A R T I C L E S
Wulff et al.
disruption of its function by the addition of a small nu-
cleophosmin-binding peptide15 leads to increased expression of
p53.16 Loss of p53 function (owing to mutation, deletion, or
hDM2 overexpression) is one of the most common features of
transformed cells, and novel approaches to restore cellular p53
function are widely sought as these have demonstrated potential
for tumor regression in vivo.17,18
While synthetic small molecules that bind to nucleophosmin
and thereby inhibit its participation in protein-protein and/or
protein-nucleic acid interactions might serve as potential leads
for the development of novel anticancer therapies,13 such
compounds are largely unknown.19-23
Figure 2. Images from fluorescence microscopy experiments with HeLa
S3 cells incubated for 2 h at 37 °C in medium containing 1 µM probe 4,
then fixed in methanol. (A) Direct fluorescence observed upon irradiation
with 365 nm light, attributed to excitation of the dansyl group of probe 4.
(B) Overlay of direct fluorescence output (green) with immunofluorescence
output from an antibody to nucleophosmin (red), used here as a nucleolar
marker.
Results and Discussion
By modifying one coupling partner in a late-stage, two-
component coupling reaction (step 15 of a 17-step synthetic
sequence),24 we have prepared more than 30 analogues of
avrainvillamide to date. For this study, we made use of the
dansyl- and biotin-conjugated probes 4 and 5 (Figure 1),
respectively, and the analogues 8-11 of Figure 4 (see Sup-
porting Information for details of their synthesis). We first
studied the antiproliferative effects of the conjugates 4 and 5
and found that both compounds inhibited the growth of T-47D
(breast cancer) cells with potencies similar to the natural product
(Figure 1). Although the biotin conjugate (5) was somewhat
less potent than the dansyl conjugate (4) in inhibiting the growth
of LNCaP (prostate cancer) cells, it did provide a GI50 value
similar to values measured with the structurally simpler analogue
2 and its biotin conjugate 3, compounds we had previously
studied and reevaluated herein as controls.3 Compounds 6 and
7 (Figure 1), which lack the unsaturated nitrone function but
contain the dansyl and biotin groups, respectively, as well as
the lipophilic tethering groups, were inactive in our assays,
suggesting that neither the tethers nor the reporter groups of
the active probes 3-5 contribute substantially to the observed
antiproliferative activities of these compounds.
Fluorescence microscopy studies conducted with the dansyl-
conjugate 4 revealed partial localization of the probe in the
nucleoli of HeLa S3 (cervical cancer) and T-47D cells, in
addition to a somewhat dispersed cytoplasmic distribution
(Figure 2 and Supporting Information).25
To identify potential binding proteins, populations of healthy
(adherent) T-47D cells were treated with the newly synthesized
biotin conjugate 5 or the structurally simpler biotin-containing
probe 3, previously studied. As a control, a separate population
of cells was treated with the biotin derivative 7, which lacks
the unsaturated nitrone function. The treated cells were incubated
with probe or control for 90 min at 37 °C and then were
harvested, washed, and lysed. The individual lysates were
exposed to an agarose resin to remove nonspecific binding
proteins. After centrifugation, the supernatants were then
exposed to a streptavidin-agarose resin. This resin was collected
by centrifugation and washed. Bound proteins were released
by heat-denaturation, separated by SDS-PAGE, and analyzed
by LC-MS/MS and Western-blot.
Nucleophosmin was initially identified by MS/MS sequencing
of a pool of proteins of broad molecular weight range obtained
using the structurally simpler probe 3. The analysis was
complicated by the presence of a number of nonspecific binding
proteins, including structural proteins such as actin, tubulin, and
myosin, as well as a number of biotinylated proteins, but the
identification of nucleophosmin in probe-treated but not control
protein samples was reproducible. With this information, MS/
MS sequencing of a protein pool of somewhat narrower
molecular weight range obtained using the more complex probe
5 also revealed a large peptide fragment with an amino acid
sequence corresponding to nucleophosmin (see Supporting
Information for details).
