Developing microcolin A analogs as biological probes

Article abstract of DOI:10.1016/j.bmcl.2005.06.020

Three microcolin A and B analogs have been synthesized. Their biological activity profiles were evaluated against several cell lines, revealing the existence of a structural determinant whose role in mediating the antiproliferative effect of the microcoli

Full text of DOI:10.1016/j.bmcl.2005.06.020

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4043–4047
Developing microcolin A analogs as biological probes
Amit K. Mandal,a John Hines,a Kouji Kuramochia and Craig M. Crewsa,b,c,
*
aDepartment of Molecular, Cell, and Developmental Biology, Yale University, New Haven, CT 06520-8103, USA
bDepartment of Chemistry, Yale University, New Haven, CT 06520-8103, USA
cDepartment of Pharmacology, Yale University, New Haven, CT 06520-8103, USA
Received 9 May 2005; revised 5 June 2005; accepted 6 June 2005
Available online 1 July 2005
Abstract—Three microcolin A and B analogs have been synthesized. Their biological activity profiles were evaluated against several
cell lines, revealing the existence of a structural determinant whose role in mediating the antiproliferative effect of the microcolins
has heretofore gone unrecognized. While previously described as potent immunosuppressive natural products, we found that these
microcolin analogs possessed no selective cytotoxicity when comparing immune cell versus nonimmune cell proliferation. In addi-
tion, the amino-terminus of microcolin A has been modified to incorporate a biotinylated polyethylene glycol linker. The biological
activity of this biotinylated microcolin A analog was determined. This affinity reagent was shown to retain limited biological activity
and thus can serve as a useful probe for elucidating the mechanism of action of microcolin A.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Pharmacological intervention aimed at suppressing the
immune system plays an important role in the manage-
ment of autoimmune diseases, the prevention of rejec-
tion after organ transplantation, and treatment of
graft versus host diseases. In recent years, many new
immunosuppressive drugs have been discovered and
developed for clinical use in organ transplantation.
The commonly used immunosuppressive drugs fall into
five groups with different mechanisms of action: (a) reg-
ulators of gene expression; (b) alkylating agents; (c)
inhibitors of de novo purine synthesis; (d) inhibitors of
de novo pyrimidine synthesis; and (e) inhibitors of
kinases and phosphatases. Despite phenomenal success
in this disease area, many of these drugs have trouble-
some side effects and high toxicity profiles. More effec-
tive and specific immunosuppressive therapy is needed
to further reduce the high morbidity due to infections,
malignancies, and graft loss due to chronic rejection
after organ transplantation. Thus, the search for more
efficacious and tolerant immunosuppressive drugs is vig-
orously pursued in both academia and pharmaceutical
industry. In our chemical biology program, we are inter-
ested in developing natural products as biological
probes to identify novel proteins for therapeutic inter-
vention (target validation) and the exploration of cell
biology (chemical genetics).1
Microcolins A and B are potent antiproliferative and
immunosuppressive natural products discovered by
Koehn and co-workers from the Venezuelan blue-green
algae Lyngbya majuscule (Scheme 1).2–4 Microcolins A
and B were found to be potent in the human two-way
mixed lymphocyte response (MLR) assay with EC50 val-
ues reported in the subnanomolar to nanomolar
range.3,5 Despite this potent activity and potential
clinical use, the mechanism of action for the microcolins
Keywords: Natural products; Immunosuppressant; Synthesis.
*
Corresponding author. Tel.: +1 203 432 9364; fax: +1 203 432
6161; e-mail: craig.crews@yale.edu
URL: http://www.yale.edu/crews
Scheme 1. Retrosynthetic analysis.
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.06.020

4044
A. K. Mandal et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4043–4047
remains unknown. Furthermore, microcolins are dis-
tinctive in structure relative to other immunosuppressive
drugs. This suggests that the immunosuppressive prop-
erties of microcolins A and B might be mediated by a
novel mechanism.
