl chloride

In an effort to realize a Friedel-Crafts alkylation between furan 111 and benzyl chloride 110, we tested conditions by screening various bases and solvents (Table 4.1). Both polar solvent and weak base was found to be essential for the alkylation. Gratifyingly, 2,6-lutidine and trifloroethanol were discovered to give 117 in 64% yield. Other common bases, such as pyridine and DMAP, were tested without an increase in yield. Assuming increasing electrophilicity of benzyl chloride could facilitate the F–C reaction, additives such as NaI (Entry 9) and AgNO3 (Entry 10) were also attempted, 117 was formed in decreased yield.

O NHTFA O OTBS O NHTFA base, + solvent O O Cl MeO MeO 111 O 110 OTBS 117

Entrya Base solvent additive Yield (%) 1 2,6-lutidine CF3CH2OH – 57 2 2,6-lutidine 90%MeCN/10%H2O – 31 3 2,6-lutidine (CF3)2CHOH – 20 4 NH4HCO3 90%MeCN/10%H2O – 23 5 NH4HCO3 CF3CH2OH – 14 6 imidazole CF3CH2OH – 39 7 pyridine CF3CH2OH – 50 8 DMAP CF3CH2OH – 51 9 2,6-lutidine CF3CH2OH NaI 48 10 2,6-lutidine CF3CH2OH AgNO3 20 11 2,6-lutidine – – 0 b 12 2,6-lutidine CF3CH2OH – 42 c 13 2,6-lutidine CF3CH2OH – 64 a Reaction condition: benzyl chloride 110 (0.0154 mmol, 1.0 eq), furan 111 (0.0185 mmol, 1.2 eq), base (0.0185 mmol, 1.2 eq), additive (0.0154 mmol, 1.0 eq), solvent (0.1M), rt; b furan 111 (1.5 eq) was used; c benzyl chloride 110 (811.1 mg, 2.62 mmol, 1.0 eq), furan 111 (850 mg, 3.14 mmol, 1.2 eq), 2,6-lutidine (0.364 mL, 3.14 mmol, 1.2 eq), CF3CH2OH (5.24 mL, 0.5 M). Table 4.1 Reaction Condition Screen for the Friedel–Crafts Alkylation

While attempting to remove TBS group, two typical reaction conditions were investigated. In the case where silyl ether 117 was reacted with tetra-n-butylammonium fluoride (TBAF), alcohol 118 was isolated in 80% yield. Treatment with 1M HCl resulted in the TBS deprotection as well as demethylation, delivering furanone 119 as the undesired product.

O NHTFA O NHTFA O TBAF, THF O (1) MeO O 80% MeO OTBS O OH

117 118

O NHTFA O NHTFA O 1M HCl, THF O (2) MeO 53% O O OTBS O OH 117 119 Scheme 4.3 Conditions for Removal of the TBS Group

The approach to macrocycle 75 is depicted in Scheme 4.4. Mesylation of alcohol 119 formed mesylate 109. Exposure of mesylate 109 to aqueous LiOH led to cyclization along with TFA removal to furnish macrocyclic furan 75. No dimer or oligomers were detected during the formation, and we were able to conduct the reaction on scales as large as 650 mg.

O NHTFA O NHTFA O MsCl, NEt LiOH HN O 3 O DCM H2O, THF O O MeO O 86% MeO O 62-85% OH OMs MeO 119 109 75 Scheme 4.4 Base-mediated Cyclization to Form the Macrocycle

4.3 Synthesis of Cephalotaxinone via Furan oxidation-Transannular Mannich Reaction

The key transformation of macrocycle 75 to cephalotaxinone (28) is shown in Table 4.2. Reaction solvent is crucial in oxidative furan openings, and the low yield can be attributed to addition of alcoholic solvent to the furan after oxidation. Based on Sayama’s research,41 sterically hindered solvents tend to give less solvolysis products, and 3-methoxy furan is known to undergo an oxidative opening in the presence of DDQ and tert-butanol. Tert-butanol (Entry 1) was first utilized as solvent in the oxidative opening/Mannich cascade reaction, to our delight, cephalotaxinone was generated, although in low yield (~7%). We suspected the solvolytic trapping of the charged reactive intermediate resulted in the low yield. The solvent screen was then limited to polar non-nucleophilic solvents. Pleasingly, trifluoroethanol at higher temperature (Entry 14) was discovered to enhance the yield to 60%. Other well-known oxidative furan opening reagents25 were tested, such as cerium ammonium nitrate (CAN), trimethylphenylammonium tribromide (PTAB), mCPBA, singlet oxygen etc., cephalotaxinone was either not observed or generated in low yield. 45

Sayama: OMe OMe O DDQ OMe Ph Ph + Ph tBuOH Ph Ph O O O Ph O 120 121, 95% 122, 0%

Our attempt: O O HN N O DDQ O O H solvent O MeO OMe 75 (±)-cephalotaxinone (28) Entry Solvent Yield(%) 1 tBuOH 7.0 2 DMSO 0.3 3 EtOH 8.0 4 iPrOH 0 5 MeOH 3.6 6 DCM 1.8 7 THF:H2O 1:1 0 8 THF:HMPA 1:1 0 9 (CF3)2CHOH 0 10 DMF 0 11 Ethylene glycol 6.7 12 CF3CH2OH 28 13b tBuOH 10 c 14 CF3CH2OH 60 a Reaction conditions: furan 75 (2.0 mg, 6.3×10-3 mmol, 1.0 eq), DDQ (1.4 mg, 6.3×10-3 mmol, 1.0 eq), solvent (0.1M), rt, HPLC yield; b furan 75 (28.0 mg, 0.0888 mmol, 1.0 eq), DDQ (20.2 mg, 0.0888 mmol, 1.0 eq), solvent (0.1M), rt, isolated yield; c furan 75 (8.2 mg, 0.026 mmol, 1.0 eq), DDQ (5.9 mg, 0.026 mmol, 1.0 eq), solvent (0.1M), 50°C, isolated yield. Table 4.2 Reaction Conditions Screened for Synthesizing Cephlotaxinone

Interestingly, when treating macrocycle 75 in a few oxidation conditions, the existence of byproducts 123 and 124 attracted our attention. Containing similar skeleton of cephalotaxine, the ester derivatives of these two compounds might be sufficient to bind with peptidyl transferase center of the ribosome. As such they may be great candidates for biological testing. Mechanistically, instead of forging a pyrrolidine, an azepane ring was constructed through an iminium ion formation to form 125. Enamine 123 was generated through tautamerization, and amine 124 was obtained from completing the Mannich reaction. Attempts to increase the yield of enamine 123 and amine 124 by 46

switching oxidation conditions and using alternative methods of preparation was unsuccessful.

O O O O HN N N N DDQ O O H O O O tBuOH MeO H O MeO O OMe MeO O 75 cephalotaxinone (28) 123 124 ~10% H+ Mannich Mannich

O HN O O O N + N O O O O O MeO MeO O 74 73 OMe 125 Scheme 4.5 Proposed Mechanism to Enamine 123 and Amine 124

As the proposed mechanism shows in Scheme 4.5, a single electron oxidation of furan 75 gives a radical cation 125, which is subsequently captured by the tethered amine to furnish 126. After the second single electron oxidation, the formed oxocarbenium ion 127 is fragmented to give iminium ion 73, which underwent Mannich cyclization to form cephalotaxinone as a single diasteromer.

O O O DDQ, HN N HN –H+ O CF3CH2OH O O O O O 60%

MeO MeO OMe 75 125 126

O O O N –e– N N O O O O H O O OMe OMe OMe 127 73 (±)-cephalotaxinone Scheme 4.6 Racemic Synthesis of Cephalotaxinone

We hypothesized that the key transformation could be rendered enantioselective. Which mechanistic step that would to be controlled is unknown. It is possible that the final Mannich step is the enantiodeterming step. Alternatively, if a conformation of intermediate 73 is stable on the timescale of the Mannich step, then the undecatrienone ring could have conformational chirality. Therefore, the initial transannular addition to 47

the radical cation 125 could be enantiodetermining. In such a scenario, good transfer of point-to-conformational chirality, and non-reversibility of the mechanistic steps is required. In either case, the enantiodeterming step is an addition to a cationic species. We decided to add a chiral counter anion to replace dichlorodicyanohydroquinone anion in the cascade reaction. Several chiral acids have been tested, four examples are shown in Scheme 4.6. Unfortunately, none of the reactions presented any enantioenrichment of cephalotaxinone. Note that conducting the reaction in a non-polar solvent (i.e. toluene) to promote lose contact of the chiral ion pair leads to very slow reaction rates and no enantioselectivity.

O O HN DDQ, N O chiral acid O O H CF CH OH 3 2 O MeO OMe 75 (–)-cephalotaxinone

iPr iPr

CH H CO CF H3C 3 3 3 O iPr OH OH O P OH O OH O iPr HO3S O (R)-(+)-MTPA (+)-CSA iPr iPr (R)-BINOL (R)-TRIP Scheme 4.7 Attempts to Synthesize (–)-Cephalotaxinone with Chiral Acids

4.4 Total Synthesis of (–)-Cephalotaxine and (–)-Homoharringtonine

In the final stages of the synthesis of (–)-cephalotaxine, we focused our attention on a kinetic resolution strategy as shown in Scheme 4.8. Gratifyingly, Noyori asymmetric 42 hydrogenation gave (–)-cephalotaxine (20) in 50% yield (krel=278) with 97% ee and (+)-cephalotaxinone (28) in 49% yield with 97% ee. To our best knowledge, this is the first example of preparing (–)-cephalotaxine using a kinetic resolution strategy. Taking advantage of the three-step sequence developed by Gin and coworkers, the ester side chain can be appended to (–)-cephalotaxine to form (–)-homoharringtonine.

iPr Ts 1) Et3N, DMAP, 128 Ph N O Cl O Ru N N Cl Ph H O Me 4 mol% Cl Cl H (–)-20 (–)-HHT 50% OH NEt3: formic acid = 5:1 BnO O 97% ee O OMe krel = 278 O (> 99% ee) + 128 (±)-cephalotaxinone (28) O (+)-28 2) MeONa, MeOH 49% 3) H2, Pd/C 97% ee Scheme 4.8 Kinetic Resolution of Cephalotaxinone

With the undesired enantiomer of cephalotaxinone in hand, racemization was attempted to recycle the compound. Mori observed racemization of cephalotaxinone during the methylation of demethylcephalotaxinone 48.23 Stoltz applied the same condition to α-hydroxy enone 130a and 130b, two diastereomers were formed in a 2:1 ratio. 43 They speculated the stereochemical interconversion resulted from an equilibrium through achiral cyclopentadienone 129. We believed that the same condition could be applied in the racemization of (+)-cephalotaxinone. To our delight, treatment of (+)-28 to dimethoxypropane and TsOH in dioxane led to racemization, enabling the recycling of the undesired enantiomer.

