Stereochemical promiscuity in artificial transcriptional activators

Article abstract of DOI:10.1021/ja0536567

Small molecule replacements of transcriptional activation domains are highly desirable targets due to their utility as mechanistic tools and their long-term therapeutic potential for a variety of human diseases. Here, we examine the ability of amphipathic

Full text of DOI:10.1021/ja0536567

Published on Web 08/18/2005
Stereochemical Promiscuity in Artificial Transcriptional Activators
Sara J. Buhrlage, Brian B. Brennan, Aaron R. Minter, and Anna K. Mapp*
Department of Chemistry and Department of Medicinal Chemistry, UniVersity of Michigan,
Ann Arbor, Michigan 48109
Received June 3, 2005; E-mail: amapp@umich.edu
The prevailing paradigm driving the discovery of small molecules
that perturb biological functions requires ligands of high affinity
and specificity for a particular macromolecule. This paradigm is
most effectively applied in cases where a single, well-defined bind-
ing site influences the function of the macromolecule. However,
many biological processes are carried out by macromolecular com-
plexes whose assembly and function are initiated through combina-
tions of weaker interactions (micromolar KDs) with functionally
redundant binding surfaces; in this way, a relatively small number
of proteins can reside in more than one complex and/or participate
in multiple functions.1 Small molecules that effectively regulate
such systems will likely need to mimic the behavior of the endo-
genous participants.
Transcriptional activation is an example of a process initiated
by a network of lower affinity interactions and is thus an ideal
system in which to explore the development of small molecule
regulators of macromolecular complexes. The activation domains
(ADs) of endogenous transcriptional activators typically exhibit a
low micromolar KD, multipartner binding profile that is likely an
essential functional contributor as the protein must mediate the
assembly of the large transcriptional machinery complex on DNA
(Figure 1a).2 Several lines of evidence suggest that the AD binding
sites within the transcriptional machinery are somewhat functionally
redundant, and particular placements of specific side chains are
not required for functionally productive binding interactions.2,3 For
example, ADs from VP16, Gcn4, and Gal4 target an overlapping
group of transcriptional machinery protein targets despite exhibiting
little sequence homology outside of a general amphipathic compo-
sition.3c,4 We recently described the first small molecule AD, an
isoxazolidine bearing functional groups seen in natural ADs (1,
Figure 1b).5 Here, we describe positional “mutagenesis” experiments
in which we evaluated analogs of 1 bearing identical side chains
in various locations within the isoxazolidine scaffold. The results
reveal that the isoxazolidine small molecule ADs mimic the
functional profile of natural ADs in that precise positioning of the
amphipathic side chains is not a critical determinant for function.
This finding has important implications for the design of future
small molecule transcriptional regulators.
Figure 1. (a) Transcriptional activator-mediated gene up-regulation, with
putative activation domain (AD) binding sites indicated in red. The DNA
binding domain (DBD) localizes the activator to a specific DNA site. (b)
1 contains functional groups commonly observed in natural ADs and
activates transcription well in vitro when localized to DNA.5 Hydrophobic
2 functions poorly, consistent with studies of natural ADs.5
The key intermediate for the preparation of 3, 5, and 6 is
isoxazoline 10, isolated as a single enantiomer in 88% yield
following a 1,3-dipolar cycloaddition reaction (Figure 2a).6
Toward 6, installation of the C3 benzyl group was accomplished
by silyl protection of the secondary alcohol of 10 followed by
addition of benzylmagnesium chloride (80% yield; 10:1 dr).6c The
major diastereomer was then treated with allyl bromide under
microwave conditions to alkylate N2 and provide isoxazolidine 13
(65% yield). Oxidative cleavage of the double bond installed the
requisite hydroxyl group on the N2 side chain, and treatment with
TBAF unmasked the 1,2-diol that was cleaved to provide an
aldehyde at C5; this sensitive intermediate was immediately
combined with Mtx (Figure 2a) and the resulting conjugate 6
isolated by reversed-phase HPLC. For diastereomers 3 and 5,
allylmagnesium chloride was employed as the nucleophile in the
initial addition reaction. Unlike the benzyl addition, the secondary
alcohol was not protected in order to reduce the diastereoselectivity
of the reaction and enable both diastereomers (11 and 12) to be
isolated (71% combined yield, 5:1 dr). The two diastereomers were
separated chromatographically, and each underwent installation of
the N2 benzyl group via alkylation (81% yield) and oxidative
cleavage of the C3 allyl group to provide 14 and 15. Straightforward
manipulations lead to the final conjugate targets 3 and 5. Isoxazo-
lidine 4 was prepared through an analogous reaction sequence
starting with the enantiomer of 9.