Boc-threonine 5 was coupled in the presence of BOP-Cl/
Et3N to commercially available valine derivative 6, fol-
lowed by acetylation (Ac2O/DMAP) to produce the
known dipeptide 76 in 62% yield. Boc deprotection of
7 and coupling of the resulting hydrochloride salt
(BOP-Cl/Et3N) to commercially available N-methyl
Boc-leucine 9 and subsequent hydrogenation furnished
the tripeptide carboxylic acid 3 in 38% yield.6 The tri-
peptide 3 was accessed in five steps in 24% yield from
commercially available starting material.
Microcolin A has been synthesized by Decicco and
Grover6 as well as by Andrus et al.7 Koehn et al.5 pre-
pared several analogs by semisynthetic modification
and/or degradation of the natural product. They showed
that the C-10 free hydroxy functionality, as opposed to
general oxygen functionality, is important for the bio-
logical activity. Striking loss of activity occurs upon
reduction of the pyrrolenone ring C2–C3 olefin. The
EC50 value for 2,3-dihydromicrocolin A is 10,000-fold
less potent than native microcolin A in both murine
and human MLR. However, the C22 OAc can be re-
placed with OH without any significant loss of activity.
It suggests that the pyrrolenone function is an important
structural element for immunosuppressive activity. De-
spite considerable research to determine a structure–ac-
tivity profile, there have been no reported studies to
determine the role of C4 methyl group and 34,36-dim-
ethyloctanoyl side chain. We were particularly interested
to know the role of 34,36-dimethyloctanoyl side chain
because we would like to design a molecular probe based
on microcolin A without impairing biological activity
for our biochemical studies. In this communication, we
report the synthesis of three microcolin analogs. We also
describe the design and synthesis of biotinylated microc-
olin A for use as a molecular probe. The pharmacolog-
ical activity profiles of each analog against different cell
lines are also discussed.
Our route to synthesize the pyrrolenone fragment was
based on a modified version of pentafluorophenyl es-
ter-based protocol reported by Andrus et al. (Scheme
3).7 Following SilvermanÕs protocol, lactone 11 was
opened up in the presence of NaN3/MeOH to give
Cbz-protected cis-4-hydroxyproline methyl ester 12 in
99% yield.8 Silylation and hydrolysis gave carboxylic
acid, which was reacted with pentafluorophenol to pro-
duce 13.
This stable activated ester 137 was reacted with imidate
14 derived from commercially available racemic 5-meth-
yl-2-pyrrolidinone to produce mixed imide 157 and 16 in
40% and 45% yields, respectively. Both diastereoisomers
were separated by flash column chromatography. Each
diastereoisomer was then converted to pyrrolenone 4a7
and 4b following the three-step sequence (Pd/C, BOC2O,
A very expedient and highly efficient synthesis of tripep-
tide was developed (Scheme 2). Commercially available
Scheme 3. Synthesis of the pyrrolenone. Reagents and conditions: (a)
Ph3P, DEAD, THF, rt, 3 h, 55%; (b) NaN3, MeOH, 45 ꢁC, 24 h, 99%;
(c) TBSCl, imidazole, DMF, rt, 36 h, 94%; (d) LiOH, THF/MeOH/
H2O, 0 ꢁC, 24 h, 81%; (e) pentafluorophenol, DCC, EtOAc, 90%; (f)
( ) 5-methyl-2-pyrrolidinone, BuLi, THF, 15 min, 13, 3 h, 15 = 40%,
16 = 45%; (g) 10% Pd/C, Boc2O, H2, MeOH, 1 h; (h) LDA, ꢀ78 ꢁC,
15 min, PhSeBr, H2O2, AcOH, H2O, 0 ꢁC, 30 min; (i) 4 N HCl,
dioxane, 0 ꢁC, 4 h, 4a = 57%, 4b = 51% over last three steps.
Scheme 2. Synthesis of the tripeptide. Reagents and conditions: (a)
BOP-Cl, Et3N, CH2Cl2, rt, 12 h, 67%; (b) Ac2O, DMAP, EtOAc, rt,
6 h, 93%; (c) 4 N HCl, dioxane, 0 ꢁC, 4 h, 89%; (d) BOP-Cl, 9, Et3N,
CH2Cl2, 8 h, 45%; (e) 10% Pd/C, H2, 96%.

