Fragmentation of Golgi membranes by norrisolide and designed analogues

Article abstract of DOI:10.1016/j.bmcl.2004.08.003

The marine natural product norrisolide was found to induce irreversible fragmentation of the Golgi membranes. A hypothesis accounting for the chemical origins of this effect is proposed based on structure/function studies of norrisolide and designed analo

Full text of DOI:10.1016/j.bmcl.2004.08.003

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5035–5039
Fragmentation of Golgi membranes by norrisolide and
designed analogues
Thomas P. Brady,a Erin K. Wallace,b Sun Hee Kim,a Gianni Guizzunti,b
a,
Vivek Malhotrab, and Emmanuel A. Theodorakis
*
*
aDepartment of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0358, USA
bDepartment of Cell and Development Biology, University of California, San Diego, 9500 Gilman Drive,
La Jolla, CA 92093-0358, USA
Received 19 July 2004; revised 3 August 2004; accepted 4 August 2004
Available online 28 August 2004
Abstract—The effect of norrisolide (4) and designed analogues on the Golgi membranes is presented. We found that 4 is the first
compound known to induce an irreversible vesiculation of these membranes. To investigate the chemical origins of this new effect
we synthesized and evaluated a series of norrisolide analogues in which the chemical functionalities present in the parent structure
were altered. Such structure/function studies suggest that the perhydroindane core of 4 is critical for binding to the target protein,
while the C21 acetate unit is essential for the irreversible vesiculation of the Golgi membranes.
ꢀ 2004 Elsevier Ltd. All rights reserved.
The transfer of ribosome-synthesized proteins from
endoplasmic reticulum to their ultimate destination (re-
ferred to as secretory pathway)1 requires a multitude of
signals and involves passage through the Golgi com-
plex.2 The Golgi complex consists of an assembly of
membranes that play a key role in the final structural
modifications of proteins, their sorting and transport
to the relevant location.3
organization is maintained by a balance of membrane
input and output. These observations have raised sev-
eral issues related to the organization of the Golgi com-
plex and its function during cell cycle. For example, the
mechanism by which Golgi membranes are organized
into stacks of cisternae as well as their fragmentation
during mitosis and reassembly in each daughter cells
are important and contentious issues.2,5
Ultrastructural analysis has revealed that the Golgi
apparatus is composed of stacks of flattened cisternae,
referred to as cis, medial and trans, and the trans Golgi
network (TGN).4 Transport vesicles originating at the
rough endoplasmic reticulum (ER) bring newly synthe-
sized and correctly folded proteins (cargo) into the cis
Golgi cisternae. The cargo then travels across the Golgi
stack in a cis to TGN direction. Within each cisterna,
the cargo undergoes specific modification such as trim-
ming and addition of oligosaccharides, phosphoryl-
ation, and sulfation. The appropriately decorated
proteins are then transferred from TGN to their respec-
tive destination, such as endosomes and plasma mem-
branes, or are secreted from the cell. It is thus obvious
that the Golgi complex is a dynamic structure whose
In the absence of genetic screens, chemical reagents, and
natural products have proven to be highly useful in elu-
cidating the function and dynamics of the Golgi appara-
tus (Fig. 1).6 For example, studies with nocodazole (1)
have revealed that microtubules are essential for the
maintenance of Golgi apparatus in the percentriolar
position.7 Brefeldin A (2) was found to cause fusion of
the Golgi membranes with the ER and helped in unrav-
eling a retrograde (Golgi to ER) pathway.8 Ilim-
aquinone (3) was shown to induce a microtubule
independent and reversible fragmentation of the Golgi
membranes into small vesicles.9 The latter may be attrib-
uted to inhibition of methylation of a Golgi-specific
protein, which is involved in activation of trimeric
G-protein and its downstream target Protein Kinase D
(PKD).10 These components are required for formation
of transport carriers from the trans Golgi network.
Keywords: Natural product; Acetylation; Golgi complex; Secretory
pathway.