Mori:

O O O N MeO OMe N HN O O O H TsOH, dioxane, HO reflux, 76% O O O OMe OMe 48 (enantioenriched) 28 (racemic) 129 Stoltz: AcO AcO AcO AcO O O MeO OMe O O N N N or N O O O + O TsOH, dioxane, H H 90℃ HO HO O O OMe 2:1 dr 130a O 130b O 131a 131b OMe

Our work:

O O N MeO OMe N , TsOH, O O H dioxane, reflux H O 72h, 99% O OMe OMe (+)-cephalotaxinone (±)-cephalotaxinone

Scheme 4.9 Racemization of (+)-Cephalotaxinone

In conclusion, we completed the total synthesis of (–)-cephalotaxine and (–)- homoharringtonine. The key steps include: 1) a furan oxidation-transannular Mannich reaction; 2) kinetic resolution of racemic cephalotaxine; 3) racemization of undesired enantiomer of cephalotaxinone. The longest linear sequence of synthesizing (–)- cephalotaxine is 9 steps and the overall yield is more than 5% chemical yield. As shown in Table 4.3, the yield of previous total synthesis of cephalotaxine and their number of longest linear sequence is summarized (formal synthesis not included). It is notable that the average LLS of total synthesis is more than 10 steps, and the yield of enantioselective synthesis is around 2%.44,24 Compared to previous syntheses, our route is one the most efficient syntheses. Yield Type Main Author Year Starting Material LLS (%) Cl O HN Weinreb 1972 8 8 O O HO

O Semmelhack 1972 N 10 12 H MeO Cl

Hanaoka 1986 O N CO2Et 7 12 MeO H O O CO2Me Kuehne 1988 O NH2 40 11

SO2Ph O O Fuchs 1988 O O 17 15 O O O Racemic

O NO2 Ikeda 1990 21 18 O O Mariano 1994 OH 12 13 O I O Nagasaka 1997 CO2Me 12 14

Chandrasekhar 2016 O 10 11 N Boc

Bn O N Fan 2017 O 7 12 O OAc 50

H O N Mori 1995 O 2 18 H tBu α-Naph Me MeO O Royer 2004 OMe 9 16 NH2 Enantioselective O Gin 2006 O 2 17 O HO TsO Ishibashi 2008 2 14 O O

Table 4.3 Previous Total Syntheses of Cephalotaxine

4.5 Future plan

Many sideRESEARCH-chain analoguesARTICLE of homoharringtonine are known and have been evaluated. However, the cephalotaxine portion has not been thoroughly studied: only around 10 S. cerevisiae G2403 E. coli derivativesa can be found in the literature. Asb the Figure 4.1 shows,G2816 lycorine, narcilasineab H. sapiens 5′ 3′ 5′ 3′ eS30 uS12 and homoharringtonine have the similar amine-substitutedMg2+ benzodioxole portion, and C G C G 25S rRNA A1754 G C G they positon differently when binding within the A site of peptidylA2820 transferase centre. C U2873 U U U U G1645 C G C G A1755 It suggests that the amine-substituted benzodioxole portion might play a specific role Geneticin 1408 A A 1645 G A in the recognition of the ribosomal binding site. Our efficient synthesis enables the A A C G 1491 C A 1754 optimization of cephalotaxine part, especially the amine-substituted benzodioxole G2874 A U U A A2821 C G A U A1756 T-2 toxin portion. The investigation of preparing potential pharmaceutical homoharringtonine3 5 3 5 P-site A-site Verrucarin A ′ ′ ′ ′ derivatives is under way in the Beaudry group. U2875 Deoxynivalenol c d Paromomycin OH O c G2403 dO G2403 Neomycin N OG2816 S. cerevisiae Ring III NH 5 G2816 O S. cerevisiae T. thermophilus ′′ Lycorine Mg2+ O OH E. coli 2+ 25S rRNA Narciclasine Mg HO G1645 Homoharringtonine OH Ring I OH A2820 A1408 A2820 U2873 OH U2873 Lycorine Narciclasine O N O 6′ C1646 A1754 H C1409 G1491 G2874 O A2821 G2874 OMe HO Nagilactone C e f A2821 O Edeine (bacteria) 18S rRNA HO CO2Me 18S rRNA U2875 25S rRNA U2875 Edeine (yeast) homoharringtonine Figure 4.1 Binding Site of Narcilasine, Lycorine and HomoharringtonineT-2 toxin efT-2 toxin S. cerevisiae Phenylalanine E. coli uS11 Cryptopleurine mRNA Pactamycin U2873 P-site codon U2504 P-site tRNA E-site codon mRNA E-site codon

A-site tRNA Figure 5 | Structures of 40S mRNA and tRNA inhibitors. a,The C2821 aminoglycosides binding site is different in eukaryotes. Secondary structure A2397 C2452 diagrams of helix 44 from bacteria (16S rRNA, left) and from yeast and human C2055 (18S rRNA, right). b, Geneticin (blue) binds to helix 44 and induces the flipping Figure 4 | Structures of 60S peptidyl transferase centre inhibitors. out of A1755 and A1756 (orange). c, The conformation of G1645 in yeast 18S a, Blasticidin S binds to the 60S tRNA P-site, whereas T-2 toxin, rRNA restricts the binding of aminoglycosides bearing a 69 amino group in ring deoxynivalenol, verrucarin A, lycorine, narciclasine, homoharringtonine, I, as shown for neomycin (pink, PDB accession code: 3QAN). d,The nagilactone C and anisomycin are clustered in the 60S tRNA A-site. b, Binding conformation of A1754 in yeast restricts the binding of aminoglycosides site of T-2 toxin (green), deoxynivalenol (orange) and verrucarin A (cyan). bearing a 599 hydroxyl group in ring III, as shown for paromomycin (blue, PDB c, Binding site of narcilasine (purple), lycorine (blue), homoharringtonine (pale 3UZ3). e, Edeine adopts a different conformation on the bacterial (blue, PDB cyan). The arrows highlight the location of the dioxol-pyrroline group. 1I95) and eukaryotic ribosome (yellow). f, Pactamycin (green) and d, Binding site of nagilactone C (blue). e, The conformation of U2504 in cryptopleurine (purple) share the same binding site in the 40S E-site. Natural bacteria (PDB accession codes: 2AVY, 2AW7) closes the binding site to mutations in the C-terminal part of protein uS11 (red) confer resistance to the eukaryotic-specific A-site inhibitors. f, A-site inhibitors hinder cryptopleurine43. The mRNA structure was taken from PDB accession code aminoacyl-tRNA positioning in the peptidyl transferase centre. The Phe–tRNA 4KZZ. structures were taken from PDB accession codes 2WDG, 2WDL. biosynthetic pathway. Nagilactone C shares all the features of A-site mRNA codons positioned in the A-site25,26. In bacteria, aminoglycosides inhibitors (Fig. 4d). Finally, homoharringtonine and anisomycin binding antibiotics alter translation accuracy and inhibit tRNA translocation by is conserved between archaea and eukaryotes; however, some variations perturbing the conformation ofthe decodingcentre nucleotides.Besides were found in the anisomycin vicinity (Extended Data Fig. 8b). their potent activity against Gram-negative bacteria, the aminoglyco- Consistent with structural data obtained from the archaeal 50S sub- side-induced suppression of premature termination holds potential for unit, the identity of 25S rRNA residue 2397 (2055) is suggested to in- the treatment of inherited disorders caused by nonsense mutations11,27. fluence the conformation of U2873 (U2504) that dictates the binding of The canonical aminoglycoside binding site is located within the either bacterial-specific or eukaryotic-specific inhibitors (Fig. 4e)14.In internal loop of helix 44 of 18S rRNA, which is part of the decoding bacteria, the residue 2397 (2055) is a cytosine, whereas an adenine is centre that contains the essential and universally conserved nucleotides found in 96% of eukaryotes. Of therapeutic interest, the remaining 4% A1755 (A1492) and A1756 (A1493). In close vicinity, two nucleotides of eukaryotes might be sensitive to antibacterial drugs instead of eukar- differ between bacteria and eukaryotes, but are identical in yeast and yotic inhibitors, such as Giardia species24. Most A-site inhibitors were humans: G1645 (A1408) and A1754 (G1491) (Fig. 5a)28,29. found to impair peptide bond formation during translation elongation The large class of aminoglycosides can be divided in three subgroups (Extended Data Table 1). Superimposition of aminoacyl-tRNA struc- according to their chemical structures: kanamycins, neomycins and tures shows that the entrance of the amino acid moiety in the peptidyl gentamicins30. We chose geneticin (G418) as a representative of kana- transferase centre is hindered by the presence of A-site inhibitors (Fig. 4f). mycins with high affinity for the eukaryotic ribosome. Geneticin binds into the aminoglycoside pocket and induces the flipping out of A1755 The decoding centre and A1756 (Fig. 5b). The structure highlights direct interactions The decoding centre of the ribosome forms a geometrically restricted between geneticin ring I and the eukaryote-specific residues G1645 pocket that accurately selects aminoacyl-tRNA in accordance with and A1754.

4.6 Supplemental Data

(CH2Cl2), and methanol (MeOH) were dried by passage through activated columns.

Triethylamine (NEt3) and 1,4-dioxane were dried and distilled over calcium hydride. All other reagents and solvents were used without further purification from commercial sources. Unless otherwise noted, melting points were obtained from material that solidified after chromatography. Instrumentation: FT-IR spectra were obtained on NaCl plates with a PerkinElmer Spectrum Vision spectrometer. Proton and carbon NMR spectra (1H NMR and 13C

NMR) were recorded in deuterated chloroform (CDCl3) unless otherwise noted on a Bruker 700 MHz Avance III Spectrometer with carbon-optimized cryoprobe and Bruker 400 MHz DPX-400 spectrometer. Multiplicities are abbreviated as follows: s = singlet, d = doublet, t = triplet, q = quartet, sept = septet, br = broad, m = multiplet. Enantiomeric excesses (ee values) were determined by HPLC analysis using a Shimadzu LC-2030 HPLC. Melting points were determined with a Cole-Parmer instrument and are uncorrected.

5-(3-((tert-Butyldimethylsilyl)oxy)propyl)furan-3(2H)-one (114). To a solution of diisopropylamine (3.67 mL, 26.2 mmol, 2.2 eq) in THF (23.8 mL, 0.50 M) at –78 °C was added n-butylithium (10.8 mL, 23.8 mmol, 2.0 eq) dropwise. The solution was stirred for 10 min, warmed to –30 °C and stirred for an additional 10 min. Then 5-((tert- butyldimethylsilyl)oxy)pentan-2-one45 (5.15 g, 23.8 mmol, 2.0 eq) was added dropwise, and the temperature was maintained for 30 min. The reaction mixture was cooled to – 50 °C, ethyl chloroacetate (1.27 mL, 11.9 mmol, 1.0 eq) was added, and the mixture 52

was slowly warmed to –30 °C over 5 h. The mixture was acidified to pH 4–5 with 2 M hydrochloric acid solution and extracted with Et2O (20 mL×3). The combined organic phases were washed with brine, dried over Na2SO4 and filtered. The solution was concentrated and used directly without further purification. To a stirred solution of the crude residue from above in THF (24.3 ml, 0.5 M) at rt was added 1,8-diazabicyclo[5.4.0]undec-7-ene (1.81 mL, 12.1 mmol, 1.02 eq) dropwise. Precipitation was observed after 5 min. After stirring for 1 h, the mixture was quenched by addition of saturated ammonium chloride solution and extracted with Et2O (30 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (4:1 hexanes:EtOAc) yielded 114 as a colorless oil (1.72 g, 57%).

Data for 114: Rf 0.40 (3:1 hexanes:EtOAc); IR (thin film) 2955, 2930, 2886, 2858, -1 1 1704, 1599, 1103 cm ; H NMR (700 MHz, CDCl3) δ 5.48 (s, 1 H), 4.49 (s, 2 H), 3.67 (t, J = 5.6 Hz, 2 H), 2.60 (t, J = 7.7 Hz, 2 H), 1.85 (quintet, J = 7.0 Hz, 2 H), 0.89 (s, 9 13 H), 0.05 (s, 6 H); C NMR (176 MHz, CDCl3) δ 201.9, 194.4, 103.2, 74.4, 60.9, 28.2,

26.5, 25.0, 17.4, –6.2; HRMS (ESI) calcd for C13H25O3Si [M+H]: 257.1573, found 257.1568. tert-Butyl(3-(4-methoxyfuran-2-yl)propoxy)dimethylsilane (111). To a slurry of sodium hydride (74.9 mg, 60 % dispersion in mineral oil, 1.87 mmol, 3.0 eq) in THF (1.46 mL) at 0 °C was added a solution of furanone 114 (160 mg, 0.624 mmol, 1.0 eq) in THF (0.62 mL) and HMPA (0.32 mL, 1.87 mmol, 3.0 eq). The mixture was stirred at 0 °C for 30 min, upon which time dimethyl sulfate (0.072 mL, 0.749 mmol, 1.2 eq) was added. The reaction mixture was stirred for additional 1.5 h. Concentrated ammonium hydroxide (1.0 mL) was slowly added to the solution and it was stirred for

30 min. The reaction mixture was extracted with Et2O (4 mL×3). The combined organic phases were washed with brine, dried over Na2SO4 and filtered. The concentrated mixture was filtered through a pad of 1:1 Celite:silica gel, washed with 20 mL hexane, the light yellow fraction was collected and concentrated to yield furan 111 as a light yellow oil (135 mg, 80%). 53

Data for 111: Rf 0.68 (4:1 hexanes:EtOAc); IR (thin film) 2954, 2931, 2898, 2858, -1 1 1622, 1102 cm ; H NMR (700 MHz, CDCl3) δ 6.95 (s, 1 H), 5.85 (s, 1 H), 3.68 (s, 3 H), 3.63 (t, J = 6.2 Hz, 2 H), 2.60 (t, J = 7.5 Hz, 2 H), 1.83-1.79 (m, 2 H), 0.90 (s, 9 H), 13 0.05 (s, 6 H); C NMR (176 MHz, CDCl3) δ 155.6, 150.8, 121.2, 99.5, 62.3, 58.0,

301.0, 26.1, 25.0, 18.5, –5.2; HRMS (ESI) calcd for C14H27O3Si [M+H]: 271.1729, found 271.1728.