The presence of nucleophosmin in probe-derived (but not
control) protein samples was readily confirmed by Western-
blotting experiments (Figure 3A, compare lane 2 with lane 3,
and lane 4 with lane 5). Strikingly, probe 5 more effectively
bound nucleophosmin than did the structurally simpler and less
potent probe 3, even when a 3-fold higher concentration of 3
was used relative to 5 (compare lane 2 of Figure 3A with lane
4). This provided the first evidence that nucleophosmin might
(14) (a) Colombo, E.; Marine, J.-C.; Danovi, D.; Falini, B.; Pelicci, P. G. Nature
Cell Biol. 2002, 4, 529-533. (b) Maiguel, D. A.; Jones, L.; Chakravarty,
D.; Yang, C.; Carrier, F. Mol. Cell. Biol. 2004, 24, 3703-3711.
(15) Szebeni, A.; Herrera, J. E.; Olson, M. O. J. Biochemistry 1995, 34, 8037-
8042.
(16) Chan, H. J.; Weng, J. J.; Yung, B. Y. M. Biochem. Biophys. Res. Commun.
2005, 333, 396-403.
(17) Hollstein, M.; Sidransky, D.; Vogelstein, B.; Harris, C. C. Science 1991,
253, 49-53.
(18) For drug development candidates targeting p53-regulating proteins, see:
(a) Vassilev, L. T.; Vu, B. T.; Graves, B.; Carvajal, D.; Podlaski, F.;
Filipovic, Z.; Kong, N.; Kammlott, U.; Lukacs, C.; Klein, C.; Fotouhi, N.;
Liu, E. A. Science 2004 303, 844-848. (b) Peng, Z. Hum. Gene Ther.
2005, 16, 1016-1027.
(19) In addition to the peptide ligand discussed in refs 15 and 16, actinomycin
D (and related compounds) may bind to nucleophosmin. See: Busch, R.
K.; Chan, P. K.; Busch, H. Life Sci. 1984, 35, 1777-1785.
(20) Several cytotoxic compounds are known to cause translocation of nu-
cleophosmin from the nucleolus to the nucleoplasm or to the cytoplasm,
but a direct interaction has not generally been inferred. See: (a) Chan, P.
K. Exp. Cell Res. 1992, 203, 174-181. (b) Lee, H.-Z.; Wu, C-H.; Chang,
S.-P. Int. J. Cancer 2005, 113, 971-976. (c) Yung, B. Y.-M.; Busch, H.;
Chan, P.-K. Cancer Res. 1986, 46, 922-925. (d) Chan, P.-K.; Aldrich, M.
B.; Yung, B. Y.-M. Cancer Res. 1987, 43, 3798-3801.
(21) The S-glutathionylation of nucleophosmin has been reported, but the
cysteine residue involved in this transformation was not determined. See:
Townsend, D. M.; Findlay, V. J.; Fazilev, F.; Ogle, M.; Fraser, J.; Saavedra,
J. E.; Ji, X.; Keefer, L. K.; Tew, K. D. Mol. Pharm. 2006, 69, 501-508.
(22) Nucleophosmin has been identified as a receptor for phosphatidylinositol
lipids, which may contribute to its regulatory activity. See: Ye, K. Cancer
Biol. Ther. 2005, 4, 918-923.
(23) In contrast to the parent protein nucleophosmin, several inhibitors of the
hybrid oncoprotein NPM-ALK have been identified, but these presumably
act upon the kinase domain. For a recent example, see: Galkin, A. V.; et
al. Proc. Nat. Acad. Sci. U.S.A. 2007, 104, 270-275.
(25) Attempted immunofluorescence and immunoprecipitation experiments using
an antibody targeting the dansyl group were not successful, which may
indicate that the dansyl group of the conjugate 4 is not recognized by the
antibody, at least in its bound state.
(24) Herzon, S. B.; Myers, A. G. J. Am. Chem. Soc. 2005, 127, 5342-5344.
9
14446 J. AM. CHEM. SOC. VOL. 129, NO. 46, 2007