Small molecule macroarray construction via palladium-mediated carbon - Carbon bond-forming reactions: Highly efficient synthesis and screening of stilbene arrays

Article abstract of DOI:10.1002/chem.200903445

(Figure Presented) Arrays of stilbenes have been prepared in high purity by using Heck reactions on planar cellulose support, and demonstrate the compatibility of the small-molecule macroarray platform (see figure) with metal-mediated reactions for the fi

Full text of DOI:10.1002/chem.200903445

DOI: 10.1002/chem.200903445
ꢀ
Small Molecule Macroarray Construction via Palladium-Mediated Carbon
Carbon Bond-Forming Reactions: Highly Efficient Synthesis and Screening
of Stilbene Arrays
Reto Frei and Helen E. Blackwell*[a]
The use of small molecule probes to interrogate biological
phenomena has become a prominent research approach in
chemical biology.[1,2] As the demand for probe molecules
continues to grow, the need to develop new methodology
for the rapid generation and evaluation of small molecules
has also become more urgent.[3,4] Our laboratory is engaged
in the development of the small molecule macroarray as
tool for the rapid parallel synthesis and biological screening
of drug-like molecules and peptidomimetics.[5–8] Macroarrays
Figure 1. A) Image of
a representative small molecule macroarray
are readily constructed via the SPOT-synthesis technique on
planar cellulose supports (i.e., laboratory filter paper)[9]
yielding spatially addressed libraries of 10–1000 unique com-
pounds (ꢁ50–200 nmol of compound/spot; Figure 1A). The
simplicity of macroarray construction, coupled to its com-
patibility with a range of biological assay formats performed
either on or off of the planar support, renders this library
synthesis technique attractive for a range of research appli-
cations. To find wider usage, however, the scope of chemical
reactions that are compatible with the macroarray platform
must be extended, and this has motivated considerable
recent research in this area.[10]
(scale=12ꢀ8 cm). B) The natural product resveratrol (1) and a general
retrosynthesis of stilbenes via the Heck reaction.
of these arrays post-synthesis revealed several stilbenes with
novel activities. This work represents both a chemical and
biological advance for the small molecule macroarray tech-
nique, and further underscores the utility of this approach
for chemical probe discovery.
We selected the Heck reaction for examination on planar
cellulose as this transformation has been well-studied in so-
lution[15,16] and, to a lesser extent, on traditional solid sup-
ports.[17] As our laboratory has an on-going interest in the
development of chemical probes to study virulence path-
ways in bacteria (e.g., quorum sensing (QS)[18–20]), we chose
a chemical scaffold for macroarray design that i) was readily
accessible via the Heck reaction and ii) has been previously
shown to modulate QS pathways. The phytoalexin resvera-
trol (1; Figure 1B) was recently shown to inhibit QS recep-
tors in Gram-negative bacteria (i.e., Proteus mirabilis,
Vibrio fischeri, and Pseudomonas aeruginosa); however, its
mechanisms of action remain unknown.[21–23] Resveratrol (1),
along with other naturally occurring stilbenes, exhibits a rich
range of additional biological activities, spanning from anti-
cancer and cardioprotective effects to life prolongation ef-
fects.[24–26] As such, the synthesis and screening of stilbenes
could provide valuable new chemical tools to examine not
only QS and virulence, but also a range of other pertinent
biological pathways. These attributes, coupled with the
ready applicability of Heck chemistry to stilbene synthesis
To date, the majority of reactions performed on planar
cellulose supports have been simple, acylation-type reac-
tions.[11–13] In view of the prevalence of metal-mediated,
ꢀ
carbon carbon bond forming reactions in both chemical
probe and pharmaceutical discovery,[14] we sought to explore
the feasibility of these reactions on the small molecule mac-
roarray. Herein, we report the application of palladium-
mediated Heck reactions for the efficient construction of
stilbene macroarrays. Preliminary bacteriological screening
[a] R. Frei, Prof. Dr. H. E. Blackwell
Department of Chemistry
University of Wisconsin-Madison
1101 University Ave., Madison, WI 53706-1322 (USA)
Fax : (+1)608 265-4534
E-mail: blackwell@chem.wisc.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903445.
2692
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2692 – 2695