N-(2-(6-((5-(3-((tert-Butyldimethylsilyl)oxy)propyl)-3-methoxyfuran-2- yl)methyl)benzo[d] [1,3]dioxol-5-yl)ethyl)-2,2,2-trifluoroacetamide (117). To a solution of furan 111 (850 mg, 3.14 mmol, 1.2 eq) and 2,6-lutidine (0.364 mL, 3.14 mmol, 1.2 eq) in 2,2,2-trifluoroethanol (5.24 mL, 0.5 M) at 0 °C was added benzyl chloride 110 (811 mg, 2.62 mmol, 1.0 eq). The mixture was stirred at rt for 5 h. Water (10 mL) was added and the aqueous layer was extracted with ethyl acetate (5 mL×3).

The combined organic layers were dried over Na2SO4, filtered and concentrated. Purification by FCC (4:1 hexanes:EtOAc) yielded 117 as a yellow film (962 mg, 68%).

Data for 117: Rf 0.55 (4:1 hexanes:EtOAc); IR (thin film) 3316, 2954, 2929, 2857, -1 1 1709, 1646, 1487, 1207, 1163 cm ; H NMR(700 MHz, CDCl3) δ 6.74 (s, 1 H), 6.60 (s, 1 H), 6.52 (br, 1 H), 5.91 (s, 3 H), 3.79 (s, 2 H), 3.71 (s, 3 H), 3.60 (t, J = 6.3 Hz, 2 H), 3.42 (q, J = 6.3 Hz, 2 H), 2.88 (t, J = 7.0 Hz, 2 H), 2.55 (t, J = 7.7 Hz, 2 H), 1.77 13 (quintet, J = 7.0 Hz, 2 H), 0.88 (s, 9 H), 0.02 (s, 6 H); C NMR(176 MHz, CDCl3) δ (157.5, 157.3, 157.1, 156.9), 146.7, 146.6, 143.6, 135.4, 130.5, 143.6, 135.4, 130.5, 128.8, (118.4, 116.8, 115.2, 113.5), 110.7, 109.7, 101.1, 98.5, 62.3, 59.6, 40.6, 31.9,

31.0, 28.7, 26.1, 25.1, 18.5, –5.2; HRMS (ESI) calcd for C26H37F3NO6Si [M+H]: 544.2342, found 544.2330.

2,2,2-Trifluoro-N-(2-(6-((5-(3-hydroxypropyl)-3-methoxyfuran-2- yl)methyl)benzo[d][1,3] dioxol-5-yl)ethyl)acetamide (118). To a solution of silyl ether 117 (500 mg, 0.920 mmol, 1.0 eq) in THF (9.20 mL, 0.1 M) at 0 °C was added 1M tetrabutylammonium floride solution (2.76 ml, 3.0 eq). The mixture was warmed 54

to room temperature and stirred for 5 h. The reaction was quenched with water (10 mL) and extracted with ethyl acetate (20 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (1:1

CH2Cl2:EtOAc) afforded 118 as a white solid (318 mg, 80%).

Data for 118: mp 72–73 °C; Rf 0.26 (1:1 hexanes:EtOAc); IR (thin film) 3330, 2955, -1 1 2931, 2897, 2859, 1715, 1580, 1488, 1209, 1162cm ; H NMR (700 MHz, CDCl3) δ 6.73 (s, 1 H), 6.71 (br, 1 H), 6.60 (s, 1 H), 5.94 (s, 1 H), 5.91 (s, 2 H), 3.79 (s, 2 H), 3.71 (s, 3 H), 3.64 (t, J = 6.3 Hz, 2 H), 3.42 (q, J = 6.3 Hz, 2 H), 2.88 (t, J = 7.0 Hz, 2 13 H), 2.60 (t, J = 7.0 Hz, 2 H), 1.82 (sept, J = 6.3 Hz, 2 H); C NMR (176 MHz, CDCl3) δ (157.6, 157.4, 157.2, 157.0), 152.9, 146.7, 146.6, 143.6, 135.7, 130.4, 128.8, (118.5, 116.8, 115.2, 113.6), 110.7, 109.7, 101.1, 98.8, 62.1, 59.5, 40.6, 31.9, 31.0, 28.7, 25.0;

HRMS (ESI) calcd for C20H23F3NO6 [M+H]: 430.1477, found 430.1474.

3-(4-Methoxy-5-((6-(2-(2,2,2-trifluoroacetamido)ethyl)benzo[d][1,3]dioxol-5- yl)methyl) furan-2-yl)propyl methanesulfonate (109). To a solution of alcohol 119

(468 mg, 1.09 mmol, 1.0 eq) in CH2Cl2 (5.45 mL, 0.2 M) at 0 °C was added NEt3 (0.380 mL, 2.72 mmol, 2.5 eq). The reaction mixture was stirred for 10 min, upon which time methanesulfonyl chloride (0.169 mL, 2.18 mmol, 2.0 eq) was added dropwise. The solution was slowly warmed to rt and stirred for 1 h. Water (5 mL) was added and the mixture was extracted with CH2Cl2 (10 mL×3). The combined organic layers were washed with brine, filtered, and dried over Na2SO4. Purification by FCC (1:1 hexanes:EtOAc) gave mesylate 109 as a pale yellow solid (477 mg, 86%).

Data for 109: mp 78–79 °C; Rf 0.41 (1:1 hexanes:EtOAc); IR (thin film) 3336, 2919, -1 1 2850, 1717, 1487, 1173 cm ; H NMR (700 MHz, CDCl3) δ 6.75 (br, 1 H), 6.70 (s, 1 H), 6.60 (s, 1 H), 5.98 (s, 1 H), 5.89 (s, 2 H), 4.20 (t, J = 6.3 Hz, 2 H), 3.79 (s, 2 H), 3.70 (s, 3 H), 3.42 (q, J = 6.3 Hz, 2 H), 2.96 (s, 3 H), 2.88 (t, J = 7.7 Hz, 2 H), 2.63 (t, 13 J = 7.0 Hz, 2 H), 2.00 (quintet, J = 7.0 Hz, 2 H); C NMR (176 MHz, CDCl3) δ (157.6, 157.3, 157.1, 156.9), 151.2, 146.7, 146.6, 143.7, 136.2, 130.4, 128.8, (118.5, 116.8, 115.2, 113.6), 110.5, 109.7, 101.1, 99.5, 69.0, 59.5, 40.7, 37.4, 31.9, 28.5, 27.7, 24.5;

HRMS (ESI) calcd for C21H25F3NO8S [M+H]: 508.1253, found 508.1242. 55

33-Methoxy-7-aza-1(5,6)-benzo[d][1,3]dioxola-3(2,5)-furanacyclononaphane (75). To a solution of mesylate 109 (21.0 mg, 0.0414 mmol, 1.0 eq) in THF (0.83 mL, 0.05 M) was added lithium hydroxide (3.0 mg, 0.12 mmol, 3.0 eq) and water (0.04 mL, 1.0 M). The mixture was heated to reflux and stirred for 36 h. The reaction was cooled to rt, diluted with water (3 mL) and extracted with ethyl acetate (4 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated.

Purification by FCC (10:1 CH2Cl2:MeOH with 10% NH4OH) yielded macrocycle 75 as a pale yellow film (12.1 mg, 85%).

Data for 75: Rf 0.26 (10:1 CH2Cl2:MeOH with 10% NH4OH); IR (thin film) 3327, 2936, -1 1 1641, 1503, 1485 cm ; H NMR (400 MHz, CDCl3) δ 6.73 (s, 1 H), 6.69 (s, 1 H), 5.89 (s, 3 H), 3.84 (s, 2 H), 3.69 (s, 3 H), 2.82 (t, J = 5.6 Hz, 4 H), 2.65 (t, J = 7.0 Hz, 2 H), 13 2.54 (t, J = 5.6 Hz, 2 H), 1.78-1.72 (m, 2H); C NMR (176 MHz, CDCl3) δ 152.5, 146.6, 145.8, 143.3, 135.6, 132.7, 130.2, 110.8, 109.5, 101.0, 99.3, 59.4, 49.0, 47.5,

31.8, 30.3, 28.2, 25.9; HRMS (ESI) calcd for C18H22NO4 [M+H]: 316.1549, found 316.1548. Large scale synthesis of 75: A solution of mesylate 109 (650 mg, 1.28 mmol, 1.0 eq), LiOH (184 mg, 7.69 mmol, 6.0 eq) in water (0.26 mL) and THF (25.6 mL, 0.05M) was refluxed for 60 h to give macrocycle 75 (250 mg, 62%).

(±)-Cephalotaxinone (28). To a solution of furan 75 (50.0 mg, 0.159 mmol, 1.0 eq) in 2,2,2-trifloroethanol (1.59 mL, 0.1 M) was added DDQ (36.0 mg, 0.159 mmol, 1.0 eq). The mixture was warmed to 50 °C and stirred for 15 h. The reaction mixture was quenched with saturated sodium bisulfite and sodium bicarbonate solution (1:1, 5 mL) and then extracted with ethyl acetate (4 mL×5). The combined organic phases were dried over Na2SO4, filtered, and evaporated. Purification by FCC (10:1 CH2Cl2:MeOH with 10% NH4OH) and recrystallization from EtOAc gave (±)-28 as a white solid (29.7 mg, 60%). 56

Data for 28: mp 181–183 °C (lit 178–180 °C); Rf 0.74 (10:1 CH2Cl2:MeOH with 10% -1 1 NH4OH); IR (thin film) 2925, 1724, 1625, 1487, 1230, 1103 cm ; H NMR (700 MHz,

CDCl3) δ 6.70 (s, 1 H), 6.64 (s, 1 H), 6.40 (s, 1 H), 5.92 (d, J = 1.4 Hz, 1 H), 5.91 (d, J = 1.4 Hz, 1 H), 3.80 (s, 1 H), 3.53 (s, 1 H), 3.11-3.08 (m, 1 H), 2.93-2.89 (m, 1 H), 2.71-2.67 (m, 1 H), 2.54-2.51 (m, 1 H), 2.45-2.43 (m, 2 H), 2.11-2.08 (m, 1 H), 1.98- 13 1.95 (m, 1 H), 1.89-1.85 (m, 2 H); C NMR (176 MHz, CDCl3) δ 201.0, 158.3, 147.4, 146.4, 130.7, 128.6, 123.9, 112.6, 110.4, 101.2, 65.5, 60.8, 57.4, 53.0, 47.8, 39.0, 31.5,

20.1; HRMS (ESI) calcd for C18H20NO4 [M+H]: 314.1392, found 314.1379. Reaction with tert-Butanol: To a solution of furan 75 (28.0 mg, 0.0888 mmol, 1.0 eq) in tert-Butanol (0.89 mL, 0.1 M) was added DDQ (20.2 mg, 0.0888 mmol, 1.0 eq). The mixture stirred for 4 d. The reaction mixture was quenched with saturated sodium bisulfite and sodium bicarbonate solution (1:1, 1 mL) and then extracted with ethyl acetate (2 mL×3). The combined organic phases were dried over Na2SO4, filtered, and evaporated. Purification by FCC (40:1 CH2Cl2:MeOH with 10% NH4OH) gave 123 as a colorless film (2.6 mg, 9%), 28 (2.8 mg, 10%) as a colorless flim and 124 (2.7 mg, 10%). 1 Data for 123: Rf 0.82 (10:1 CH2Cl2:MeOH with 10% NH4OH); H NMR (700 MHz,

CDCl3) δ 6.59 (s, 1 H), 6.53 (s, 1 H), 5.89 (s, 2 H), 5.65 (s, 1 H), 5.32 (s, 1 H), 3.78 (s, 3 H), 3.46 (m, 2 H), 3.41 (m, 2 H), 3.06 (m, 2 H), 2.72 (m, 2 H), 1.88 (m, 2 H); 13 C NMR (176 MHz, CDCl3) δ 199.2, 167.9, 146.1, 145.0, 141.0, 132.2, 130.1, 109.6, 109.1, 108.4, 103.8, 100.8, 56.5, 56.0, 53.0, 52.1, 38.1, 37.7, 29.7, 25.0; HRMS (ESI) calcd for C18H20NO4 [M+H]: 314.1392, found 314.1385. 1 Data for 124: Rf 0.46 (10:1 CH2Cl2:MeOH with 10% NH4OH); H NMR (700 MHz,

CDCl3) δ 6.85 (s, 1 H), 6.58 (s, 1 H), 5.94 (s, 1 H), 5.89 (s, 1 H), 5.26 (s, 1 H), 4.00 (s, 3 H), 3.26 (d, J = 15 Hz, 2 H), 3.02 (m, 4H), 2.95 (m, 1 H), 2.90 (m, 1 H), 2.83 (d, J = 14.5 Hz, 1 H), 2.74 (d, J = 11.1 Hz, 1 H), 2.50 (d, J = 12.7 Hz, 1H) 2.34 (d, J = 13.7 13 Hz, 1 H), 2.16 (m, 1 H); C NMR (176 MHz, CDCl3) δ 200.5, 191.4, 146.9, 146.0, 132.7, 129.1, 112.1, 109.0, 100.8, 100.4, 75.2, 63.6, 59.2, 56.1, 53.2, 37.6, 33.9, 29.7,

25.2; HRMS (ESI) calcd for C18H20NO4 [M+H]: 314.1392, found 314.1392.