A. K. Mandal et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4043–4047
4045
H2; LDA, PhSeBr, H2O2,9 and HCl). Pyrrolenone 4c
was synthesized from Cbz-protected proline following
Andrus et al. protocol (Scheme 4). Enantiomerically
pure 5-methyl-2-pyrrolidinone was easily synthesized
from commercially available 18 according to the condi-
tions of Amstutz et al.10
The tripeptide carboxylic acid 3 was coupled with 4a, 4b,
and 4c separately to produce 21, 22, and 23, respectively,
in 67%, 73%, and 56% yields (Scheme 5).6 Hydrochloric
acid treatment produced the salt, which was then cou-
pled with octanoic acid in the presence of BOP-Cl/
Et3N to yield 24,11 25,12 and 26.13
Once these compounds were in hand, we next investigat-
ed their activity on different cell lines (Table 1). We have
found that the microcolin A analog 24 potently inhibits
both primary lymphocytes and immune cell lines with
EC50 values in the low nanomolar range. In fact, 24 is
similarly potent at inhibiting the proliferation of all cell
lines tested in this study, including nontransformed
(mouse) C2C12 myoblast cells and even primary (bo-
vine) endothelial cells. Rather curiously, this is the first
study ever to test whether microcolin A actually is selec-
tive for immune cells (all previous studies have relied
heavily on the mixed lymphocyte reaction and exclusive-
ly on immune-derived cells and cell lines to assess
microcolin A activity). This finding raises significant
doubts about the therapeutic potential of microcolins
as immunosuppressants, and rather points toward their
more likely usefulness as antineoplastic agents.
Scheme 5. Synthesis of microcolin A analogs. Reagents and condi-
tions: (a) BOP-Cl, Et3N, CH2Cl2, rt, 12 h, 21 = 67%, 22 = 73%,
23 = 56%; (b) 4 N HCl, dioxane, 0 ꢁC, 4 h; (c) BOP-Cl, octanoic acid,
Et3N, CH2Cl2, 12 h, 24 = 63%, 25 = 50%, 26 = 43% over two steps.
Analogs 25 and 26 displayed striking loss of activity: 26,
which lacks the C10 prolylhydroxyl group, was 100-fold
less potent than 24, thereby confirming the importance
of this functional group in mediating the antiprolifera-
tive effects of the microcolins. Interestingly, 25, which
Table 1. Biological activity profiles of microcolin A analogs 24, 25,
and 26
Entry Cell name
IC50
24
Mouse mixed lymphocyte reaction 32.8 nM
Mouse PMA ionomycin-stimulated lymphocytes 30.3 nM
EL-4 (mouse thymoma)
45.6 nM
39.4 nM
66.7 nM
35.9 nM
43.4 nM
36.7 nM
Jurkat (human T-cell leukemia)
P-388 (mouse T-cell leukemia)
N1E-115 (mouse neuroblastoma)
C2C12 (mouse myoblast)
BAE (bovine aortic endothelial)
25
26
EL-4 (mouse thymoma)
P-388 (mouse T-cell leukemia)
6 lM
7 lM
EL-4 (mouse thymoma)
P-388 (mouse T-cell leukemia)
6 lM
5 lM
differs from 24 in the stereochemistry of the C4 methyl
group, had a 100-fold reduced potency as well. More-
over, in a competition assay, the potency of 24 to inhibit
cell proliferation was unaffected by the presence of mo-
lar excess 25, indicating that the proper stereochemistry
of the C4 methyl group is important for target protein
binding. This finding, therefore, has identified a novel
structural determinant important for the antiprolifera-
Scheme 4. Synthesis of pyrrolenone. Reagents and conditions: (a)
Ph3P, CBr4, MeCN, 4 h, 73%; (b) Bu3SnH, AIBN, PhH, 70 ꢁC, 3 h,
61%; (c) BuLi, THF, 15 min; (d) 10% Pd/C, Boc2O, H2, MeOH, 1 h,
90%; (e) LDA, ꢀ78 ꢁC, 15 min, PhSeBr, H2O2, AcOH, H2O, 0 ꢁC,
30 min, 60%; (f) 4 N HCl, dioxane, 0 ꢁC, 4 h, 50%.