*
Herein we report our studies on the Golgi fragmentation
using norrisolide (4), a natural product containing an
Corresponding authors. Tel.: +1 858 822 0456; fax: +1 858 822 0386
(E.A.T.); e-mail: etheodor@chem.ucsd.edu
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.003

5036
T. P. Brady et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5035–5039
but lack the acetoxy functionality at the C19 center. In
compound 7 the unusual fused c-lactone-c-lactol motif
of the natural product has been truncated to a monocy-
clic system, while analogue 8 was further modified by
replacing the acetoxy unit with a more stable carbonyl
group. In compounds 9 and 10 the perhydroindane core
of norrisolide has been replaced by a tert-butyl group.
Analogues 11, 12 and 13, 14 representing the two hemi-
spheres of the natural product were also synthesized and
screened.
Compounds 5, 6, and 11–14 were synthesized during our
campaign toward the synthesis of norrisolide.12 The syn-
thesis of analogues 7 and 8 is highlighted in Scheme 1.
Crucial to this effort was the synthesis of aldehyde 18
representing the precursor of the truncated side chain
of the natural product. Compound 18 was prepared
from butenolide 1613 via a vinylcuprate addition14 at
the C11 center (norrisolide numbering) followed by
Figure 1. Chemical structures of selected Golgi-disturbing agents.
uncommon rearranged spongiane skeleton.11 We have
found, and report herein, that treatment of Normal
Rat Kidney (NRK) cells with norrisolide causes an irre-
versible fragmentation of Golgi complex, which is inde-
pendent of the microtubule structure. Using a series of
designed norrisolide analogues, shown in Figure 2, we
determined that the observed irreversibility stems from
the unique reactivity of the fused c-lactone-c-lactol
motif of 4. A hypothesis regarding the mode of action
of norrisolide based on our observations is proposed.
Evaluation of the structure of norrisolide as a function
of its activity in Golgi vesiculation was performed with
analogues 5–14, in which crucial functionalities of the
parent structure were partially deleted (Fig. 2). Com-
pounds 5 and 6 have an identical structural motif to 4
Scheme 1. Reagents and conditions: (a) vinylmagnesium bromide
(1.1equiv), CuBrÆMe2S (0.15equiv), THF, ꢀ78ꢁC, 1h, 73%; (b)
DIBAL-H (1.1equiv), CH2Cl2, ꢀ78ꢁC, 30min, 99%, (c) MeOH, HCl
(cat), 25ꢁC, 15min, 99%; (d) OsO4 (2.5%), NMO (1.2equiv), pyridine
(cat), acetone, 25ꢁC, 12h, 92%; (e) Pb(OAc)4 (1equiv), CH2Cl2, 25ꢁC,
20min, 89%; (f) NiCl2 (cat), CrCl2 (10equiv), DMF, 1h, 25ꢁC, 71%;
(g) Dess–Martin periodinane (1.5equiv), CH2Cl2, 25ꢁC, 8h, 95%; (h)
10% Pd/C, H2 (1atm), MeOH, 25ꢁC, 12h, 99%; (i) MeLi (10equiv),
DME, 0ꢁC, 30min; (j) SOCl2 (3equiv), pyridine (10equiv), CH2Cl2,
25ꢁC, 30min, 71% (over two steps); (k) TBAF (1.1equiv), THF, 25ꢁC,
6h, 99%; (l) IBX (2equiv), CH3CN, reflux, 2h, 99%; (m) methylmag-
nesium bromide (5equiv), THF, 0ꢁC, 30min, 65%; (n) Dess–Martin
periodinane (1.5equiv), CH2Cl2, 25ꢁC, 12h, 95%; (o) CrO3 (10equiv)
acetic acid/H2O/acetone (8/1/1), 25ꢁC, 2h, 83%; (p) mCPBA
(1.3equiv), NaHCO3, CH2Cl2, 0ꢁC, 45min, 84%.
Figure 2. Structures of designed norrisolide analogues.