(–)-Cephalotaxine (20) and (+)-Cephalotaxinone (28). A thick walled reaction tube containing racemic 28 (13.0 mg, 0.0415 mmol, 1.0 eq) was evacuated and backfilled with argon. Then a solution of freshly prepared Ru(p-cymene)-(S,S)-TsDPEN 42 catalyst (1.0 mg, 4 mol%) in a 5:1 NEt3:formic acid mixture (0.166 mL, 0.25 M) was added. The reaction mixture was sealed, heated to 40 °C and stirred for 48 h. The mixture was cooled to room temperature, diluted with saturated sodium bicarbonate solution (2 mL) and extracted with ethyl acetate (4 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered, and evaporated.

Purification by FCC (50:1 to 10:1 CH2Cl2:MeOH with 10% NH4OH) gave (–)-20 and (+)-28, the latter as a colorless film (6.4 mg, 49%, 97% ee). Recrystallization of (–)-20 from diethyl ether gave a white solid. (6.5 mg, 50%, 97% ee).

46 . . Calculation of krel = = = 278 [ ] [. . ]

47 Data for 20: mp 115–117 °C (lit 118–120 °C); Rf 0.34 (15:1 CH2Cl2:MeOH with 10% -1 1 NH4OH); IR (thin film) 3376, 2933, 2883, 1651, 1503, 1487, 1224 cm ; H NMR (700

MHz, CDCl3) δ 6.68 (s, 1 H), 6.65 (s, 1 H), 5.91 (d, J= 2.5 Hz, 2 H), 4.93 (s, 1 H), 4.77 (dd, J = 9.4, 3.4 Hz, 1 H), 3.73 (s, 3 H), 3.68 (d, J = 9.4 Hz, 1 H), 3.37-3.32 (m, 1 H), 3.08 (td, J = 9.0, 4.3 Hz, 1 H), 2.92 (td, J = 11.6, 7.2 Hz, 1 H), 2.61-2.56 (m, 2 H), 2.36 (dd, J = 14.5, 7.0 Hz, 1 H), 2.04-1.99 (m, 1 H), 1.89-1.85 (m, 1 H), 1.77-1.71 (m, 2 H), 13 1.61 (d, J = 3.4 Hz, 1 H); C NMR (176 MHz, CDCl3) δ 160.7, 147.1, 146.3, 134.5, 128.2, 112.8, 110.5, 101.1, 97.9, 73.5, 70.7, 58.3, 57.3, 54.1, 48.7, 43.9, 31.9, 20.6; HRMS (ESI) calcd for C18H22NO4 [M+H]: 316.1549, found 316.1551; [α] = –172.1

(c 0.21, CHCl3); HPLC (DAICEL, Chiralcel OD-RH, 30:70 H2O:MeOH with 2%

NH4OH, 0.5 mL/min, 25 °C): tR[min] = 12.44 (98.67%), 14.42 (1.33%), 97% ee.

O N O H O BnO OMe O O

O S1 58

(1S,3aR,14bS)-2-Methoxy-1,5,6,8,9,14b-hexahydro-4H-[1,3]dioxolo[4',5':4,5] benzo[1,2-d]cyclopenta[b]pyrrolo[1,2-a]azepin-1-yl (R)-2-((E)-4-(benzyloxy)-4- methylpent-2-en-1-yl)-4-oxooxetane-2-carboxylate (S1). To a solution of β-lactone 25 128 (10.0 mg, 0.033 mmol, 2.0 eq) and NEt3 (0.015 mL, 0.11 mmol, 6.6 eq) in CH2Cl2 (0.12 mL) was added 2,4,6-trichlorobenzoyl chloride (5.7 µL, 0.036 mmol, 2.2 eq). The resulting solution was stirred for 1 h and then transferred via syringe to a solution of cephalotaxine (20) (5.2 mg, 0.017 mmol, 1.0 eq) and N,N-dimethylaminopyridine (2.2 mg, 0.018 mmol, 1.1 eq) in CH2Cl2 (0.12 mL). This solution was then stirred for 3 h and directly loaded onto a pH 7.0 buffered silica gel25 column. Purification by FCC

(30:1 CH2Cl2:MeOH with 10% NH4OH) afforded S1 as a light yellow oil (8.5 mg, 84%).

Data for S2: Rf 0.37 (20:1 CH2Cl2:MeOH with 10% NH4OH); IR (neat film) 2931, -1 1 2807, 1841, 1748, 1655, 1507, 1224 cm ; H NMR (700 MHz, CDCl3) δ 7.32-7.24 (m, 5 H), 6.60 (s, 1H), 6.59 (s, 1H), 5.86 (d, J = 12.2 Hz, 3 H), 5.71 (d, J=16.1 Hz, 1 H), 5.49-5.45 (m, 1 H), 5.08 (s, 1H), 4.30 (s, 2H), 3.81 (d, J = 9.7 Hz, 1 H), 3.69 (s, 3 H), 3.10-3.06 (m, 2 H), 2.98 (d, J = 16.5 Hz, 1 H), 2.94-2.92 (m, 1 H), 2.63 (d, J = 16.7 Hz, 1 H), 2.60-2.56 (m, 3 H), 2.41 (dd, J = 14.4, 7.1 Hz, 1 H), 2.35 (dd, J = 14.4, 6.7 Hz, 1 H), 2.06-2.01 (m, 1 H), 1.91-1.88 (m, 1 H), 1.77-1.73 (m, 2 H), 1.32 (s, 6 H); 13C

NMR (176 MHz, CDCl3) δ 168.1, 165.5, 156.6, 147.1, 146.0, 143.1, 139.7, 133.6, 128.4, 127.90, 127.6, 127.4, 127.3, 120.5, 113.4, 109.9, 101.3, 101.2, 75.9, 75.6, 75.3, 70.8, 65.1, 57.4, 56.5, 54.1, 48.7, 46.2, 43.6, 37.7, 31.6, 26.5, 26.3, 20.5; HRMS (ESI) calcd for C35H40NO8 [M+H]: 602.2754, found 602.2742.

O N O H O BnO OMe O HO O S2 OMe 1-((1S,3aR,14bS)-2-Methoxy-1,5,6,8,9,14b-hexahydro-4H-[1,3]dioxolo[4',5':4,5] benzo[1,2-d]cyclopenta[b]pyrrolo[1,2-a]azepin-1-yl) 4-methyl (R)-2-((E)-4- (benzyloxy)-4-methylpent-2-en-1-yl)-2-hydroxysuccinate (S2). To a solution of β- lactone S1 (5.6 mg, 0.0092 mmol, 1.0 eq) in MeOH (0.092 mL, 0.1 M) was added a 59

freshly prepared solution of 0.5 M NaOMe in MeOH (0.020 mL, 0.011 mmol, 1.1 eq). After 10 min the solution was quenched with saturated ammonia chloride solution (2 mL) diluted with CH2Cl2 (2 mL). The phases were separated and the aqueous phase was extracted with CH2Cl2 (3 mL×3). The combined organic phases were dried over

Na2SO4, filtered and concentrated. The crude ester S2 was used directly in next step without further purification.

Data for S2: Rf 0.63 (15:1 CH2Cl2:MeOH with 10% NH4OH); IR (neat film) 3520, -1 1 2927, 1745, 1655, 1504, 1487, 1364 cm ; H NMR (700 MHz, CDCl3) δ 7.34-7.23 (m, 5 H), 6.61 (s, 1 H), 6.55 (s, 1 H), 5.94 (d, J = 9.7 Hz, 1 H), 5.85 (d, J = 13.5 Hz, 2 H), 5.61 (d, J = 15.8 Hz, 1 H), 5.51-5.47 (m, 1 H), 5.05 (s, 1 H), 4.32 (s, 2 H), 3.77 (d, J= 9.7 Hz, 1 H), 3.67 (s, 3 H), 3.58 (s, 3 H), 3.47 (s, 1 H), 3.14-3.11 (m, 2 H), 3.97-2.92 (m, 1 H), 2.59 (m, 2 H), 2.39 (dd, J =12.6, 5.2 Hz, 1 H), 2.29 (d, J = 16.3 Hz, 1 H), 2.23-2.16 (m, 2 H), 2.04-2.01 (m, 1 H), 1.96 (d, J = 16.1 Hz, 1 H), 1.90 (m, 1 H), 1.76- 13 1.74 (m, 2 H), 1.33 (d, J = 4.9 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 173.5, 170.4, 146.8, 145.9, 141.0, 139.7, 128.3, 127.5, 127.1, 122.9, 112.9, 109.7, 100.9, 100.2, 75.3, 75.25, 74.61, 65.15, 57.49, 56.16, 54.15, 51.80, 48.81, 43.56, 41.92, 41.77, 31.63, 26.69,

26.52, 20.50; HRMS (ESI) calcd for C36H44NO9 [M+H]: 634.3016, found 634.3007.

(–)-Homoharringtonine (21). To a solution of crude allylic benzyl ether S2 (7.0 mg, 0.011 mmol, 1.0 eq) in MeOH (0.11 mL) was added 10% Pd/C (1.4 mg, 20% by wt).

A H2 balloon was applied to the vessel and the suspension was stirred at rt for 48 h.

Glacial acetic acid (11 µL) was added via syringe and the solution was stirred under H2 for 22 h. Further 10% Pd/C (0.7 mg, 10% by wt) and glacial acetic acid (11 µL) were added and the suspension was stirred under H2 for 48 h. The mixture was filtered through a plug of Celite and the solvent was evaporated. Purification by FCC (20:1

CH2Cl2:MeOH with 10% NH4OH) afforded (–)-21 as a colorless film (2.4 mg, 40% over 2 steps).

Data for (–)-21: Rf 0.38 (10:1 CH2Cl2:MeOH with 10% NH4OH); IR (neat film) 2920, -1 1 2852, 1746, 1654, 1505, 1488, 1366, 1225 cm ; H NMR (700 MHz, CDCl3) δ 6.62 (s, 1 H), 6.54 (s, 1 H), 6.00 (d, J = 10.1 Hz, 1 H), 5.87 (d, J = 2.3 Hz, 2 H), 5.05 (s, 1 H), 60

3.78 (d, J = 9.8 Hz, 1 H), 3.68 (s, 3 H), 3.57 (s, 3 H), 3.53 (s, 1 H), 3.13-3.08 (m, 2 H), 2.96-2.93 (m, 1 H), 2.60-2.57 (m, 2 H), 2.38 (dd, J = 14.4, 6.8 1 H), 2.26 (d, J = 16.5 Hz, 1 H), 2.06-2.01 (m, 1 H), 1.91 (d, J =16.3 Hz, 1 H), 1.91-1.89 (m, 1 H), 1.77-1.74 (m, 2 H), 1.42-1.38 (m, 5 H), 1.27 (s, 1 H), 1.19 (d, J = 3.4 Hz, 6H); 13C NMR (176

MHz, CDCl3) δ 174.1, 170.6, 157.8, 146.8, 146.0, 133.5, 128.5, 112.8, 128.5, 112.8, 109.8, 101.0, 100.4, 74.9, 74.8, 71.1, 70.7, 57.6, 56.0, 54.1, 51.7, 48.8, 43.9, 43.5, 42.7,

39.3, 31.5, 29.5, 29.1, 20.4, 18.0; HRMS (ESI) calcd for C29H40NO9 (M+H) 545.2703 found 545.2703; [α] = –110.5 (c 0.24, CHCl3).

