- Email: [email protected]
The total synthesis of (±)-friedelin, an unsymmetrical, pentacyclic triterpene
TetrahedronLetters No. 14, pp 1071 - 1074, 1976. Pergamon Press. Printed in Great Britain.
THE TOTAL SYNWESIS OF (?)-FRIEDELIN,p4VLMSYMETRTUL,
PEmACYCLIC TRITERPENE
Robert E. Ireland* and David M. Walbal Division of Chemistry and Uremical Engineering California Instituteof Technology,Pasadena, CA 91125 ContributionNo. 5261 (Ree~ivedin USA 20 January 1976; received in UK for publication27 February 1976) In Previous reports' from these laboratoriesan efficient route to the pentacyclicdiether 1 was described. A pentacyclicdiaromatic compound of this type, with the trans-anti-transBCD ring structure and the correct array of angular methyl groups, was envisaged as a useful key intermediatefor the total synthesis of pentacyclictriterpenesof the friedelin (2)' structural group. The successfulcompletionof the total synthesisof (?I-alnusenoneQ'
m30&o:
establishedthe
d3
viability of this approach, and the present report describes the results of a continuationof this effort which has culminated in the first total synthesisof (?)-friedclin(2) itself. Certain modificationsof the previous (-+)-alnusenone (2) synthesiswere deemed wise at the outset of the present work. In the alnusenonescheme' the aromatic E ring of the diether f was reduced and methylated before reductionand methylationof the aromatic A ring. The known2'4 ease of the latter transformationdictated this strategy. For the Eriedelin (lOJ work such a sequence has disadvantages. First, the overall yield observed' for the alterationof the aromatic E ring was only 14%, principallyas a result of the initial Birch reduction. It was felt that this could be improved in the friedelin scheme by reversingthe aromatic ring modification sequence--i.e. , 4 before E rather than E before .A. Secondly, the process envisaged for the alterationof the aromaticA ring entailed a cationic cyclizationas one step (eq. 617). Had this reaction been performed on the acetylenicalcohol precursor i in which the severe steric strain in the CDE ring system had already been introduced by prior modificationof the aromatic E ring, there was a distinct possibilitytflatbackbone rearrangement3would compete effectivelywith cyclization. Thus, it seemed wise to undertake modificationof the aromaticA ring first; completionmight then lead to a higher yield in the
1071
1072 No. 14
=%
SCHEMEI: TOTALSmSIS
sev.
OF (k)-FRIEDELIN (lOla
steps2
25% c2H, CzHs
(31 steps,0.3% overallyield) 's,(C,H5),PLi,THF5; b,Li,NH3,tME,EtOH; c,C&I,DME;cJ,S$C1,EtOH,C6H6; e_,H,O,,aq.NaOH, ~BOH; ~_,~T~NHNH~,HOAC,CH~C~~~; g,CH,Li,EtpO; h,CF3C02H,(CF3CO)zO; i,LDA,Tw:; i,Zn-Ag, MP12,THF7;~,Li(OtBu)&lH,THF,C6H6,00; &DHP,POC13,CH2C:2; m_,Li(OtBu)BA1H,THF,CGH6, reflux;~,C~O~.ZP~,CH~C~Z; g,KOtBu,CH3I,THF; ~,Li,h&,THF,tBuOH; ~,ClPO(NMe2)2,DME,M@A, nBuLi@;r,Li,EtNH,,tBuOH; ~,pTsCH,MaOH,THF.
1073
No. 14
Birch reduction but, more important,would relegate the introductionof the principal source of steric strain--thecis-D/E ring fusion--tothe later, less sensitive stages.
