Studies toward luzopeptins: Assembly of the elusive serine-PCA dipeptide

Full text of DOI:10.1021/jo9701952

2320
J . Org. Chem. 1997, 62, 2320-2321
Sch em e 1a
Stu d ies tow a r d Lu zop ep tin s: Assem bly of
th e Elu sive Ser in e-P CA Dip ep tid e
Marco A. Ciufolini and Ning Xi
Department of Chemistry, MS 60, Rice University,
6100 Main Street, Houston, Texas 77005-1892
Received February 4, 1997
Luzopeptins A-C (BBM928, 1a -c, Chart 1) are me-
tabolites of Actinomadura luzonensis.1 Interest in these
substances, originally engendered by their extreme po-
tency against tumors,2 has been magnified by the dis-
covery of their inhibitory action against reverse tran-
scriptase (RT),3 a crucial enzyme for retroviral replication,
and consequently a target for intervention against HIV.
It is noteworthy that while antitumor effectiveness
decreases in the order A, B, C, antiretroviral activity
follows the opposite trend, and it is most pronounced in
the weakly cytotoxic C series at noncytopathic doses.4
No synthetic studies toward luzopeptins have been
reported, though the preparation of individual compo-
nents of 1 and of systems related to it have been
investigated.5 However, none of these model systems
contained the hallmark tetrahydro-4-hydroxypyridazine-
3-carboxylic acid 2 (“PCA”).5b Herein, we describe the
synthesis of a D-serine-PCA dipeptide,6 a goal that has
heretofore proved elusive7 due to the instability and poor
reactivity of 2.5b
a
(a) TBSCl, imidazole, DMF, rt, 16 h, 90%; (b) H2, 10% Pd/C,
cyclohexane, rt, 95%; (c) DMSO, (COCl)2, -78 °C; Et3N, CH2Cl2,
92%; (d) ethylene glycol, cat. PPTS, benzene, reflux, 3 h, 88%; (e)
TBAF, THF, rt, 4 h, 85%; (f) Cbz-NdN-Cbz, LDA, THF, -78 °C,
61%; (g) Ac2O, pyridine, rt, 16 h, 95%; (h) BOC2O, H2, 10% Pd/C,
rt, 8 h, 97%; (i) sym-collidine, CH2Cl2, 0 °C, 0.5 h, 60%; (j) 9:1 TFA/
H2O, 0.5 h, 97%.
The known scalemic ester 38 was advanced to com-
pound 8 in a straightforward manner, as shown in
Scheme 1.9 Gennari-Evans-Vederas10 reaction of 8
(1) (a) Ohkuma, H.; Sakai, F.; Nishiyama, Y.; Ohbayashi, M.;
Imanishi, H.; Konishi, M.; Miyaki, T.; Koshiyama, H.; Kawaguchi, H.
J . Antibiot. 1980, 33, 1087. (b) Tomita, K.; Hoshino, Y.; Sasahira, T.;
Kawaguchi, H. J . Antibiot. 1980, 33, 1098. (c) Arnold, E.; Clardy, J . J .
Am. Chem. Soc. 1981, 103, 1243.
(2) The most active member of the family, luzopeptin A, is reported
to be over 100 times more potent than mitomycin C against P388
leukemia (ref 1a,b).
(3) (a) Inouye, Y.; Take, Y.; Nakamura, S. J . Antibiot. 1987, 40, 100.
(b) Take, Y.; Inouye, Y.; Nakamura, S.; Allaudeen, H. S.; Kubo, A. J .
Antibiot. 1989, 44, 107.
(4) Luzopeptin C completely suppressed replication of HIV-1 in
infected MT-4 cells at around 2.5-5.0 µg/mL without significant
cytopathic effects (ref 3).
(5) First synthesis of PCA: (a) Hughes, P.; Clardy. J . J . Org. Chem.