(±)-Cephalotaxinone (28). To a solution of (+)-28 (1.0 mg, 0.0032 mmol, 99% ee, 1.0 eq) in dioxane (0.32 mL, 0.01 M) and 2,2-dimethoxypropane (0.32 mL, 0.01 M) in thick walled reaction tube was added TsOH·H2O (2.4 mg, 0.013 mmol, 4.0 eq). The suspension was heated to reflux and kept stirring for 3 d. The reaction mixture was cooled to rt, quenched with saturated sodium bicarbonate solution (2 mL) and extracted with ethyl acetate (3 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (50:1

CH2Cl2:MeOH with 10% NH4OH) afforded (±)-28 as a pale yellow film (1.0 mg, 100%, 5.0% ee). Larger scale racemization of 28: Using the same procedure above, (+)-2 (8.5 mg, 0.027 mmol, 97% ee) afforded (±)-28 (6.8 mg, 80%, 22% ee) after refluxing for 82 h.

61 Depicted 1H and 13C Spectra 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94

39 Winkler, J. D.; Oh, K.; Asselin, S. M. Org. Lett. 2005, 7, 387–389. 40 Onozaki, Y.; Kurono, N.; Senboku, H.; Tokuda, M.; Orito, K. J. Org. Chem. 2009, 74, 5486–5495. 41 (a) Sayama, S. Heterocycles 2005, 65, 1347–1358; (b) Sayama, S. Synth. Commun. 2007, 37, 3067–3075. 42 Haack, K. J.; Hashiguchi, S.; Fujii, A.; Ikariya, T.; Noyori, R. Angew. Chem. Int. Ed. Engl. 1997, 36, 285 – 288; Angew. Chem. 1997, 109, 297–300. 43 Liu, Q.; Ferreira, E. M.; Stoltz, B. M. J. Org. Chem. 2007, 72, 7352–7358. 44 a) Abdelkafi, H.; Nay, B. Nat. Prod. Rep. 2012, 29, 845–869; b) Ma, X.-Y.; An, X.- T.; Zhao, X.-H.; Du, J.-Y.; Deng, Y.-H.; Zhang, X.-Z.; Fan, C.-A. Org. Lett. 2017, 19, 2965−2968. 45 Liu, L.; Floreancig, P. E. Org. Lett. 2009, 11, 3152−3155. 46 Jacobsen, E. N. Acc. Chem. Res. 2000, 33, 421–431. 47 Nagasaka, T.; Sato, H.; Saeki, S.-i. Tetrahedron: Asymmetry 1997, 8, 191–194. 95

Chapter 5 Introduction and Synthetic Studies toward Bazzanin K

5.1 Conformational Chirality

Molecular chirality plays a critical role in chemistry, biology, and medicine. 48 Identification of chirality in molecules which possess sp3-hybridized stereogenic carbon atoms is straightforward. However, for molecules that show conformational chirality by virtue of restricted rotations of a sigma bond, the determination is not- trivial, particularly when attempting to distinguish between achiral and racemic compounds. Prediction and determination of conformational chirality in compounds is often difficult, and the ideal method of verifying conformational chirality is still total synthesis. Exploration of conformational chirality is a long term goal in the Beaudry group, we have already synthesized and studied lots of compounds which might possess conformational chirality. Diarylether heptanoid (DAEH) natural products were synthesized to determine the chirality, and their racemization barriers were predicted by computational studies.49 A chiral cyclophane, cavicularin was synthesized by an enantioselective Diels–Alder reaction.12 Free energy of activation for racemization and racemization half-life was mersured after finishing the total synthesis of Arundo donax alkaloids, such as arundamine.14 Isolated as optically active, russuphelol (133) was determined to be achiral and optically inactive after its synthesis.50 We speculated that the optical activity might be caused by impurities in its isolation.

HO HO O Cl MeNH OMe Cl O Cl O O N HO NMe2 MeO OH Cl O Cl OH OH MeO OH N H Cl

juglanin A (132) HO cavicularin (10) arundamine (15) MeO russuphelol (133) - unknown chiral properties -believed to be chiral -believed to be achiral -believed to be chiral -no [α] reported -optical active in nature -optical inactive in nature -optical active, [α]D =-3.2 -chiral, stable to 200℃ -chiral, stable to 180℃ -chiral, slow racemization at 25℃ -achiral Figure 5.1 Recent Conformational Chirality Researches in Beaudry Group

5.2 Introduction of Macrocyclic Bisbibenzyls

Macrocyclic bis(bibenzyls) (MBBs) are phenolic natural products isolated from liverwort species, and over 70 biological active MBBs have been found so far.51 Biosynthetically, all MBBs originate from dimerization of lunularin to form perrottetin E or isoperrottetin A as shown in Figure 5.2.52 The following esterification and biaryl coupling affords four different kinds of MMBs. Depending on connectivity of four aromatic rings, the MBBs can be categorized in three types: “C,O-type”, “C,C-type” and “O,O-type”.53

A O OH A O OH A O OH

D D D B B B o/p o/p OH OH HO HO HO HO C HO C O C A D perrottetin E C,O-type O,O-type +

A OH A OH A OH OH OH OH HO B HO C

B D B D B D lunularin o/p HO HO O o/p HO C HO C HO C isoperrottetin A C,C-type C,O-type Esterification and biaryl coulping in biosynthesis Figure 5.2 Three Types of MBBs

Further modification in biosynthesis, such as oxidative coupling and methylation, leads to new MBBs structures. Asterelin A was isolated from liverwort Asterella angusta and exhibits antifungal activities with MIC value of 128 µg/ml. 54 It possesses a dibenzofuran moiety in macrocyclic ring system due to a biphenyl and diaryl ether linkage between the B and C ring. Cavicularin is a cyclophane natural product isolated from Cavicularia densa and displays conformational chirality. The dihydrophenanthrene unit was formed by an extra biphenyl bond between the C and D rings, leading to significant strain to the molecule. Its A ring is distorted into a boat- like configuration, 15º out of planarity. Bazzanin K has a phenathrene moiety due to an additional biaryl linkage between C and D rings and unsaturation of dimethylene bridge.55 It also contains aryl chlorides which makes its synthesis more challenging. 97

Owing to the interesting structure and possibility of conformational chirality, our group started the investigation of this natural product.

OH O OH HO HO Cl O A A A D D D

C C BC Cl HO B B O HO OMe HO OH asterelin A (134) cavicularin (10) bazzanin K (135) Figure 5.3 Structure of Asterelin A, Cavicularin and Bazzanin K

5.3 Introduction of Bazzanin K

Bazzanin K (135) is a chlorinated “C,C-type” bisbibenzyl whose racemization barrier remains to be determined. It was isolated from the liverwort Bazzania trilobata by 20 Berker and coworkers in 1997. The reported optical rotation [α]D is +180º, indicating bazzanin K exists in an enantioenriched form.55 However, demonstrated by our group previously, optical rotation is not always a reliable method for determination of conformational chirality due to the possible existence of impurities in an isolation sample.47 The diastereotopic protons at two methylene position suggest a slow interconversion of the rigidified macrocycle on the NMR time scale. The structure of bazzanin K is similar to cavicularin. Moreover, the A ring of bazzanin K is a meta- substituted aromatic ring with respect to the alkyl tether and the biaryl bond, leading to a twelve membered macrocycle core which is two carbons smaller than that of cavicularin. We believe bazzanin K displays a strained molecular architecture and shows conformational chirality. To test this hypothesis, we are currently working towards the total synthesis of bazzanin K.

Cl OH HO Cl HO HO A D O HO α’ α C OMe HO OMe Cl B OMe OH Cl bazzanin K (135) HO cavicularin (10) Figure 5.4 Structure of Bazzanin K and Cavicularin 98

5.4 Retrosynthetic Analysis The retrosynthetic analysis of bazzanin K inspired by our group’s synthesis of cavicularin12 (Scheme 5.1). Bazzanin K (135) could be forged by an intramolecular Diels–Alder reaction of pyrandione 137. The A ring could be constructed by elimination of phenylsulfinic acid of oxabicyclo[2.2.2]octene intermediate 136 followed by a retro-Diels–Alder reaction. We envisoned that the regioselectivity would result from interactions between the the nucleophilic C2 position of the pyrone motif and electrophilic C3 position of the vinyl sulfone moiety. It is worth noting that utilizing 5-hydroxypyrone would set phenolic A ring as well as making an enantioselective variant possible with the use of the cinchona alkaloids catalysts. Pyrandione 137 could be furnished by Stille coupling between pyrone-stannane 138 and triflate 139. The substituted phenanthrene moiety could be accessed by ring closing metathesis of terphenyl 140. The terphenyl architecture suggested a Suzuki coupling, and our first object was development of a one-pot, three-component, double Suzuki reaction between boronic esters 141, 142 and dibromostyrene 143.

OH O HO Cl 1 2 O 6 1 OH OiPr Cl 6 O iPr 5 O O Cl A D 5 5 3 2 6 D 3 2 4 PhO2S 4 3 PhO2S 1 D 4 O C Cl C C Cl Cl B OMe B OMe OH B OMe OiPr OiPr bazzanin K (135) 136 137

Br Cl Cl O D D C O + iPrO O Br C C 143 Cl OTf O OMe OMe SnBu3 OMe O OMOM OiPr OiPr Cl Bpin D PhO2S 138 B B B O Cl Cl OiPr Bpin 139 140 141 142 Scheme 5.1 Retrosynthetic Analysis of Bazzanin K

In order to achieve the desired connectivity, several issues need to be overcome in the double Suzuki reaction. According to our group’s previous study, dibromostyrene 143 can react regioselectively in Suzuki coupling reactions: bromides with proximal vinyl 99

groups undergo oxidative addition at reduced rates.56 However, it is still unknown whether the regioselectivity could be controlled when applied to more complex boronic esters. Additionally, the presence of different halogens in the Suzuki coupling requires chemoselectivity for bromine over chlorine to give the desired biphenyl. Also, the ortho,ortho-disubstituted aryl boronic ester is sterically hindered, which may reduce the reaction rate of transmetallation during Suzuki coupling.

5.5 Synthesis of Boronic Esters

The strategy starts with the preparation of two boronic esters 141 and 142. The synthetic route was developed my previous colleagues. Due to the steric hindrance of ortho,ortho-disubstituted position, conversion from bromide to boronic ester was challenging in original route. After optimization of the conditions, both boronic esters were prepared in good yield. My colleague Patrick Salvo developed the synthetic route to boronic ester 141 as shown in Scheme 5.2. Starting from commercially available 3-hydroxybenzaldehyde, a regioselective bromination 57 followed by isopropyl protection gave bromobenzaldehyde 145. A regioselective chlorination was performed para to the bulky isopropoxy group. A Wittig olefination of 146 afforded corresponding styrene 147. A lithium-halogen exchange with subsequent trap with iPrOBPin delivered the aryl boronic ester 141. The original step was conducted in the presence of nBuLi, unfortunately, the yield is only 38%. After switching the alkylithium reagent to tBuLi and lower the temperature to -98°C, the yield was increased to 60% and the reaction is repeatable.

1. Br , Fe, NaOAc Br OH 2 NCS, pTSA O HOAc, 23% OiPr DMF O 2. iPrBr, K2CO3 64% 99% 144 145

Br Br Bpin NaH, THF tBuLi, THF OiPr OiPr Ph PCH Br iPrOBpin OiPr O 3 3 96% 60% Cl Cl Cl 146 147 141 Scheme 5.2 Preparation of Boronic Ester 141

100

My colleagues Dr. Frank Dyer and Marshall Allen developed a synthetic route to aryl boronic ester 142. A regioselective chlorination of vanillin was performed when treating with sodium hydride and N-chlorosuccinimide. In the presence of aluminum trichloride, demethylation happened to afford catechol 149 in 73% over two steps. A regioselective bromination followed by direct protection of catechol generated benzodioxole 150. A Wittig olefination delivered the styrene 151. Finally, a lithium halogen exchange and treatment with iPrOBPin generated the aryl boronic ester 142 in moderate yield. After carefully maintaining the temperature at -98°C and increasing the amount of nBuLi added, boronic ester 142 could be prepared in 71% yield.