T. P. Brady et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5035–5039
5037
reduction/protection at the C20 and oxidative cleavage
of the pendant alkene of 17. Nozaki coupling15 between
18 and 19,12 followed by oxidation of the resulting alco-
hol and hydrogenation of the C8–C9 alkene produced
compound 20, which after a two step olefination proce-
dure gave rise to adduct 21.16 The latter was functional-
ized at the C19 and C20 centers to furnish methyl ketone
8 that, after Baeyer–Villiger oxidation,17 produced ad-
duct 7.18 Analogues 9 and 10 were synthesized in an
analogous manner.19
All cells were visualized by immunofluorescence micro-
scopy with antibodies specific to Golgi membranes and
microtubules. The nuclei were visualized using the
DNA specific dye Hoechst (H3334, Molecular Probes).
Under these conditions the Golgi membranes appear
red (Alexa fluor 594 goat anti rabbit), the microtubules
green (Alexa fluor 488 goat anti mouse) and the nucleus
blue.21
The results of the above study are summarized in Figure
3. Within 60min of incubation with norrisolide (4) the
Golgi membranes were fragmented with little change
in microtubule organization (Fig. 3, column 2, row
1).22 The fragmentation of the Golgi apparatus was irre-
versible since removal of norrisolide by extensive wash-
ing did not promote its reassembly (Fig. 3, column 3,
row 1). Under these conditions the cells died, presum-
ably due to irreversible Golgi fragmentation.
The effects of norrisolide (4) and analogues 5–14 in the
Golgi apparatus were investigated using normal rat kid-
ney cells (NRK cells).20 Cells growing on coverslips in
complete growth medium were treated with a 30lM
stock solution of the compounds in DMSO, and then
incubated for 60min at 37ꢁC. The cells were then di-
vided in two groups. One set of coverslips (Fig. 3, col-
umn 2) was fixed after 60min of incubation and
processed for immunofluorescence. The other set (Fig.
3, column 3) was subjected to washout with PBS, to
eliminate all traces of the compounds. The cells were
then allowed to recover in fresh growth medium for
90min at 37ꢁC prior to the staining procedure for fluo-
rescence microscopy. Control cells (Fig. 3, column 1)
were treated only with DMSO and processed as above.
Compounds 5–8 exhibited partial Golgi vesiculation
without inducing any microtubule disassembly (Fig. 3,
column 2, row 2). However, in contrast with norrisolide,
in these cases the fragmentation was reversible and after
the washing protocol the cells recovered completely
(Fig. 3, column 3, row 2). NRK cells treated with
compounds 9–14 were indistinguishable from control
Figure 3. Effect of norrisolide and analogues on Golgi membranes. Column 1: untreated NRK cells; column 2: NRK cells treated with norrisolide
and analogues and incubated for 60min; column 3: NRK cells treated as in column 2 and then washed with buffer and incubated for an additional
90min. The Golgi apparatus is shown in red color, microtubules in green, and nucleus in blue.

5038
T. P. Brady et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5035–5039
untreated cells, suggesting that these analogues have no
effect on Golgi membranes (Fig. 3, columns 2 and 3,
row 3).
Acknowledgements
Financial support from the NIH (CA 086079) is grate-
fully acknowledged.
The Golgi fragmentation induced by norrisolide (4) and
analogues 5–8 is independent of the microtubule struc-
ture, as evidenced by fluorescence microscopy. Nonethe-
less, to exclude the possibility of undetectable changes in
the microtubules structure, NRK cells were preincu-
bated for 30min with 10lg/mL of Taxolꢂ, a microtu-
bule stabilizing agent, and then treated with
norrisolide and analogues as described above. Under
these conditions, compounds 4–8 still cause Golgi mem-
branes to vesiculate (data not shown). Thus, norrisolide
(4) and analogues 5–8 break down Golgi membranes
even in the presence of taxol-stabilized microtubules.
References and notes
1. For selected reviews on the secretory pathway see: Glick,
B. S. In Protein Targeting, Transport Translocation;
Dalbey, R. E., Von Heijne, G., Eds.; Academic: New York
2002; pp 358–376; Ostermann, J.; Stauber, T.; Nilsson, T.
In Protein Targeting, Transport Translocation; Dalbey, R.
E., Von Heijne, G., Eds.; Academic: New York, 2002; pp
377–401; Trombetta, E. S.; Parodi, A. J. Ann. Rev. Cell
Develop. Biol. 2003, 19, 649–676; Harter, C.; Reinhard, C.