4046
A. K. Mandal et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4043–4047
chemistry in mediating the activity of these compounds.
Despite the impaired biological activity of biotinylated
microcolin A, it can still be useful as molecular probe
for biochemical studies. We now aim to identify and
characterize the molecular target of microcolin A by
employing affinity chromatographic techniques. We will
report our progress in this regard in due course.
Acknowledgments
This work was supported by NIH Grant GM62120
(C.M.C.) and a Leukemia Lymphoma Society Award
(J.H.).
Scheme 6. Synthesis of biotinylated microcolin A derivative. Reagents
and conditions: (a) dPEG4TM-biotin acid, BOP-Cl, Et3N, CH2Cl2, rt,
12 h, 40%.
References and notes
1. Crews, C. M.; Splittgerber, U. Trends Biochem. Sci. 1999,
24, 317.
2. Koehn, F. E.; Longley, R. E.; Reed, J. K. J. Nat. Prod.
1992, 55, 613.
3. Zhang, L.-H.; Longley, R. E.; Koehn, F. E. Life Sci. 1997,
60, 751.
Table 2. Biological activity of biotinylated microcolin A
Entry
Cell name
IC50 (lM)
27
EL-4 (mouse thymoma)
P-388 (mouse T-cell leukemia)
10
10
4. Zhang, L.-H.; Longley, R. E. Life Sci. 1999, 64, 1013.
5. Koehn, F. K.; McConnell, O. J.; Longley, R. E.; Sennett,
S. H.; Reed, J. K. J. Med. Chem. 1994, 37, 3181.
6. Decicco, C. P.; Grover, P. J. Org. Chem. 1996, 61, 3534.
7. Andrus, M. B.; Li, W.; Keyes, R. F. J. Org. Chem. 1997,
62, 5542.
8. Gomez-Vidal, J. A.; Silverman, R. B. Org. Lett. 2001, 3,
2481.
9. Reich, H. J.; Renga, J. M.; Reich, I. L. J. Am. Chem. Soc.
1975, 97, 5434.
tive activity of the microcolins, and further corroborates
the notion that the pyrrolylproline functionality is the
active pharmacophore.
Biotinylated derivatives of the epoxide containing natu-
ral products fumagillin, parthenolide, epoxomicin, and
eponemycin have proven useful in the identification of
their intracellular targets.14–17 Structure–activity rela-
tionship studies by Koehn et al. and our studies involv-
ing the C4 methyl group of microcolin A established
that the pyrrolylproline C-terminus is the pharmaco-
phore. Our studies with the octanoyl analog demon-
strated that 34,36-dimethyl side chain plays an
anciliary role in imparting pharmacological activity.
Therefore, we decided to attach a biotin handle to the
amino-terminus of microcolin A (Scheme 6). The hydro-
chloride salt derived from 21 was coupled with commer-
cially available PEG-biotin carboxylic acid to furnish
biotinylated microcolin A 27 in modest 40% yield.18
10. Amstutz, R.; Ringdahl, B.; Karlen, B.; Roch, M.; Jenden,
D. J. J. Med. Chem. 1985, 28, 1760.
11. Compound 24. The substrate 21 (5.3 mg, 7.80 lmol) was
treated with 4 N HCl in dioxane (0.23 ml, 0.92 mmol) at
0 ꢁC and stirred for 4 h. The solvent was evaporated under
reduced pressure to give crude hydrochloride salt, which
was carried to the next step without further purification. A
solution of hydrochloride salt and octanoic acid (1.6 ll,
10 lmol) in CH2Cl2 (1 ml) was treated with BOP-Cl
(2.55 mg, 10 ll) and Et3N (1.4 ll) at 0 ꢁC. It was then
warmed to room temperature and stirred for 12 h. The
reaction was quenched by the addition of aqueous
saturated NaHCO3. The aqueous layer was extracted with
ethyl acetate (2 · 2 ml). The organic layers were combined,
washed with brine, dried (Na2SO4), filtered and evaporat-
ed under reduced pressure. The residue was purified by
preparative TLC (90% EtOAc in hexane) to give com-
Once the synthesis was complete, we measured the
inhibitory potency of the biotinylated microcolin A 27
(Table 2). Of all the analogs tested in this study, 27
had the least biological activity, with IC50 values be-
tween 200- and 300-fold less potent than 24 when tested
against P-388 cells and EL-4 cells, respectively. Never-
theless, at these readily achievable higher concentra-
tions, this reagent may prove useful in identifying the
target protein of these small molecules.