To implementthis scheme, it was necessary to prepare the pentacyclicdiether kin
which
the location of the ethoxy and methoxy groups were interchangedfrom their locations in the (i-)-alnusenone precursorlJSCHEME I). This was accomplishedin 25% overall yield from the appropriateprecursors shown by the same procedurespreviouslyreported.2 Selective reduction of the aromaticA ring did occur in superioryield in this case, and modificationof this ring along the same lines as reported'previously in the total synthesis of (+)-shiononeled through the acetylenicalcohol 6 to the enol trifluoroacetate7. Methylenation of the enolate derived from this materlal2 byyreatment with lithium diisopr:pylamide provided the cyclopropylalcohol 8. In contrast to the experiencrin the (+)-shiononesynthesis,'cleavage of the cyclopropyl alcohol 8 was not particularlyefficient at this state. While both base (KOtBu,EtpO) and acid (S&Cl, ?tOH, C6HS) afforded predominatelythe desired 4-methylketone,in neither case was the yield better than 75%, and when coupled with the subsequentBirch reduction of the aromatic E ring, the enedione2 was available only in 35% overall yield. Since the complicationin the more favorableacid catalyzed cleavage appeared to entail backbone rearrangement3of the pentacyclicsystem not seen in the (i)-shiononework, an adequate solution to the problem was found when the aromatic E ring was reduced first and then the resulting dihydrocyclopropyl alcohol cleaved by acid treatment. Completionof the synthesis from the enedione2 by the introductionof the requisite E ring methyl groups followed the pattern used in the previous (+)-alnusenonesynthesis'after provision was made to protect the C3 oxygen function. (c)-friedelin(lA), mp 246.5-248',so formed gave tic and glpc behavior and infrared,pmr and mass spectra identicalto those of an authentic sample of the natural product isolated from cork.3 Acknowledgment Support of this work by grants from the National Science Foundationand the Hoffmann-LaRocht Foundation is gratefullyacknowledgeJ.
no. 14
References 1) National Defense EducationAct Trainee, 1971, 1974. 2) R. E. Ireland,M. I. Dawson, S. C. Welch, A. Hagenbach,J. Bordner and B. Trus, J. Amer. Chem. Sot., z,
7829 (1973).
3j E. J. Corey and J. J. Ursprung,J. Amer. Chem. Sot., 78, 5041 (1956);G. Brownlie, F. S. Spring, R. Stevensonand W. S. Strachen,3. Chem. Sot., 2419 (1956). 4) R. E. Ireland and L. N. Mander, J. Org. Chem., z,
689 (1967).
5) F. G. Mann and M. J. Pragnell, Chem. Ind. (London),1386 (1964). 6) D. Felix, J. Schreiber,G. Ohloff and A. Eschenmoser,Helv. Chim. Acta, 2,
2896 (1971).
7) H. W. Whitlock, Jr. and L. E. Overman, J. Org. Chem., 34_,1962 (1969);J. M. Denis, C. Girard and .J.M. Cania, Synthesis, 549 (1972). 8) R. E. Ireland,D. C. Muchmore and U. Hengartner,J. Amer. Chem. Sot., 2,
5098 (1972).
9) R. E. Ireland, C. A:Lipinski, C. J. Kowalski,J. W. Tilley and D. M. Walba, J. Amer. Chem. Sot., z,
3333 (1974).
THE TOTAL SYNWESIS OF (?)-FRIEDELIN,p4VLMSYMETRTUL,
PEmACYCLIC TRITERPENE
Robert E. Ireland* and David M. Walbal Division of Chemistry and Uremical Engineering California Instituteof Technology,Pasadena, CA 91125 ContributionNo. 5261 (Ree~ivedin USA 20 January 1976; received in UK for publication27 February 1976) In Previous reports' from these laboratoriesan efficient route to the pentacyclicdiether 1 was described. A pentacyclicdiaromatic compound of this type, with the trans-anti-transBCD ring structure and the correct array of angular methyl groups, was envisaged as a useful key intermediatefor the total synthesis of pentacyclictriterpenesof the friedelin (2)' structural group. The successfulcompletionof the total synthesisof (?I-alnusenoneQ'
m30&o:
establishedthe
d3
viability of this approach, and the present report describes the results of a continuationof this effort which has culminated in the first total synthesisof (?)-friedclin(2) itself. Certain modificationsof the previous (-+)-alnusenone (2) synthesiswere deemed wise at the outset of the present work. In the alnusenonescheme' the aromatic E ring of the diether f was reduced and methylated before reductionand methylationof the aromatic A ring. The known2'4 ease of the latter transformationdictated this strategy. For the Eriedelin (lOJ work such a sequence has disadvantages. First, the overall yield observed' for the alterationof the aromatic E ring was only 14%, principallyas a result of the initial Birch reduction. It was felt that this could be improved in the friedelin scheme by reversingthe aromatic ring modification sequence--i.e. , 4 before E rather than E before .A. Secondly, the process envisaged for the alterationof the aromaticA ring entailed a cationic cyclizationas one step (eq. 617). Had this reaction been performed on the acetylenicalcohol precursor i in which the severe steric strain in the CDE ring system had already been introduced by prior modificationof the aromatic E ring, there was a distinct possibilitytflatbackbone rearrangement3would compete effectivelywith cyclization. Thus, it seemed wise to undertake modificationof the aromaticA ring first; completionmight then lead to a higher yield in the
1071
1072 No. 14
=%
SCHEMEI: TOTALSmSIS
sev.