Subcell. Biochem. 2000, 34, 1–38; Ellgaard, L.; Molinari,
M.; Helenius, A. Science 1999, 286, 1882–1888; Hendriks,
R. J. M.; Fuller, S. D. Subcell. Biochem. 1994, 22, 101–149;
Dempski, R. E.; Imperiali, B. Curr. Opin. Chem. Biol.
2002, 6, 844–850; Zhou, A.; Webb, G.; Zhu, X.; Steiner,
D. F. J. Biol. Chem. 1999, 274, 20745–20748; Zapun, A.;
Jakob, C. A.; Thomas, D. Y.; Bergeron, J. M. Structure
1999, 7, R173–R182; Haigh, N. G.; Johnson, A. E. In
Protein Targeting, Transport Translocation; Dalbey, R.
E., Von Heijne, G., Eds.; Academic: New York, 2002; pp
74–106.
2. For selected reports on the effect of Golgi during cell cycle
see: Maag, R. S.; Hicks, S. W.; Machamer, C. E. Curr.
Opin. Cell Biol. 2003, 15, 456–461; Lippincott-Schwartz,
J.; Zaal, K. J. M. Histochem. Cell Biol. 2000, 114, 93–103;
Acharya, U.; Mallabiabarrena, A.; Acharya, J. K.; Mal-
hotra, V. Cell 1998, 92, 183–192; Colanzi, A.; Deerinck, T.
J.; Ellisman, M. H.; Malhotra, V. J. Cell Biol. 2000, 149,
331–339; Lucocq, J. M.; Warren, G. EMBO J. 1987, 6,
3239–3246; Nelson, W. J. J. Cell Biol. 2000, 149, 243–248;
Sutterlin, C.; Lin, C.-Y.; Feng, Y.; Ferris, D. K.; Erikson,
R. L.; Malhotra, V. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
9128–9132; Lin, C.-Y.; Madsen, M. L.; Yarm, F. R.; Jang,
Y.-J.; Liu, X.; Erikson, R. L. Proc. Natl. Acad. Sci. U.S.A.
2000, 97, 12589–12594; Lowe, M.; Gonatas, N. K.;
Warren, G. J. Cell Biol. 2000, 149, 341–356.
Evaluation of the above data suggests that the perhydro-
indane core of 4 is essential for activity and appears to
function as the target recognition element. This could
explain why compounds 9–14, lacking the perhydroin-
dane core, have no effect on the Golgi membranes, while
analogues 4–8 induce a visible fragmentation. Among
the latter analogues, the fragmentation is more pro-
nounced with compounds 4–6 that contain the entire
C1–C20 backbone of the natural product, and is some-
what attenuated with compounds 7 and 8 that lack the
c-lactone subunit (data not shown). Of particular signif-
icance is the finding that norrisolide (4) induces an irre-
versible fragmentation of the Golgi membranes, while
the partially reduced analogues 5–8 induce a reversible
fragmentation. This observation invites the hypothesis
that a covalent modification of the target protein occurs
in the case of norrisolide (4), presumably due to the
highly electrophilic nature of the C21 acetyl group.
A possible scenario that rationalizes these findings
would involve selective binding of the perhydroindane
core of 4 into a hydrophobic pocket of a protein that
is involved in the signal transduction related to the
Golgi fragmentation. This binding could bring the c-lac-
tol-c-lactone side chain in close proximity to a nucleo-
phile, which can then undergo acetylation. The
identification of norrisolide dependent acetylation of a
Golgi specific protein will help reveal the significance of
acetylation dependent regulation of Golgi organization.
3. For selected reports on this topic see: Roth, J. Chem. Rev.
2002, 102, 285–303; Arnold, S. M.; Kaufman, R. J. New
Compreh. Biochem. 2003, 38, 411–432; van Vliet, C.;
Thomas, E. C.; Merino-Trigo, A.; Teasdale, R. D.;
Gleeson, P. A. Progr. Biophys. Mol. Biol. 2003, 83, 1–45;
Schulein, R. Rev. Physiol. Biochem. Pharmacol. 2004,
151, 45–91; Sifers, R. N. Science 2003, 299, 1330–
1331.