O O 1) NCS, NaH MeO HO 1) PyHBr , MeOH H 2) AlCl3, pyridine H 3 73%(2 steps) 2) CH2Br2, Cs2CO3, HO HO DMF, 65% (2 steps) 148 Cl 149

Br O Br Bpin nBuLi, THF O NaH, THF H O iPrOBpin O Ph3PCH3Br O 99% O 71% O Cl Cl Cl 150 151 142 Scheme 5.3 Preparation of Boronic Ester 142

5.6 Sequential Suzuki Cross Couplings

Having two boronic esters in hand, we decided to investigate the Suzuki coupling sequentially before the one-pot three-component Suzuki reaction. Based on our previous studies of dibromobenzene 143, the alkene slows the rate of oxidative addition of palladium through coordinative saturation of the agostic Pd(0) complex, thus the bromide ortho to styrene is less reactive compared to the bromide para to the styrene. Boronic ester 141 was subjected to bromostryene 143 in a typical Suzuki coupling condition, which involves Pd(PPh3)4 as the catalyst, aqueous sodium hydroxide as a base, and 1,4-dioxane as the solvent. Gratifyingly, the desired regioselective and chemoselective compound 152 was obtained in 54% as a single biaryl product. The major byproduct observed was the protodeboronation compound of aryl boronic ester 141, which is commonly found in sterically hindered Suzuki cross couplings. In attempt to increase the yield of the coupling reaction and consume bromide 143, 101

various reaction conditions were screened. Considering the instability of boronic ester 143, it was added protionwise (Entry 2) during the reaction until the consumption of dibromobenzene, and the yield was increased to 60%. Dimethoxyethane (Entry 5) has also been shown to be useful in sterically demanding Suzuki couplings,58 to our delight, the yield was slightly improved to 65%. Mechanistically, reaction pH might influence the rate of protodeboronation.59 Other bases such as potassium carbonate, potassium phosphate and cesium carbonate were tested, unfortunately, these bases proved to be less effective than sodium hydroxide in Suzuki cross coupling. It is notable that protodeboronation of 141 is slow in the presence of potassium phosphate, however, increasing the reaction temperature helped with cross coupling reaction but also accelerated the decomposition of boronic ester 141.

Br Br Bpin Pd(PPh3)4 OiPr KBr, base OMe + Br solvent (5% H2O) OiPr Cl OMe 62℃ 143 141 Cl 152 Entrya Solvent Base Boronic ester (eq) Yield (%) 1 1,4-dioxane NaOH 1.0 54 2 1,4-dioxane NaOH 1.0+0.5 60 3b 1,4-dioxane NaOH 1.0+0.5 50 4 DME NaOH 1.0 63 5 DME NaOH 1.0+0.2+0.2 65 6 DME K2CO3 1.0+0.2+0.2 19 7 DME K3PO4 1.0+0.2+0.2 46 c 8 DME K3PO4 1.0+0.2+0.2 56 c 9 DME Cs2CO3 1.0+0.2+0.2 30 a Bromide 143 (18.7 mg, 0.0642 mmol, 1.0 eq), boronic ester 141 (20.7 mg, 0.0642 mmol, 1.0 eq), Pd(PPh3)4 (7.4 mg, 0.00642 mmol, 10mol%), KBr (30.5 mg, 0.257 mmol, 4.0 eq), NaOH (10.3 mg, 0.257 mmol, 4.0 eq), solvent b c (5% H2O) 0.1M, 62℃, 42h; 1,4-dioxane 0.3 M; 70℃. Table 5.1 Reaction Conditions Screen for First Suzuki Coupling

With biphenyl 152 in hand, the second Suzuki coupling was performed using the same condition, but at an increased-temperature. Terphenyl was isolated as two diastereomers, 140a and 140b, in 53% overall yield. We hypothesized that the bulky substituents of terphenyl prohibits the rotation about C–C biaryl axis, causing the existence of two atropisomers.60 Some variants of reaction condition have been tested. 102

Similar to the first Suzuki coupling, dimethoxyethane can facilitate the cross coupling compared to 1,4-dioxane. Temperature is crucial for the second cross coupling. The reaction time was elongated due to the lower temperature (Entry 4), resulting in occurrence of side reactions prior to the coupling reaction. Boronic ester 142 can easily decompose under high temperature (Entry 3), therefore, heating at refluxing temperature did not help with the formation of terphenyl. Pleasingly, using 1.5 equivalent boronic ester all at once delivered terphenyls 140 in 69% yield, which is acceptable for second Suzuki coupling. Potassium phosphate (Entry 6) has also been tried, affording terphenyl 140 in comparable yield.

Cl Cl Br Cl Pd(PPh ) O 3 4 O O OMe KBr, NaOH O O + OMe + OMe OiPr O DME(5% H2O) 80℃, 69% OiPr OiPr Bpin Cl Cl Cl 152 142 140a 140b Entrya Solvent Base Temp(°C) Boronic ester (eq) Yield (%) 1 DME NaOH 80 1.0 53 2 1,4-dioxane NaOH 80 1.0+0.2+0.2 49 3 DME NaOH 85 1.0 43 4 DME NaOH 60 1.0+0.5 36 5 DME NaOH 80 1.5 69 6 DME K3PO4 80 1.0+0.3 64 a Bromide 152 (6.0 mg, 0.0147 mmol, 1.0 eq), boronic ester 142 (4.5 mg, 0.0147 mmol, 1.0 eq), Pd(PPh3)4 (1.7 mg, 0.00147 mmol, 10mol%), KBr (7.0 mg, 0.0589 mmol, 4.0 eq), NaOH (2.4 mg, 0.0589 mmol, 4.0 eq), solvent (5% H2O) 0.1 M, 80℃, 24h Table 5.2 Reaction Conditions Screen for Second Suzuki Coupling

A 2D NOESY correlation study was performed to determine the stereochemical relationship of the isopropyl group and olefin group (Figure 5.5). The significant correlation of terphenyl 140a was observed between methyl protons of isopropyl group and all the olefin protons of ring D. No correlation between isopropyl group and olefins was found in terphenyl 140b. Instead, methyl protons of isopropyl group could observe methoxy protons, methylenedioxy protons and one proton from ring C.

103

Cl Cl Cl Cl O Cl O O O O OMe O OMe O O O O OiPr O OiPr O Cl Cl Cl 140a 140b Figure 5.5 Stereochemical Analysis of Terphenyl via NOESY Correlation Spectroscopy

Atropisomers interconvert depending on time and temperature via bond rotation, and atropisomers which possess a half-life of 1000 seconds or longer would have have sufficient conformational stability for separation.61 Noticing the slow conversion from one single diastereomer to a mixture of two diastereomers in 1H NMR, we speculated that two terphenyls would equilibrate at high temperature to overcome the barrier of rotation. After heating terphenyl 140a and 140b side by side in deuterated benzene at 80°C, 1.15:1 distereomeric ratio was observed in 1H NMR.

Cl Cl Cl

O O C6D6 O O O O OMe + 80℃ OMe OMe OiPr OiPr OiPr

Cl Cl Cl 140a or 140b 140a:140b=1.15:1 Scheme 5.4 Atropisomerization of Terphenyls

5.7 Double Suzuki Coupling

The regioselective and chemoselective sequential cross coupling indicates the possibility of a double Suzuki reaction. 143 was coupled with 141 using the standard Suzuki conditions. While the consumption of dibromide was indicated on TLC, boronic ester 142 was simply added and the reaction was allowed to proceed to completion. To our delight, the first trial of a one-pot, three-component Suzuki coupling was successful, albeit in yield (~15%). Our current goal is increasing the yield of double Suzuki coupling by optimization of conditions. 104

Cl Br Cl Pd(PPh ) O 3 4 Cl Bpin KBr, NaOH O + + O Br O DME(5% H2O) OMe OiPr 62-80℃ OMe Bpin OiPr 143 141 142 Cl 140 Scheme 5.5 Double Suzuki Coupling

5.8 Ring Closing Metathesis

Two diastereomeric terphenyls were treated with Grubbs 2nd generation catalyst, both of them delivered the substituted phenanthrene 153 as spot to spot.

Cl Cl Cl

O O O O Grubbs cat. 2nd O Grubbs cat. 2nd O OMe OMe OMe DCM, 45℃ OiPr DCM, 45℃ OiPr OiPr quant. 90%

Cl Cl Cl 140a 153 140b Scheme 5.6 Ring Closing Metathesis of Terphenyls

In summary, the phenathrene moiety of Bazzanin K was successfully synthesized, featuring a one-pot, three-component double Suzuki reaction. Due the low yield of current double Suzuki coupling, optimization of conditions is still under investigation.

105

5.9 Supplemental Data

(CH2Cl2), and methanol (MeOH) were dried by passage through activated columns. 1,4-dioxane, dimethoxyethane (DME) were dried and distilled over sodium. All other reagents and solvents were used without further purification from commercial sources. Unless otherwise noted, melting points were obtained from material that solidified after chromatography. Instrumentation: FT-IR spectra were obtained on NaCl plates with a PerkinElmer Spectrum Vision spectrometer. Proton and carbon NMR spectra (1H NMR and 13C

2-Bromo-3-isopropoxybenzaldehyde (145). A solution of 2-bromo-3- hydroxybenzaldehyde (4.15 g, 20.6 mmol, 1.0 eq) and K2CO3 (5.70 g, 41.3 mmol, 2.0 eq) in DMF (41.2 mL, 0.5 M) was stirred for five minutes before the addition of iPrBr (3.80 g, 30.9 mmol, 1.5 eq) and the reaction was heated to 55 °C. After 6 h, the reaction was cooled to rt and quenched with aqueous LiCl and extracted with Et2O (3×90 mL).

The combined organic layers were washed with aqueous LiCl (40 mL), H2O (40 mL), and brine (40 mL). The organic phase was dried over Na2SO4, filtered, and concentrated to yield 145 (4.98 g, 99%) as a yellow oil.

Data for 145: Rf 0.76 (3:1 hexanes:EtOAc); IR (thin film) 3070, 2978, 2933, 2871, -1 1 1690, 1567, 1267, 1237, 828 cm ; H NMR (400 MHz, CDCl3) δ 10.42 (s, 1 H), 7.50 106

(d, J = 7.6 Hz, 1 H), 7.33 (t, J = 8.0 Hz, 1 H), 7.14 (d, J = 8.0 Hz, 1 H), 4.60 (sept, J = 13 6.0 Hz, 1 H), 1.41 (d, J = 6.0 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 192.6, 155.0, 135.0, 128.1, 121.5, 121.7, 120.5, 119.0, 72.7, 21.9; HRMS (TOF MS ES+) cald for

C10H11O2Br [M+H]: 241.99424, found 241.99408.

2-Bromo-6-chloro-3-isopropoxybenzaldehyde (146). A solution of 2-bromo-3- isopropoxybenzaldehyde (1.55 g, 6.3 mmol, 1.0 eq) and pTsOH-H2O (2.43 g, 12.7 mmol, 2.0 eq), in CH3CN (31.5 mL, 0.2 M) were stirred and heated to 70 °C before the addition of N-chlorosuccinimide (0.85 g, 6.3 mmol, 10 eq). After 5 h, the reaction mixture was cooled to rt and quenched with aqueous Na2S2O3 solution and extracted with Et2O (3×60 mL). The organic layers were washed twice with aqueous Na2S2O3, twice with H2O, and once with brine. The organic phase was then dried over Na2SO4, filtered, and concentrated. Purification by FCC (12:1 hexanes:EtOAc) yielded 146 (1.13 g, 64%) as a yellow oil.

Data for 146: Rf 0.38 (12:1 hexanes:EtOAc); IR (thin film) 3075, 2978, 2932, 2872, 1 1704, 1557, 1445, 1285, 1108, 804 cm-1; H NMR (400 MHz, CDCl3) δ 10.33 (s, 1 H), 7.35 (d, J = 8.8 Hz, 1 H), 7.00 (d, J = 8.8 Hz, 1 H), 4.56 (sept, J = 6.0 Hz,1 H), 1.40 (d, 13 J = 6.0 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 190.5, 153.9, 133.2, 130.3, 126.2,

118.8, 116.4, 73.0, 21.8; HRMS (TOF MS ES+) cald for C10H11O2ClBr [M+H]: 276.9631, found 276.9630.