The function of the isoxazolidines was measured by their ability
to up-regulate transcription in a standard in vitro transcription assay
employing HeLa (human) nuclear extracts with the natural AD
ATF14 as a positive control (Figure 2c).5 The activity of enanti-
omers 3 and 4 is indistinguishable from that of AD 1 containing
both enantiomers of the isoxazolidine ring (Figure 2c). Isoxazo-
lidines 5-7 more significantly differ in the presentation of the
amphipathic functional groups due to stereochemical changes (5)
or positional changes within the ring (6 and 7). Nonetheless, all
function well as transcriptional ADs, in line with our prediction.
Isoxazolidine 7 showed the only noteworthy attenuation in activity,
with 35% lower functional levels relative to 1 (∼4-fold). In sum,
In our original experiments, all isoxazolidines were prepared as
racemates and were tested as stereoisomeric mixtures.5 We thus
targeted each enantiomer of the original isoxazolidine (3 and 4) as
well as a diastereomer (5) and two positional isomers (6 and 7) for
this study (Figure 2a). The compounds contain the same functional
groups found in the original active compound (1) but in varying
three-dimensional orientations since significantly altering the
hydrophobic and polar content of the molecules was found to
decrease function.5 Analogous to the natural system, we hypoth-
esized that all of the molecules would function as transcriptional
ADs.
9
12456
J. AM. CHEM. SOC. 2005, 127, 12456-12457
10.1021/ja0536567 CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. (a) Isoxazolidines with varying spatial orientations of polar and hydrophobic functional groups. (b) Synthetic scheme for isoxazolidines 3, 5, and
6. An analogous series of reactions were used for the preparation of 4 and 7.5 (c) Results from in vitro transcription assays. The DBD is the fusion protein
LexA-DHFR; the high affinity interaction between DHFR and methotrexate localizes the Mtx-tagged small molecules (50 nM) to DNA.11 Each activity is
the average of at least three independent experiments, with the indicated error (SDOM). The maximal activation is 7-fold relative to background. See
Supporting Information for additional details.
the data indicate that precise positioning of functional groups is
not the most important determinant of activator function in our
system.
were supported by the UM CBI training program (GM08597).
A.K.M. is an Alfred P. Sloan Fellow. We thank Prof. G. Glick for
helpful discussions.
The conserved activity across amphipathic, isomeric isoxazo-
lidines 3-7 parallels the functional behavior of the endogenous
amphipathic ADs that this molecular class was originally designed
to mimic. For example, the activity of 3 and 4 is consistent with
an earlier report that the D and L-enantiomers of the natural AD
ATF29 stimulate similar transcription levels in a cell-free system.7
Among peptidic ADs, a variety of combinations of polar and
hydrophobic amino acids function as ADs, but a hydrophobic/polar
balance is conserved.2,3a-c Further, like our small molecules,
endogenous ADs share a common structural motif; for natural ADs,
structural studies suggest that formation of a helix occurs upon
binding to a number of transcriptional machinery targets, although
other secondary structures may play a role.8 Also similar to the
isoxazolidines,5 mutations in natural ADs that disrupt the hydro-
phobic surface significantly decrease activation potential.9 One
remaining question is whether the similarities between the small
molecules and natural ADs extend to the binding surfaces within
the transcriptional machinery. Although the aggregate data are
suggestive of an affirmative answer, cross-linking experiments will
be required to provide a more definitive conclusion.