4. Rios, R. M.; Bornens, M. Curr. Opin. Cell Biol. 2003, 15,
60–66; Short, B.; Barr, F. A. Curr. Biol. 2000, 10, 583–585;
Farquar, M. G.; Palade, G. E. Trends Cell Biol. 1998, 8,
2–10.
5. Colanzi, A.; Suetterlin, C.; Malhotra, V. Curr. Opin. Cell
Biol. 2003, 15, 462–467; Warren, G.; Malhotra, V. Curr.
Opin. Cell Biol. 1998, 10, 493–498; Glick, B. S. Curr. Opin.
Cell Biol. 2000, 12, 450–456; Shorter, J.; Warren, G. Rev.
Cell Develop. Biol. 2002, 18, 379–420.
6. Dinter, A.; Berger, E. G. Histochem. Cell Biol. 1998, 109,
371–390.
7. Turner, J. R.; Tartakof, A. M. J. Cell Biol. 1989, 109,
2081–2088.
In conclusion, we present herein the effects of norrisolide
(4) and designed analogues on the Golgi membranes.
We found that the natural product induces an irrevers-
ible vesiculation of these membranes without affecting
the microtubules structure. To the best of our knowl-
edge, this is the first compound to have such an effect.
We have also synthesized a series of norrisolide ana-
logues in which several functionalities found in the
structure of this natural product were sequentially de-
leted or altered. This allowed us to investigate the chem-
ical origins of the irreversible fragmentation of the Golgi
membranes that is produced by 4. Such structure/func-
tion studies suggest that the perhydroindane core of 4
is critical for binding to the target protein, while the
C21 acetyl group unit is essential for the irreversible
vesiculation of the Golgi membranes.
8. Pelham, H. R. Cell 1991, 67, 449–451; Pelham, H. R.
Trends Biochem. Sci. 1990, 15, 483–486; Klausner, R. D.;
Donaldson, J. G.; Lippincott-Schwartz, J. J. Cell Biol.
1992, 116, 1071–1080; Reaves, B.; Banting, G. J. Cell Biol.
1992, 116, 85–94; Schiaky, N.; Presley, J.; Smith, C.; Zaal,

T. P. Brady et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5035–5039
5039
K. J. M.; Cole, N.; Moreira, J. E.; Terasaki, M.; Siggia, E.;
Lippincott-Schwartz, J. J. Cell Biol. 1997, 139, 1137–1155.
9. Takizawa, P. A.; Yucel, J. K.; Veit, B.; Faulkner, J. D.;
Deernick, T.; Soto, G.; Ellisman, M.; Malhotra, V. Cell
1993, 73, 1079–1090; Veit, B.; Yucel, J. K.; Malhotra, V. J.
Cell Biol. 1993, 122, 1197–1206.
10. Radeke, H. S.; Digits, C. A.; Casaubon, R. L.; Snapper,
M . L.Chem. Biol. 1999, 6, 639–647; Radeke, H. S.;
Snapper, M. L. Bioorg. Med. Chem. 1998, 6, 1227–1232;
Casaubon, R.; Snapper, M. L. Bioorg. Med. Chem. Lett.
2001, 11, 133–136; Jamora, C.; Takizawa, P. A.; Zaarour,
R. F.; Denesvre, C.; Faulkner, D. J.; Malhotra, V. Cell
1999, 91, 617–626.
(c 0.3, CH2Cl2); 1H NMR (400MHz, CDCl3) d 6.24 (d,
J = 5.6Hz, 1H), 5.42 (s, 1H), 4.87 (s, 1H), 4.48 (d,
J = 10.8Hz, 1H), 3.25 (m, 1H), 3.07 (t, J = 10.4Hz,
1H), 2.68 (dd, J = 3.2, 18.8Hz, 1H), 2.52 (dd, J = 10.8,
19.2Hz, 1H), 2.23 (s, 3H), 1.03 (s, 9H); 13C NM R
(100MHz, CDCl3) d 204.2, 174.5, 151.2, 113.1, 107.7,
85.1, 45.6, 42.8, 36.6, 30.3, 28.8, 27.2, 19.9, 14.1; HRMS
calcd for C14H20O4 (M+Na+) 275.1260, found 275.1281.