2-Bromo-4-chloro-1-isopropoxy-3-vinylbenzene (147). To a slurry of tBuOK (4.03 g, 36.0 mmol, 2.0 eq) in THF (180 mL, 0.1 M) was added methyltriphenylphosphonium bromide (14.1 g, 39.6 mmol, 2.2 eq) slowly, the reaction mixture was stirred for 10 min. A solution of aldehyde 146 (5.0 g, 18.0 mmol, 1.0 eq) in THF (20 mL, 0.9 M) was added slowly, the mixture was stirred for 30 min. Water (50 mL) was added and the aqueous layer was extracted with ethyl acetate (50 mL×3). The combined organic layers were dried over Na2SO4, filtered and concentrated. Purification by FCC (40:1 hexanes:EtOAc) yielded 147 as a colorless oil (4.74 g, 96%). 107

Data for 147: Rf 0.68 (30:1 hexanes:EtOAc); IR (thin film) 2979, 1557, 1446, 1384, -1 1 1281 cm ; H NMR (700 MHz, CDCl3) δ 7.29 (d, J = 8.8 Hz, 1 H), 6.77 (d, J = 8.8 Hz, 1 H), 6.65 (dd, J = 17.8, 11.6 Hz, 1 H), 5.68 (dd, J = 11.6, 1.2 Hz, 1 H), 5.65 (dd, J = 17.8, 1.2 Hz, 1 H), 4.52 (sept, J = 6.0 Hz, 1 H), 1.40 (d, J = 6.1 Hz, 6 H); 13C NMR

(176 MHz, CDCl3) δ 153.6, 138.2, 133.7, 128.9, 125.0, 122.5, 116.0, 114.6, 72.7, 22.0;

HRMS (TOF MS ES+) cald for C11H13OClBr [M+H]: 274.9838, found 274.9823.

2-(3-Chloro-6-isopropoxy-2-vinylphenyl)-4,4,5,5-tetramethyl-1,3,2- dioxaborolane (141). To a solution of bromide 147 (100 mg, 0.36 mmol, 1.0 eq) in THF (3.63 mL, 0.1 M) at -98 °C was added tBuLi (0.43 mL, 0.80 mmol, 2.2 eq) dropwise, the solution was stirred for 10 min. Then iPrOBPin (0.30 mL, 1.45 mmol, 4.0 eq) was added dropwise, and mixture was kept strring for additional 30 min. The reaction was quenched with water (3 mL) and extracted with ethyl acetate (5 mL×3).

The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (30:1 hexanes:EtOAc) afforded 141 (71.4 mg, 60%) as a white solid.

Data for 141: mp 94–96 °C; Rf 0.50 (30:1 hexanes: EtOAc); IR (thin film) 2978, 1578, 1 1443, 1326, 1143; H NMR (700 MHz, CDCl3) δ 7.22 (d, J = 9.1 Hz, 1 H), 6.95 (dd, J = 11.2, 6.3 Hz, 1 H), 6.69 (d, J = 9.1 Hz, 1 H), 5.59 (d, J = 17.5 Hz, 1 H), 5.40 (d, J = 11.2 Hz, 1 H), 4.51 (sept, J = 6.3 Hz, 1 H), 1.35 (s, 12 H), 1.32 (d, J = 4.9 Hz, 6 H); 13 C NMR (176 MHz, CDCl3) δ 159.7, 141.4, 135.9, 130.4, 124.4, 119.7, 112.5, 83.9,

70.5, 24.9, 22.1; HRMS (TOF MS ES+) cald for C17H25O3ClB [M+H]: 323.1585, found 323.1601.

4-Bromo-7-chlorobenzo[d][1,3]dioxole-5-carbaldehyde (150). To a solution of aldehyde 149 (7.50 g, 43.5 mmol, 1.0 eq) in MeOH (150 mL, 0.30 M) at 0 °C was added pyridine hydrobromide (18.6 g, 116.2 mmol, 2.7 eq), the reaction mixture was warmed up to rt and stirred for 1h. The solvent was evaporated and the residue was diluted with 1% H2SO4 (150 mL) and EtOAc (150 mL). After separation, the aqueous 108

layer was extracted with EtOAc (150 mL×2). The combined organic phases were washed with saturated Na2S2O3 solution, dried over MgSO4 and filtered. The solution was concentrated and used directly without further purification. To a solution of crude residue from above (1.00 g, 3.97 mmol, 1.0 eq) in DMF (9.92 mL, 0.4 M) was added Cs2CO3 (3.22 g, 9.92 mmol, 2.5 eq), the mixture was stirred at rt for 15 min before the dropwise addition of CH2Br2 (0.31 mL, 4.37 mmol, 1.1eq). The reaction mixture was warmed to 65 °C and kept stirring for 4h. After cooling to rt, the mixture was quenched with H2O (8 mL) and extracted with EtOAc (10 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (6:1 hexanes:EtOAc) delivered 150 (530 mg, 78%) as white solid.

Data for 150: mp 164–165 °C; Rf 0.55 (6:1 hexanes:EtOAc); IR (thin film) 1680, 1592, -1 1 1453, 1435, 1337, 1260 cm ; H NMR (700 MHz, CDCl3) δ 10.13 (s, 1H), 7.58 (s 1 13 H), 6.25 (s, 2H); C NMR (176 MHz, CDCl3) δ 188.6, 149.0, 147.6, 128.4, 126.3,

114.6, 103.3; HRMS (TOF MS ES+) cald for C8H5O3ClBr [M+H]: 262.9111, found 262.9110.

4-Bromo-7-chloro-5-vinylbenzo[d][1,3]dioxole (151). To a slurry of NaH (167.6 mg, 60 % dispersion in mineral oil, 4.19 mmol, 1.2 eq) in THF (23.3 mL, 0.15 M) at 0 °C was added methyltriphenylphosphonium bromide (1.497 g, 4.19 mmol, 1.2 eq), the mixture was stirred for 20 min. Then a solution of aldehyde 150 (920.0 mg, 3.49 mmol, 1.0 eq) in THF (11.6 mL, 0.3 M) was added dropwise, the reaction mixture was stirred for 4 h, quenched by H2O (20 mL) and extracted with EtOAc (20 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (6:1 hexanes:EtOAc) yielded 151 (901.8 mg, 99%) as white solid.

Data for 151: mp 78–80 °C; Rf 0.74 (6:1 hexanes: EtOAc); IR (thin film) 2922, 2851, -1 1 1457, 1435, 1247, 1060, 1039, 933 cm ; H NMR (700 MHz, CDCl3) δ 7.09 (s, 1 H), 6.88 (dd, J = 17.5, 11.2 Hz, 1 H), 6.13 (s, 2 H), 5.62 (d, J = 17.5 Hz, 1 H), 5.32 (d, J = 13 11.2 Hz, 1H); C NMR (176 MHz, CDCl3) δ 147.0, 143.6, 133.6, 132.5, 120.2, 116.5, 109

113.5, 102.3, 100.8; HRMS (TOF MS ES+) cald for C9H7O2ClBr [M+H]: 260.9318, found 260.9324.

2-(7-Chloro-5-vinylbenzo[d][1,3]dioxol-4-yl)-4,4,5,5-tetramethyl-1,3,2- dioxaborolane (142). To a solution of bromide 151 (300.0 mg, 1.15 mmol, 1.0 eq) in THF (5.74 mL, 0.2 M) at -98 °C was added nBuLi (0.57 mL, 1.26 mmol, 1.1 eq) dropwise, the solution was stirred for 10 min. Then iPrOBpin (0.702 mL, 3.44 mmol, 3.0 eq) was added dropwise, and mixture was kept strring for additional 2.5 h. The reaction was quenched with saturated NH4Cl (3 mL) and extracted with ethyl acetate

(6 mL×3). The combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated. Purification by FCC (10:1 hexanes:EtOAc) gave 142 (250.7 mg, 71%) as white solid.

Data for 142: mp 95–97 °C; Rf 0.30 (30:1 hexanes: EtOAc); IR (thin film) 2980, 1608, -1 1 1 1415, 1329, 1131cm ; H NMR (700 MHz, CDCl3) δ H NMR (700 MHz, CDCl3) δ 7.15 (dd, J = 17.6, 11.2 Hz, 1 H), 7.08 (s, 1 H), 6.06 (s, 1 H), 5.54 (dd, J = 17.2, 0.8 Hz, 13 1 H), 5.18 (dd, J = 10.8, 0.8 Hz, 1 H), 1.37 (s, 12 H); C NMR (176 MHz, CDCl3) δ 153.8, 142.8, 138.2, 136.3, 119.5, 116.2, 114.6, 101.7, 84.1, 24.8; HRMS (TOF MS

ES+) cald for C15H19O4ClB: 309.1065, found 309.1071.

Br Br Pd(PPh3)4 Cl Bpin KBr, NaOH OMe Br DME(5% H2O) OiPr OMe OiPr 62℃, 62% 28 26 Cl 37 5'-Bromo-3-chloro-6-isopropoxy-2'-methoxy-2,4'-divinyl-1,1'-biphenyl (152). A flask containing bromide (18.7 mg, 0.0642 mmol, 1.0 eq), boronic ester (20.7 mg,

0.0642 mmol, 1.0 eq), Pd(PPh3)4 (7.4 mg, 0.00642 mmol, 10 mol%), KBr (30.5 mg, 0.257 mmol, 4.0 eq) and NaOH (10.3 mg, 0.257 mmol, 4.0 eq) was degassed 3 times (evacuation followed by backfill with Ar), then dimethoxyethane (0.64 mL, 0.1 M) and

H2O (0.032 mL, 5% v/v) was added under argon and the mixture was heated to 62 °C for 16 h upon which time TLC indicated consumption of the boronic ester. Boronic 110

ester 141 (4.1 mg, 0.0128 mmol, 0.2 eq) was added to the mixture and continued heating for another 8 h. Boronic ester 141 (4.1 mg, 0.0128 mmol, 0.2 eq) was added to the mixture and continued heating for another 12 h upon which time TLC indicated completion of the reaction. The reaction mixture was cooled to rt, filtered and concentrated. Purification by FCC (100:1 hexanes:EtOAc) yielded 152 (16.9 mg, mmol, 65%) as a colorless oil.

Data for 152: Rf 0.62 (30:1 hexanes:EtOAc); IR (thin film) 2979, 2934, 1447, 1372, -1 1 1274, 1115 cm ; H NMR (700 MHz, CDCl3) δ 7.31 (d, J = 8.8 Hz, 1 H), 7.22 (s, 1 H), 7.08 (dd, J = 17.4, 10.9 Hz, 1H), 7.06 (s, 1 H), 6.82 (d, J = 8.8 Hz, 1 H), 6.47 (dd, J = 17.9, 11.6 Hz, 1H), 5.75 (dd, J = 17.4, 0.84 Hz, 1H), 5.39 (dd, J = 11.0, 0.91 Hz, 1 H), 5.24 (dd, J = 11.6, 1.5 Hz, 1 H), 5.12 (dd, J = 17.9, 1.5 Hz, 1 H), 4.33 (sept, J = 6.0 Hz, 13 1H), 3.75 (s, 3H), 1.14 (dd, J = 20.9, 6.1 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 156.5, 154.4, 137.2, 136.8, 136.1, 135.5, 132.9, 129.5, 128.2, 128.1, 124.8, 120.8, 116.1, 114.6, 113.8, 108.4, 71.5, 55.6, 22.0, 21.9; HRMS (TOF MS ES+) calcd for

C20H20O2ClBr [M+H]: 406.03351, found 406.03454.