20
1
pound 24 (3.5 mg, 63%). ½aꢁD ꢀ125.4 (c 0.22, CHCl3); H
(500 MHz, CDCl3) dH: 7.29 (1H, dd, J = 8.1 and 2.2), 6.89
(1H, d, J = 10.0), 6.10 (1H, dd, J = 6.0 and 1.5), 5.68 (1H,
dd, J = 10 and 2.0), 5.25 (3H, m), 5.05 (1H, d, J = 11.0),
4.97 (1H, dd, J = 9.5 and 3.5), 4.83 (1H, m), 4.40 (1H, br
s), 3.85 (2H, m), 3.13 (3H, s), 2.94 (3H, s), 2.49 (1H, m),
2.37 (2H, m), 2.22 (1H, m), 2.03 (3H, s), 2.01 (1H, m), 1.67
(4H, m), 1.49 (3H, d, J = 6.5), 1.31 (10H, br m), 1.18 (3H,
d, J = 7.0), 1.01 (3H, d, J = 6.5), 0.95 (3H, d, J = 6.5), 0.90
(6H, m) and 0.83 (3H, d, J = 7.0); 13C NMR (125 MHz,
CDCl3) dc: 175.1, 174.6, 171.6, 170.2, 170.2, 154.5, 125.7,
72.3, 69.0, 59.2, 58.6, 58.0, 56.8, 54.2, 51.9, 37.1, 36.3, 33.9,
30.8, 30.5, 29.7, 29.3, 29.6, 27.2, 25.5, 24.0, 23.1, 23.6, 21.9,
22.1, 19.0, 18.3, 17.8, 17.4, 14.3. IR (neat) 3420, 3011;
HRMS (FAB+): calcd for C37H62N5O9 (M+H)+ 720.4548;
found 720.4550.
In conclusion, we have synthesized microcolin analogs
as well as designed and synthesized biotinylated microc-
olin A as a biological probe. It was also demonstrated
that microcolin A analog 24 is comparably potent to
the native microcolin A 1 and that, unexpectedly, shows
no immune cell selectivity in its antiproliferative activity
when tested on a wide panel of cell types. Our study has
revealed the importance of the C4 methyl group stereo-

A. K. Mandal et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4043–4047
4047
12. Compound 25. Compound 22 (12.2 mg, 21 lmol) was
treated with 4 N HCl in dioxane (0.5 ml, 2.0 mmol) and
allowed to stir at 0 ꢁC. After 4 h, the solvent was removed
to give the crude. To a solution of octanoic acid (5 ll,
32 lmol) and the crude amine in CH2Cl2 (3 ml) triethyl-
amine (4 ll, 29 mmol) was added followed by BOP-Cl
(6.7 mg, 26 lmol) at room temperature. After being stirred
for 16 h, the mixture was quenched by the addition of
H2O. The mixture was partitioned between ethyl acetate
and water. The organic layer was washed with brine, dried
with sodium sulfate, filtrated, and evaporated under
reduced pressure. The residue was purified by preparative
TLC (17% hexanes in ethyl acetate) to yield 25 (7.5 mg,
(1H, m), 2.03 (3H, s), 2.01 (1H, m), 1.97 (1H, m), 1.88 (1H,
m), 1.69 (4H, m), 1.48 (3H, d, J = 6.5), 1.33 (9H, m), 1.18
(3H, d, J = 6.5), 1.01 (3H, d, J = 7.0), 0.95 (3H, d, J = 6.5),
0.91 (3H, d, J = 6.0), 0.89 (3H, t, J = 6.0), 0.82 Hz (3H, d,
J = 6.5); 13C NMR (125 MHz, CDCl3) dc: 174.6, 172.4,
171.6, 170.2, 168.7, 154.2, 125.9, 69.3, 60.4, 59.7, 58.5, 54.6,
52.6, 48.4, 36.8, 34.4, 32.1, 32.2, 31.0, 30.1, 29.8, 29.5, 29.3,
27.7, 25.5, 25.4, 25.0, 23.5, 23.0, 23.0, 22.4, 21.4, 19.3, 18.8,
17.6, 17.5, 14.5. IR (neat) 3417, 3015; HRMS (FAB+): calcd
for C37H62N5O8 (M+H)+ 704.4598; found 704.4590.