25
Compound 10: Rf = 0.4 (50% ether in hexanes); ½aꢁD
+9.18, (c 0.43, CH2Cl2); 1H NMR (400MHz, CDCl3) d
6.44 (d, J = 5.2Hz, 1H), 5.65 (d, J = 6Hz, 1H), 4.84 (s,
1H), 4.57 (s, 1H), 2.75 (m, 1H), 2.22 (m, 1H), 2.13 (dd,
J = 5.2, 18Hz, 1H), 1.65 (dd, J = 10.8, 18.8Hz, 1H), 1.54
(s, 3H), 0.69 (s, 9H); 13C NMR (100MHz, CDCl3) d 173.7,
168.6, 151.5, 136.1, 112.1, 106.8, 102.1, 47.3, 40.8, 36.3,
30.1, 28.4, 27.0, 20.3; HRMS calcd for C14H20O5
(M+Na+) 291.1209, found 291.1228.
11. Hochlowski, J. E.; Faulkner, D. J.; Matsumoto, J.;
Clardy, J. J. Org. Chem. 1983, 48, 1141–1142.
12. Brady, T. P.; Kim, S. H.; Wen, K.; Theodorakis, E. A.
Angew. Chem., Int. Ed. 2004, 43, 739–742.
13. Compound 16 is readily available from D-mannitol using
the procedures reported in the following publications:
Schmid, C. R.; Bryant, J. D.; Dowlatzedah, M.; Phillips, J.
L.; Prather, D. E.; Shantz, R. D.; Sear, N. L.; Vianco, C.
S. J. Org. Chem. 1991, 56, 4056–4058; Mann, J.; Wey-
mouth-Wilson, A. Carbohydr. Res. 1991, 216, 511–515;
Fazio, F.; Scheider, M. P. Tetrahedron: Asymmetry 2000,
11, 1869–1876.
14. Sahlberg, C. Tetrahedron Lett. 1992, 33, 679–682.
15. Takai, K.; Tagashira, M.; Kuroda, T.; Oshima, K.;
Utimoto, K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108,
6048–6050.
16. All new compounds exhibited satisfactory spectroscopic
and analytical data. Yields refer to spectroscopically and
chromatographically homogeneous materials.
17. (a) For selected reports on the Baeyer–Villiger reaction
see: Hudlicky, M. In Oxidations in Organic Chemistry;
Am. Chem. Soc.: Washington, DC, 1990; pp 186–195;
Mislow, K.; Brenner, J. J. Am. Chem. Soc. 1953, 75,
2318–2322; Goodman, R. M.; Kishi, Y. J. Am. Chem. Soc.
1998, 120, 9392–9393.
20. Similar results were obtained with experiments in HeLa
cells.