1989, 54, 3260. Synthesis and chemistry of PCA: (b) Ciufolini, M. A.;
Xi, N. J . Chem. Soc., Chem. Commun. 1994, 1867 and references cited
therein. See also: Greck, C.; Bischoff, L.; Genet, J . P. Tetrahedron:
Asymmetry 1995, 8, 1989. Recent work on piperazic acids: (c) Schmidt,
U.; Braun, C.; Sutoris, H. Synthesis 1996, 223. Synthesis of 3-hydoxy-
valines: (d) Ciufolini, M. A.; Swaminathan, S. Tetrahedron Lett. 1989
30, 2037. (e) Shao, H; Goodman, M. J . Org. Chem. 1996, 61, 2582.
Synthesis of the quinaldic acid unit: (f) Boger, D. L.; Chen J . H. J .
Org. Chem. 1995, 60, 7369. Total synthesis of sandramycin, a congener
of 1 lacking the PCA unit: (g) Boger, D. L.; Chen, J .-H. J . Am. Chem.
Soc. 1993, 115, 11624. (h) Boger, D. L.; Chen, J .-H.; Saionz K. W. J .
Am. Chem. Soc. 1996, 118, 1629. (i) Unnatural analogues of 1: Olsen,
R. K.; Apparao, S.; Bhat, K. L. J . Org. Chem. 1986, 51, 3079.
(6) Portions of this work dealing with the preparation of hydrazine
8 by the route shown here and the use of our serinyl chlorides (12) for
the creation of PCA-serine dipeptides related to 14 were presented
at the 210th National Meeting of the American Chemical Society,
Chicago, IL, Aug 1995.
with dibenzyl azodicarboxylate furnished an 18:1 mixture
of anti (9, major, desired) to syn diastereomers. Sequen-
tial O-acetylation and catalytic debenzylation of the
emerging 10 in the presence of BOC2O yielded mono-BOC
hydrazine 11.11 No hydrogenolytic damage of the pre-
sumed free hydrazine intermediate, a fragile substance,
was incurred during this step. The delicate mono-BOC
hydrazine 11 condensed with the reactive, yet well-
behaved, protected D-serinyl chloride 1212 in the presence
of sym-collidine to afford 13.13 This material cyclized
cleanly and in high yield to blocked serine-PCA dipep-
tide 14 under aqueous acidic conditions.
Deacetylation of 14 could be effected cleanly and
quantitatively with hydrazine hydrate in acetonitrile.14
Subsequent attempts to convert 15 to 20 by Kunieda
cleavage15 of the oxazolone were complicated by the
extreme sensitivity of PCA esters related to 15 to
(10) (a) Gennari, C.; Colombo, L.; Bertolini, G. J . Am. Chem. Soc.
1986, 108, 6394. (b) Evans, D. A.; Britton, T. C.; Dorow, R. L.; Dellaria,
J . F. J . Am. Chem. Soc. 1986, 108, 6395. (c) Trimble, L. A.; Vederas,
J . C. J . Am. Chem. Soc. 1986, 108, 6397. See also: (d) Guanti, G.; Banfi,
L.; Narisano, E. Tetrahedron 1988, 44, 5553.
(11) The same transformation may also be achieved by BOC-
derivatization of the terminal nitrogen in 10 (BOC2O, 4-DMAP, Et3N,
CH2Cl2, rt, 96%) followed by hydrogenolysis (99%). The protocol
described in the text removes the need for a separate N-BOC formation.
(12) Xi, N.; Ciufolini, M. A. Tetrahedron Lett. 1995, 36, 6595.
(13) Other coupling methods based on DCC, DCC-HOBt, BOP-Cl,
mixed anhydrides, active esters, etc., failed completely to deliver a
serinyl derivative of 11. This may be attributed not only to the innate
lack of nucleophilicity of the N-2-atom (ref 5b), exacerbated in the
present case by the presence of an electron-withdrawing acyl group
on N-1, or to steric congestion in its surroundings, but also to the ease
of â-elimination of the OH functionality (or a protected variant thereof)
in serine derivatives wherein strong activation has been provided to
the COOH terminus.
(7) Cf. (a) Rebert, N. W. Dissertation, Utah State University, Logan,
UT, 1987. (b) Schmidt, U.; Riedl, B. Synthesis 1993, 809. (c) Schmidt,
U.; Riedl, B. Synthesis 1993, 815.
(8) Brooks, D.; Kellogg, R. P.; Cooper, C. S. J . Org. Chem. 1987, 52,
192.