14. Sin, N.; Meng, L.; Wang, M. Q.; Wen, J. J.; Bornmann,
W. G., et al. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 6099.
15. Kwok, H. B.; Koh, B.; Ndubuisi, M.; Elofsson, M.;
Crews, C. M. Chem. Biol. 2001, 8, 759.
16. Meng, L.; Mohan, R.; Kwok, B. H. K.; Elofsson, M.; Sin,
N., et al. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10403.
17. Meng, L.; Kwok, B. H. B.; Sin, N.; Crews, C. M. Cancer
Res. 1999, 59, 2798.
20
50%) as white foam. ½aꢁD ꢀ135.2 (c 0.21, CHCl3); 1H
NMR (400 MHz, CDCl3) dH: 7.28 (1H, dd, J = 6.1 and
2.0), 6.90 (1H, d, J = 8.9), 6.10 (1H, dd, J = 6.1 and 1.5),
5.78 (1H, dd, J = 10.1 and 2.0), 5.25 (1H, m), 5.22 (1H, m),
5.02 (1H, d, J = 11.1), 4.95 (1H, dd, J = 8.8 and 3.4), 4.85
(1H, qt, J = 6.7 and 1.8), 4.40 (1H, br s), 3.85 (2H, m), 3.11
(3H, s), 2.92 (3H, s), 2.50 (1H, m), 2.35 (2H, m), 2.26 (1H,
m), 2.01 (3H, s), 1.98 (1H, m), 1.65 (4H, m), 1.44 (3H, d,
J = 6.7), 1.25 (10H, br s), 1.17 (3H, d, J = 6.5), 1.03 (3H, d,
J = 6.6), 0.94 (3H, d, J = 6.6), 0.89 (3H, d, J = 6.6), 0.88
(3H, t, J = 6.0), and 0.82 (3H, d, J = 6.6). 13C NMR
(125 MHz, CDCl3) dc: 174.3, 173.6, 171.2, 170.0, 169.9,
169.1, 154.5, 125.3, 71.7, 68.6, 59.2, 58.6, 58.0, 56.8, 54.2,
51.9, 37.1, 36.3, 33.9, 30.8, 30.5, 29.7, 29.4, 29.1, 27.2, 25.1,
25.0, 23.1, 22.6, 21.9, 21.1, 19.0, 18.4, 17.7, 17.4, 14.1. IR
(neat) 3420, 3011; HRMS (FAB+): calcd for C37H62N5O9
(M+H)+ 720.4548; found 720.4539.