In sum, our data suggest that isoxazolidines are unlikely to be
the only suitable scaffolds for the construction of small molecule
transcriptional activation domains. Rather, a variety of appropriately
functionalized conformationally constrained small molecules should
also function well, a prediction currently under investigation. This
strategy obviates the need to identify high affinity ligands for single
protein targets and takes advantage of the remarkable functional
flexibility of the endogenous transcriptional regulatory system.
Given the increased interest in small molecule ADs as mechanistic
probes and therapeutic agents,10 this approach may find wide
application.
Supporting Information Available: Synthetic details for com-
pounds 3-7 and complete refs 1b and 1c. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Prudencio, M.; Ubbink, M. J. Mol. Recognit. 2004, 17, 524-39. (b)
Gavin, A.-C. et al. Nature 2002, 415, 141-7. (c) Figeys, D. et al. Nature
2002, 415, 180-3.
(2) Ptashne, M.; Gann, A. Genes & Signals; Cold Spring Harbor Laboratory:
New York, 2001.
(3) For example: (a) Ma, J.; Ptashne, M. Cell 1987, 51, 113-9. (b) Giniger,
E.; Ptashne, M. Nature 1987, 330, 670-2. (c) Melcher, K. J. Mol. Biol.
2000, 301, 1097-112. (d) Bryant, G. O.; Ptashne, M. Mol. Cell 2003,
11, 1301-9. (e) Cantin, G. T.; Stevens, J. L.; Berk, A. J. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 12003-8.
(4) Brown, C. E.; Howe, L.; Sousa, K.; Alley, S. C.; Carrozza, M. J.; Tan,
S.; Workman, J. L. Science 2001, 292, 2333-7.
(5) Minter, A. R.; Brennan, B. B.; Mapp, A. K. J. Am. Chem. Soc. 2004,
126, 10504-5.
(6) (a) Kanemasa, S.; Nishiuchi, M.; Kamimura, A.; Hori, K. J. Am. Chem.
Soc. 1994, 116, 2324-39. (b) Bode, J. W.; Fraefel, N.; Muri, D.; Carreira,
E. M. Angew. Chem., Int. Ed. 2001, 40, 2082-5. (c) Minter, A. R.; Fuller,
A. A.; Mapp, A. K. J. Am. Chem. Soc. 2003, 125, 6846-7.
(7) Nyanguile, O.; Uesugi, M.; Austin, D. J.; Verdine, G. L. Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 13402-6.
(8) For example: (a) Kussie, P. H.; Gorina, S.; Marechal, V.; Elenbaas, B.;
Moreau, J.; Levine, A. J.; Pavletich, N. P. Science 1996, 274, 948-53.
(b) Radhakrishnan I.; Perez-Alvarado, G. C.; Parker, D.; Dyson, H. J.;
Montminy, M. R.; Wright, P. E. Cell 1997, 91, 741-52. (c) Uesugi, M.;
Nyanguile, O.; Lu, H.; Levine, A. J.; Verdine, G. L. Science 1997, 277,
1310-3. (d) Parker, D.; Jhala, U. S.; Radhakrishnan, I.; Yaffe, M. B.;
Reyes, C.; Shulman, A. I.; Cantley, L. C.; Wright, P. E.; Montminy, M.
Mol. Cell 1998, 2, 353-9.
(9) (a) Regier, J. L.; Shen, F.; Triezenberg, S. J. Proc. Natl. Acad. Sci. U.S.A.
1993, 90, 883-7. (b) Drysdale, C. M.; Duenas, E.; Jackson, B. M.;
Reusser, U.; Braus, G. H.; Hinnebusch, A. G. Mol. Cell. Biol. 1995, 15,
1220-33. (c) Uesugi, M.; Verdine, G. L. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 14801-6.
(10) Ansari, A. Z.; Mapp, A. K. Curr. Opin. Chem. Biol. 2002, 6, 765-72.
(11) Baker, K.; Bleczinksi, C.; Lin, H. N.; Salazar-Jimenez, G.; Sengupta, D.;
Krane, S.; Cornish, V. W. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
16537-42.
Acknowledgment. The Burroughs Wellcome Fund and the
March of Dimes provided support for this work. A.R.M. and S.J.B.
JA0536567
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12457