(9) A noteworthy aspect of this sequence is that hydrogenolytic
debenzylation of 4 to 5 was best conducted in nonpolar cyclohexane.
Polar solvents such as EtOAc favored cyclization of 5 to the lactone.
Furthermore, we found that minimization of the number of steps after
introduction of the sensitive hydrazine unit (cf. 8 f 9) was crucial for
maximum efficiency. Therefore, it was best to advance 3 to 8 prior to
the Gennari-Evans-Vederas reaction.
S0022-3263(97)00195-3 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 8, 1997 2321
Ch a r t 1
Sch em e 2a
â-elimination of the oxygen functionality, regardless of
the type of protection provided to the OH and of the basic
agent employed.16 This problem was corrected by cre-
ation of the considerably less R-C-H-acidic secondary
amide 17.17 To that end, ester 15 was carefully hydro-
lyzed to acid 16, which in turn delivered 17 upon reaction
with BuNH2 and O-benzotriazol-1-yl-N,N,N′,N′-tetra-
methylisouronium hexafluorophosphate (“HBTU”) in DMF.
Interestingly, hydrolysis of 15 proceeded well with NaOH,
but not with the milder LiOH (much â-elimination).18
O-Acetylation of 17 served to ensure high regioselectivity
in favor of the oxazolone nitrogen upon reaction of 18
with BOC2O.19 Exposure of the resultant 19 to metha-
nolic Cs2CO3 prompted clean, rapid decarboxylative
opening of the oxazolone ring to give 20 (Scheme 2).
In light of the above observations, it is likely that a
suitable glycinyl derivative of 16 should be amenable to
conversion into a serine-PCA-glycine tripeptide suitable
for incorporation into advanced luzopeptin intermediates.
This suggests a plausible strategy for the construction
a
(a) NH2NH2, CH3CN, rt, 80%; (b) NaOH, THF/H2O, rt, 90%;
(c) BuNH2, HBTU, DMF, rt, 70%; (d) Ac2O, pyridine, rt, 3h, 95%;
(e) BOC2O, cat. DMAP, Et3N, CH2Cl2, 0 °C, 15min, 100%; (f)
Cs2CO3, MeOH, rt, 15min, 90%.
(14) Attempted deacetylation with other nitrogen (morpholine,
piperidine, n-butylamine, diethylamine), carbon (KCN/MeOH or MeCN),
or oxygen (LiOH; NaOH; MeOH/K2CO3 or Cs2CO3) nucleophiles
promoted â-elimination of the oxygenated functionality in PCA and
subsequent decomposition through aromatization to a pyridazine.
Reaction with pyrrolidine (ref 12) resulted not only in deacetylation
but also in formal substitution of the OH group by pyrrolidine, probably
by Michael addition of the amine into an intermediate R,â-unsaturated
ester.
of the challenging framework of 1a -c. Additional de-
velopments in this area will be reported in due course.
Ack n ow led gm en t. We acknowledge the NIH (CA-
55268), the NSF (CHE 95-26183), and the Robert A.
Welch Foundation (C-1007) for their support of our
research program. M.A.C. is a Fellow of the Alfred P.
Sloan Foundation, 1994-1998.
(15) Ishizuka, T.; Kunieda, T. Tetrahedron Lett. 1987, 28, 4185.
(16) Hydroxyl protecting groups examined included acetate, Et3Si,
and tBuMe2Si; bases included LiOH, NaOH, MeOH/K2CO3, or Cs2CO3.
(17) For an important literature analogy see: Askin, D.; Verhoeven,
T. R.; Liu, T. M.-H.; Shinkai, I. J . Org. Chem. 1991, 56, 4929.
(18) We attribute this to coordination of Li+ by the â-OH group, an
event that would facilitate departure of the oxygen functionality in
an elimination reaction.
(19) Reaction of unprotected 17 with BOC2O resulted in preferential
OH derivatization; moreover, variable amounts of the N-BOC deriva-
tive of the N-butylamide were formed. This problem was no longer
apparent with acetate 18.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedure and spectral data for selected compounds (12 pages).
J O9701952