18. Compound 27. Compound 21 (26.7 mg, 38 lmol) was
treated with 4 N HCl (1.5 ml, 6.0 mmol) and allowed to
stir at 0 ꢁC. After 4 h, the solvent was removed to give the
crude product. To a solution of dPEG4-biotin acid
(28.4 mg, 58 lmol) and the crude amine in CH2Cl2
(0.8 ml) triethylamine (20 ll, 144 lmol) was added fol-
lowed by BOP-Cl (15.2 mg, 60 lmol) at room tempera-
ture. After the mixture was stirred for 12.5 h, the solvent
was removed under reduced pressure. The residue was
purified by preparative TLC (14% methanol in chloro-
20
form) to yield 27 (16.3 mg, 40%) as white foam. ½aꢁD ꢀ43.6
(c 0.50, CHCl3); 1H NMR (400 MHz, CDCl3) dH: 7.27
(1H, dd, J = 6.0 and 2.0), 7.17 (1H, d, J = 8.6), 6.80 (1H,
br s), 6.73 (1H, br s), 6.21 (1H, br s), 6.09 (1H, dd, J = 6.0
and 1.5), 5.62 (1H, dd, J = 9.7 and 2.9), 5.23 (2H, m), 5.01
(1H, d, J = 11.1), 4.96 (1H, m), 4.81 (1H, m), 4.51 (1H, m),
4.38 (1H, br s), 4.33 (1H, m), 3.82 (2H, m), 3.73 (2H, m),
3.71–3.63 (14H, m), 3.57 (2H, br t, J = 7.8), 3.44 (2H, m),
3.16 (1H, m), 3.11 (3H, s), 2.93 (3H, s), 2.90 (1H, dd,
J = 12.8 and 3.8), 2.75 (1H, d, J = 12.8), 2.70 (1H, m), 2.61
(1H, m), 2.53 (1H, m), 2.27 (1H, m), 2.23 (1H, t, J = 7.4),
1.99 (3H, s), 1.97 (1H, m), 1.76–1.62 (6H, m), 1.47 (3H, d,
J = 6.7), 1.24 (2H, m), 1.17 (3H, d, J = 6.4), 0.99 (3H, d,
J = 6.4), 0.93 (3H, d, J = 6.6), 0.87 (3H, d, J = 6.5), and
0.81 Hz (3H, d, J = 6.7). 13C NMR (125 MHz, CDCl3) dc:
175.5, 174.1, 173.4, 172.0, 171.1, 170.1, 170.0, 169.9, 168.8,
154.2, 125.4, 71.4, 70.6, 70.5, 70.2, 69.9, 68.9, 67.3, 66.6,
59.3, 58.6, 58.2, 56.6, 54.4, 52.3, 51.7, 39.2, 36.6, 35.8, 34.9,
34.1, 31.9, 31.0, 30.6, 29.7, 28.2, 27.2, 25.5, 25.0, 23.1, 21.9,
21.8, 21.1, 19.2, 18.4, 17.4, 17.0. IR (neat) 3420, 3054;
HRMS (FAB+): calcd for C50H83N5O15 (M+H)+
1067.5699; found 1067.5718.
13. Compound 26. Compound 23 (20.0 mg, 30 lmol) was
treated with 4 N HCl (1.0 ml, 4.0 mmol) and allowed to
stir at 0 ꢁC. After 3 h, the solvent was removed to give the
crude. To a solution of octanoic acid (8 ll, 50 lmol) and the
crude amine in CH2Cl2 (1.5 ml) triethylamine (7 ll,
50 mmol) was added followed by BOP-Cl (15.2 mg,
60 lmol) at room temperature. After being stirred for
18.5 h, the mixture was quenched by the addition of H2O.
The mixture was partitioned between ethyl acetate and
water. The organic layer was washed with brine, dried with
sodium sulfate, filtrated, and evaporated under reduced
pressure. The residue was purified by preparative TLC (17%
hexanes in ethyl acetate) to yield 26 (9 mg, 43%) as colorless
20
oil. ½aꢁD ꢀ126.0 (c 0.53, CHCl3); 1H NMR (500 MHz,
CDCl3) dH: 7.25 (1H, dd, J = 6.0 and 2.0), 6.87 (1H, d,
J = 8.5), 6.08 (1H, dd, J = 6.0 and 1.5), 5.48 (1H, dd, J = 8.5
and 5.5), 5.26 (1H, dq, J = 6.5 and 2.0), 5.23 (1H, dd, J = 9.0
and 6.5), 5.06 (1H, d, J = 11.0), 4.98 (1H, dd, J = 9.0 and
6.5), 4.79 (1H, qt, J = 7.0 and 1.5), 3.83 (1H, m), 3.74 (1H,
m), 3.14 (3H, s), 2.93 (3H, s), 2.45 (1H, m), 2.37 (2H, m), 2.29