Cl Br Cl Pd(PPh ) O 3 4 O OMe KBr, NaOH O + OMe OiPr O DME(5% H2O) 80℃, 53% OiPr Bpin Cl 25 37 27 Cl 7-chloro-4-(3'-chloro-6'-isopropoxy-6-methoxy-2',4-divinyl-[1,1'-biphenyl]-3-yl)- 5-vinylbenzo[d][1,3]dioxole (140a & 140b). A flask containing bromide (6.0 mg,

0.0147 mmol, 1.0 eq), boronic ester (6.8 mg, 0.0221 mmol, 1.5 eq), Pd(PPh3)4 (1.7 mg, 0.00147 mmol, 10 mol%), KBr (7.0 mg, 0.0589 mmol, 4.0 eq) and NaOH (2.4 mg, 0.0589 mmol, 4.0 eq) was degassed 3 times (evacuation followed by backfill with Ar), then dimethoxyethane (0.15 mL, 0.1 M) and H2O (0.0075 mL, 5% v/v) was added under argon and the mixture was heated to 80 °C for 14 h upon which time TLC indicated consumption of the bromide. The reaction mixture was cooled to rt, filtered and concentrated. Purification by FCC (70:1 hexanes:EtOAc) yielded 140a (2.8 mg, 37%) as a colorless film and FCC (30:1 hexanes:EtOAc) yielded 140b (2.5 mg, 32%) as a colorless film. 111

Data for 140a: Rf 0.44 (30:1 hexanes: EtOAc); IR (thin film) 2924, 2853, 1454, 1378, -1 1 1261, 1115 cm ; H NMR (700 MHz, CDCl3) δ 7.30 (d, J = 8.8 Hz, 1 H), 7.21 (s, 1 H), 7.16 (s, 1 H), 6.87 (s, 1 H), 6.82 (d, J = 8.8 Hz, 1 H), 6.50 (dd, J = 16.9, 11.7 Hz, 1H), 6.47 (dd, J = 17.4, 11.1 Hz, 1 H), 6.37 (dd, J = 17.5, 11.0 Hz, 1H), 5.99 (dd, J = 17.4, 1.3 Hz, 2 H), 5.74 (d, J = 17.5 Hz, 1 H), 5.36 (dd, J = 17.5, 0.84 Hz, 1 H), 5.23 (dd, J = 11.7, 1.5 Hz, 1 H), 5.39 (dd, J = 11.0, 0.91 Hz, 1 H), 5.19 (dd, J = 11.0, 0.77 Hz, 1 H), 5.18 (dd, J = 17.9, 1.5 Hz, 1 H), 5.03 (dd, J = 11.1, 0.84 Hz, 1 H), 4.32 (sept, J = 13 6.1 Hz, 1H), 3.79 (s, 3H), 1.11 (dd, J = 6.0, 3.2 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 156.8, 154.4, 146.8, 143.0, 137.1, 136.5, 134.6, 134.5, 133.5, 132.9, 132.5, 129.2, 129.2, 126.3, 124.9, 124.1, 120.6, 119.9, 119.1, 114.9, 114.7, 114.3, 113.2, 106.7, 101.9,

71.4, 55.6, 22.0, 21.9; HRMS (TOF MS ES+) calcd for C29H27O4Cl2 [M+H]: 509.1286, found 509.1274.

Data for 140b: Rf 0.33 (30:1 hexanes: EtOAc); IR (thin film) 2926, 1451, 1382, 1241, -1 1 1115 cm ; H NMR (700 MHz, CDCl3) δ 7.29 (d, J = 8.8 Hz, 1 H), 7.20 (s, 1 H), 7.14 (s, 1 H), 6.84 (s, 1 H), 6.82 (d, J = 8.8 Hz, 1 H), 6.53 (dd, J = 17.8, 11.6 Hz, 1H), 6.48 (dd, J = 17.5, 11.1 Hz, 1 H), 6.29 (dd, J = 17.5, 11.1 Hz, 1H), 6.01 (d, J = 1.3 Hz, 1 H), 5.95 (d, J = 1.3 Hz, 1 H), 5.74 (dd, J = 17.5 Hz, 1 H), 5.49 (dd, J = 17.4, 0.91 Hz, 1 H), 5.24 (dd, J = 11.6, 1.7 Hz, 1 H), 5.20 (dd, J = 11.0, 0.8 Hz, 1 H), 5.08 (dd, J = 17.9, 1.6 Hz, 1 H), 5.04 (dd, J = 11.0, 0.9 Hz, 1 H), 4.29 (sept, J = 6.1 Hz, 1H), 3.81 (s, 3H), 13 1.11 (dd, J = 21.8, 6.0 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 157.0, 154.6, 146.7, 143.0, 137.4, 136.5, 134.7, 134.6, 133.7, 133.3, 132.7, 129.2, 129.1, 126.4, 125.1, 124.1, 120.4, 119.9, 119.2, 115.3, 114.7, 114.5, 113.2, 106.5, 101.8, 71.7, 55.5, 22.0, 21.9;

HRMS (TOF MS ES+) calcd for C29H27O4Cl2 [M+H]: 509.1286, found 509.1262.

Cl Br Cl Pd(PPh ) O 3 4 Cl Bpin KBr, NaOH O + + O Br O DME(5% H2O) OMe OiPr 62-80℃ OMe Bpin OiPr 143 141 142 Cl 140 112

7-chloro-4-(3'-chloro-6'-isopropoxy-6-methoxy-2',4-divinyl-[1,1'-biphenyl]-3-yl)- 5-vinylbenzo[d][1,3]dioxole (140). A flask containing bromide (18.7 mg, 0.0642 mmol, 1.0 eq), boronic ester (20.7 mg, 0.0642 mmol, 1.0 eq), Pd(PPh3)4 (7.4 mg, 0.00642 mmol, 10 mol%), KBr (30.5 mg, 0.257 mmol, 4.0 eq) and NaOH (10.3 mg, 0.257 mmol, 4.0 eq) was degassed 3 times (evacuation followed by backfill with Ar), then dimethoxyethane (0.64 mL, 0.1 M) and H2O (0.032 mL, 5% v/v) was added under argon and the mixture was heated to 62 °C for 16 h upon which time TLC indicated consumption of the boronic ester. Boronic ester 141 (10.4 mg, 0.0321 mmol, 0.5 eq) was added to the mixture and continued heating for another 10 h. Boronic ester 142 (19.8 mg, 0.0642 mmol, 1.0 eq) was added to the mixture and continued heating for another 38 h upon which time TLC indicated completion of the reaction. The reaction mixture was cooled to rt, filtered and concentrated. Purification by FCC (60:1 hexanes:EtOAc) yielded 140 (4.9 mg, 15%) as a colorless film.

Cl Cl

O O O Grubbs cat. 2nd O OMe OMe OiPr DCM, 45℃ OiPr

Cl Cl 140a 153 4-Chloro-10-(3-chloro-6-isopropoxy-2-vinylphenyl)-9-methoxyphenanthro[3,4- d][1,3]dioxole (153). To a solution of terphenyl 140a (2.0 mg, 0.00393 mmol, 1.0 eq) in DCM (0.39 mL, 0.01 M) was added Grubbs 2nd catalyst (0.1 mg, 0.000157 mmol, 4 mol%), the reaction mixture was warmed to 45 °C and stirred for 16 h. The mixture was cooled to rt, filter and evaporated. Purification by FCC (60:1 hexanes:EtOAc) yielded 153 (1.9 mg, quant.) as a colorless film.

Data for 153: Rf 0.28 (30:1 Hexanes: EtOAc); IR (thin film) 2924, 2853, 1509, 1465, -1 1 1424, 1284, 1100, 1059 cm ; H NMR (700 MHz, CDCl3) δ 8.64 (s, 1 H), 7.55 (dd, J = 11.9, 9.1 Hz, 2 H), 7.44 (s, 1 H), 7.37 (d, J = 9.1 Hz, 1 H), 7.22 (s, 1 H), 6.89 (d, J = 8.4 Hz, 1 H), 6.52 (dd, J = 18.2, 11.9 Hz, 1 H), 6.38 (d, J = 10.5 Hz, 2 H), 5.11 (dd, J = 25.9, 11.9 Hz, 2 H), 4.35 (sept, J = 6.3Hz, 1 H), 3.87 (s, 3 H), 1.08 (dd, J = 17.5, 6.3 13 Hz, 6 H); C NMR (176 MHz, CDCl3) δ 156.4, 154.8, 143.7, 142.1, 137.5, 133.2, 113

130.4, 130.0, 129.2, 128.2, 127.0, 126.3, 125.7, 124.9, 121.6, 121.5, 120.6, 115.5, 114.8, 113.4, 107.1, 102.0, 71.5, 55.5, 22.0, 21.9;

Cl Cl

O O O Grubbs cat. 2nd O OMe OMe OiPr DCM, 45℃ OiPr

Cl Cl 140b 153

4-Chloro-10-(3-chloro-6-isopropoxy-2-vinylphenyl)-9-methoxyphenanthro[3,4- d][1,3]dioxole (153). To a solution of terphenyl 140b (2.0 mg, 0.00393 mmol, 1.0 eq) in DCM (0.39 mL, 0.01 M) was added Grubbs 2nd catalyst (0.1 mg, 0.000157 mmol, 4 mol%), the reaction mixture was warmed to 45 °C and stirred for 10 h. The mixture was cooled to rt, filter and evaporated. Purification by FCC (60:1 hexanes:EtOAc) yielded 153 (1.7 mg, 90%) as a colorless film.

114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138

48 Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds, Wiley, 1994. 49 Salih, M. Q.; Beaudry, C. M. Org. Lett. 2012, 14, 4026–4029. 50 Martini, U.; Zapp, J.; Becker, H. Phytochemistry 1998, 47, 89–96. 51 Asakawa, Y . Phytochem. 2004, 65, 623–669. 52 a) Y. Asakawa, Prog. Chem. Org. Nat. Prod. 1995, 65, 335–351; b) M. To- yota, T. Kinugawa, Y. Asakawa, Phytochemistry 1994, 37, 859–862; c) M. Toyota, M. Tori, K. Takikawa, Y. Shiobara, M. Kodama, Y. Asakawa, Tetra- hedron Lett. 1985, 26, 6097–

6100. d) Faisal A. Almalki and David C. Harrowven Eur. J. Org. Chem. 2016, 5738– 5746. 53 Keserü, G. M.; Nógrádi, M. Nat. Prod. Rep. 1995, 12, 69–75. 54 Qu, J.; Xie, C.; Guo, H.; Yu, W.; Lou, H. Phytochemistry, 2007, 68, 1767–1774. 55 Martini, U.; Zapp, J.; Becker, H. Phytochemistry 1998, 47, 89–96. 56 Zhao, P.; Young, M. D.; Beaudry, C. M. Org. Biomol. Chem., 2015, 13, 6162–6165. 57 Stavrakov, G.; Keller, M.; Breit, B. Eur. J. Org. Chem., 2007, 5726–5733. 58 Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett, 1992, 3, 207–210. 59 Cox, P. A.; Leach, A. G.; Campbell, A. D.; Lloyd-Jones, G. C, J. Am. Chem. Soc. 2016, 138, 9145−9157. 60 Peter W. Glunz, P. W. Bioorg. Med. Chem. Lett. 2018, 28. 53–60. 61 Oki, M. Top. Stereochem. 1983, 14, 1–81.

139

Chapter 6 Summary and Future Work

Dearomatization has enormous potential as a strategy in chemical synthesis. The highly functionalized atoms of arenes could be masked by aromaticity, and upon dearomatization, those reactive atoms can be readily applied to cascade bond formation and further manipulations. Chapter 1 provided a brief introduction of aromatization reaction and a detailed overview of the utilization of aromatization in classic total synthesis of strychnine and research in the Beaudry group. In chapter 2, an introduction of cephalotaxine and the mode of action of homoharringtonine in chronic myeloid leukemia was described. And it also reviewed some representative syntheses of cephalotaxine, organized by the construction of its center tricyclic ring system. The most common approach involves sequential construction of tricyclic ring system by first building the 1-azaspiro[4.4]nonane framework and then the azepine ring. Transannular cyclization of forming pyrrolobenzazepine portion is generally seen in one step synthesis of two rings of cephalotaxine. In an attempt to apply dearomatization of furan toward the synthesis of complex natural products, we chose to construct cephalotaxine using a furan oxidation-transannular Mannich reaction. Chapter 3 described the first approach to synthesizing macrocycle, the precursor to our key transformation. Due to the lack of electrophicility of ester functionality of furan, the formation of macrocycle was not observed. In chapter 4, the total synthesis of (–)-cephalotaxine and (–)-homoharringtonine was described. The key features involve an oxidative furan opening with spontaneous transannular Mannich reaction and a Noyori asymmetric hydrogenation. The undesired enantiomer of cephalotaxinone can be recycled through acid-mediated racemization. Molecular chirality plays a critical role in chemistry, biology, and medicine. Identification of chirality in molecules without sp3-hybridized stereogenic carbon atoms is not straightforward. Our group has been focusing on the synthesis of molecules with conformational chirality and the long-term aim is developing an efficient method of determining and predicting conformational chirality in compounds. Bazzanin K is a macrocyclic bis(bibenzyls) with diastereotopic protons at its two 140

methylene position, indicating the possibility of conformational chirality. In chapter 5, a synthetic approach toward bazzanin K was described. A sequential Suzuki coupling and ring closing metathesis of preparing its phenanthrene moiety has been described. And we are currently working on the one-pot three-component Suzuki coupling reaction. Future work includes finishing the total synthesis of bazzanin K, separation of the potential enantiomers using chiral HPLC and measurement the energy barrier of racemization.