OF (k)-FRIEDELIN (lOla
steps2
25% c2H, CzHs
(31 steps,0.3% overallyield) 's,(C,H5),PLi,THF5; b,Li,NH3,tME,EtOH; c,C&I,DME;cJ,S$C1,EtOH,C6H6; e_,H,O,,aq.NaOH, ~BOH; ~_,~T~NHNH~,HOAC,CH~C~~~; g,CH,Li,EtpO; h,CF3C02H,(CF3CO)zO; i,LDA,Tw:; i,Zn-Ag, MP12,THF7;~,Li(OtBu)&lH,THF,C6H6,00; &DHP,POC13,CH2C:2; m_,Li(OtBu)BA1H,THF,CGH6, reflux;~,C~O~.ZP~,CH~C~Z; g,KOtBu,CH3I,THF; ~,Li,h&,THF,tBuOH; ~,ClPO(NMe2)2,DME,M@A, nBuLi@;r,Li,EtNH,,tBuOH; ~,pTsCH,MaOH,THF.
1073
No. 14
Birch reduction but, more important,would relegate the introductionof the principal source of steric strain--thecis-D/E ring fusion--tothe later, less sensitive stages.
To implementthis scheme, it was necessary to prepare the pentacyclicdiether kin
which
the location of the ethoxy and methoxy groups were interchangedfrom their locations in the (i-)-alnusenone precursorlJSCHEME I). This was accomplishedin 25% overall yield from the appropriateprecursors shown by the same procedurespreviouslyreported.2 Selective reduction of the aromaticA ring did occur in superioryield in this case, and modificationof this ring along the same lines as reported'previously in the total synthesis of (+)-shiononeled through the acetylenicalcohol 6 to the enol trifluoroacetate7. Methylenation of the enolate derived from this materlal2 byyreatment with lithium diisopr:pylamide provided the cyclopropylalcohol 8. In contrast to the experiencrin the (+)-shiononesynthesis,'cleavage of the cyclopropyl alcohol 8 was not particularlyefficient at this state. While both base (KOtBu,EtpO) and acid (S&Cl, ?tOH, C6HS) afforded predominatelythe desired 4-methylketone,in neither case was the yield better than 75%, and when coupled with the subsequentBirch reduction of the aromatic E ring, the enedione2 was available only in 35% overall yield. Since the complicationin the more favorableacid catalyzed cleavage appeared to entail backbone rearrangement3of the pentacyclicsystem not seen in the (i)-shiononework, an adequate solution to the problem was found when the aromatic E ring was reduced first and then the resulting dihydrocyclopropyl alcohol cleaved by acid treatment. Completionof the synthesis from the enedione2 by the introductionof the requisite E ring methyl groups followed the pattern used in the previous (+)-alnusenonesynthesis'after provision was made to protect the C3 oxygen function. (c)-friedelin(lA), mp 246.5-248',so formed gave tic and glpc behavior and infrared,pmr and mass spectra identicalto those of an authentic sample of the natural product isolated from cork.3 Acknowledgment Support of this work by grants from the National Science Foundationand the Hoffmann-LaRocht Foundation is gratefullyacknowledgeJ.
no. 14
References 1) National Defense EducationAct Trainee, 1971, 1974. 2) R. E. Ireland,M. I. Dawson, S. C. Welch, A. Hagenbach,J. Bordner and B. Trus, J. Amer. Chem. Sot., z,
7829 (1973).
3j E. J. Corey and J. J. Ursprung,J. Amer. Chem. Sot., 78, 5041 (1956);G. Brownlie, F. S. Spring, R. Stevensonand W. S. Strachen,3. Chem. Sot., 2419 (1956). 4) R. E. Ireland and L. N. Mander, J. Org. Chem., z,
689 (1967).
5) F. G. Mann and M. J. Pragnell, Chem. Ind. (London),1386 (1964). 6) D. Felix, J. Schreiber,G. Ohloff and A. Eschenmoser,Helv. Chim. Acta, 2,
2896 (1971).
7) H. W. Whitlock, Jr. and L. E. Overman, J. Org. Chem., 34_,1962 (1969);J. M. Denis, C. Girard and .J.M. Cania, Synthesis, 549 (1972). 8) R. E. Ireland,D. C. Muchmore and U. Hengartner,J. Amer. Chem. Sot., 2,
5098 (1972).
9) R. E. Ireland, C. A:Lipinski, C. J. Kowalski,J. W. Tilley and D. M. Walba, J. Amer. Chem. Sot., z,
3333 (1974).