21. Reagents and cells: NRK cells were plated on 12mm glass
coverslips coated with Pronectin F (Sigma) and grown in
complete medium (250lL per coverslip), consisting of
alpha MEM medium (GIBCO) with 10% fetal calf serum,
2mM L-glutamine and 25mMHepes pH7.4, at 37 ꢁC in a
5% CO2 cell incubator. Stock solution (5mg/mL) of
norrisolide and analogues were made in DMSO and
stored at ꢀ20ꢁC. The working concentration of the
compounds was 30lMfor each coverslip. To half of the
coverslips (70% confluent) were added norrisolide or
analogues (2.5lL of the stock solutions). To the other
half were added 2.5lL of DMSO as negative control. Both
groups of cells were incubated at 37ꢁC for 60min. Part of
the treated cells (Fig. 1, column 2) and part of the control
cells (Fig. 1, column 1) were then fixed with 4% formal-
dehyde and processed for immunofluorescence micro-
scopy. The remaining cells were washed four times with
phosphate-buffered saline (PBS) (150mMNaCl, 1.8mM
NaH2PO4, 8.4mMNa 2HPO4). The cells were incubated in
fresh complete medium at 37ꢁC for 90min, then fixed with
4% formaldehyde and processed for immunofluorescence
microscopy (Fig. 3, column 3). Immunofluorescence
microscopy: For fluorescent labeling, cells were incubated
in blocking buffer (PBS containing 2.5% fetal bovine
serum and 0.1% Tween 20) for 30min at room tempera-
ture. The cells were then incubated for 1h at room
temperature in primary antibody diluted in blocking
buffer. Rat tubulin antibody (1:75) (Accurate Chemicals)
was used to detect microtubules; rabbit Mannosidase II
antibody (1:2000) (a gift from Dr. Kelly Moreman,
Vanderbilt University, TN) was used to visualize Golgi
apparatus. The cells were then washed three times with
PBS and incubated with secondary antibody, diluted in
blocking buffer, for 1h at room temperature. Alexa fluor
488 goat anti mouse (1:500) and Alexa Fluor 594 goat anti
rabbit (1:500) from Molecular Probes were used. Cells
were washed three times with PBS containing Hoescht
(1:100,000) (H33342, Molecular Probes) to stain DNA.
Coverslips were then mounted onto glass slides and
visualized using a Nikon micophot-FXA fluorescence
microscope at 60· magnification.
18. Spectroscopic and analytical data for compounds 7–8.
25
Compound 7: Rf = 0.4 (50% ether in hexanes); ½aꢁD +1.67,
(c 0.3, CH2Cl2); IR (film) mmax 2924, 1802, 1765, 1646,
1458, 1370, 1260, 1213, 1142, 1014, 704cmꢀ1 1H NM R
;
(400MHz, C6D6) d 6.45 (d, J = 1.2Hz, 1H), 4.82 (s, 1H),
4.68 (s, 1H), 2.68 (d, J = 8Hz, 1H), 2.41 (dd, J = 8.4,
17.6Hz, 1H), 2.11 (dd, J = 2.8, 17.6Hz, 1H), 1.90 (t,
J = 10.4Hz, 1H), 1.48 (s, 3H), 1.39–0.77 (m, 11H), 0.80 (s,
3H), 0.79 (s, 3H), 0.43 (s, 3H); 13C NMR (100MHz, C6D6)
d 174.1, 168.2, 146.7, 112.3, 99.1, 58.3, 58.0, 46.3, 43.7,
41.6, 39.6, 33.3, 33.2, 32.6, 30.2, 26.7, 25.1, 20.7, 20.3, 14.1;
HRMS calcd for C20H30O4 (M+Na+) 357.2042, found
357.2019. Compound 8: Rf = 0.6 (50% ether in hexanes);
25
½aꢁD +2.9, (c 0.7, CH2Cl2); IR (film) mmax 2945, 2855, 1791,
; R
1723, 1459, 1362, 1147, 1063, 899cmꢀ1 1H NM
(400MHz, CDCl3) d 5.13 (s, 1H), 5.05 (s, 1H), 4.56 (d,
J = 4.4Hz, 1H), 3.07 (m, J = 4.4Hz, 1H), 2.68 (dd, J = 9.2,
18Hz, 1H), 2.52 (dd, J = 5.2, 18Hz, 1H), 2.28 (s, 3H), 2.07
(t, J = 8Hz, 1H), 1.72–0.85 (m, 11H), 0.85 (s, 6H), 0.64 (s,
3H); 13C NMR (100MHz, CDCl3) d 204.8, 175.3, 148.5,
112.0, 88.0, 58.5, 58.1, 43.6, 43.2, 41.4, 41.3, 39.6, 34.2,
33.2, 33.1, 26.7, 25.3, 20.5, 19.9, 14.1; HRMS calcd for
C20H30O3 (M+H+) 319.2273, found 319.2298.
22. The fragmentation of the Golgi membranes is evidenced
by the appearance of Golgi vesicles (shown as red spots) in
the cytoplasm.
19. Spectroscopic and analytical data for compounds 9–10.
25
Compound 9: Rf = 0.5 (60% ether in hexanes). ½aꢁD +13.7,