COMMUNICATION
(Figure 1B), caused us to select
stilbenes as our scaffold for this
study.
Macroarray
commenced with the selection
of robust linker system
construction
a
(Scheme 1). We chose the
Rink-amide linker, as this
linker has been shown to be
compatible with
a range of
solid-phase syntheses.[10,27] Prior
to the attachment of the linker
to planar cellulose, we derivat-
ized the cellulose (Whatman 1
CHR filter paper, 2) with a
flexible diamine spacer unit 3
to yield support 4. The spacer
was installed by first converting
the primary hydroxyl groups of
Scheme 1. Initial planar cellulose support modification and Rink linker installation to generate support 6.
cellulose membrane 2 to tosyl groups at 258C, then reacting
this tosylated support with the neat diamine 3 at 708C.
Next, the Rink-amide linker (5, 4-[(2,4-dimethoxyphenyl)-
(Fmoc-amino)methyl]phenoxyacetic acid) was attached to
ly-addressed cleavage of 10a,b using TFA vapor gave trans-
stilbenes 11a,b in excellent conversions and purities (ꢁ97
and ꢁ93%, respectively; as determined by HPLC analy-
sis).[28] We found that the Heck reaction protocol did not re-
quire a tertiary phosphine ligand,[29,30] further streamlining
support
4 via a standard N,N’-diisopropylcarbodiimide
(DIC) coupling reaction at 708C. Notably, we applied linker
5 to support 4 in an arrayed “spot” format using a micropip-
ette (0.3 cm2 per spot) to minimize the use of this relatively
expensive reagent. Following linker loading, the membrane
was acetylated (to cap any unreacted spacer amines), and
subjected to basic Fmoc deprotection to generate amine
support 6. Using this method, Rink linker loadings of
ꢁ550 nmolcmꢀ2 were routinely generated (as determined by
UV Fmoc quantification) in 75 min reaction times (2!6).
To generate stilbenes on support 6 via the Heck reaction,
we required an aryl alkene component that could be directly
attached to the Rink linker. para-Vinyl and meta-vinyl ben-
zoic acids (7a,b) met this criterion as our initial building
blocks (Scheme 2), and were ef-
this synthesis method. PdACTHUNGTRENNU(G OAc)2 was used in an excess of *
1.4 equivalents with respect to immobilized styrenes 8a,b,
suggesting a non-catalytic process. We note, however, that
the small scale of macroarray synthesis (nanomolar) does
not demand a catalytic process. Increasing the palladium
loading led to lowered stilbene purities, while lowering the
loading resulted in poorer stilbene conversions. Importantly,
palladium contamination of the stilbenes products was
found to be minimal (ppb level; see Supporting Informa-
tion).
Using the optimized reaction conditions (Scheme 2), we
prepared a 100-member stilbene macroarray (10a,b) on a
18ꢀ15 cm section of support 6 using the two benzoic acids
ficiently coupled to linker
“spots” on support 6 using spa-
tially addressed DIC reactions
(loading level ꢁ350 nmolcmꢀ2).
We next turned to the aryl
halide component for the Heck
reaction, and examined aryl io-
dides 9 as the second building
block for stilbene macroarray
synthesis. Careful optimization
work with a small set of aryl io-
dides revealed that spotting a
mixture of 3 mol% PdACTHNUTRGNEUNG(OAc)2
(relative to 9) and 1.2 equiv
Na2CO3 (relative to 9) in N-
methyl-2-pyrrolidone
(NMP)
onto styrene support 8a,b led
to the productive formation of
Scheme 2. Styrene loading and Heck reaction to yield stilbene macroarrays 10a,b (stilbenes 11a,b after TFA
stilbene support 10a,b. Spatial- cleavage).
Chem. Eur. J. 2010, 16, 2692 – 2695
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2693

R. Frei and H. E. Blackwell
7a,b and 50 structurally divergent aryl-iodides (9; see Sup-
porting Information for structures) as library building
blocks. The stilbene products (11a,b) were isolated in
ꢁ100 nmol per spot quantities, which was more than suffi-
cient for analytical testing and numerous biological assays.[5]
A subset of the library (20 randomly selected compounds)
was evaluated by HPLC, and high product conversions and
purities were obtained in almost all cases (Table 1).[31] These
results are significant, as they represent to our knowledge
the first demonstration of metal-mediated coupling reactions
in small molecule macroarray synthesis. Notably, we were
able to prepare the 100 stilbenes (11a,b) in under 12 h start-
ing from unfunctionalized cellulose 2 (including all reac-
tions, washing, drying, cleavage, and compound elution), fur-
ther highlighting the efficiency of this synthesis method for
the generation of libraries of biologically relevant mole-
cules.
Figure 2. Luminescence (LuxR) inhibition data for the five most potent
stilbenes in library 11a,b (1:1 against OHHL; 5 mm compound). Two of
the most active stilbenes are shown. Strain: V. fischeri ES114 (D-luxI).
Neg. control=DMSO. Pos. control=OHHL. Assays performed in tripli-
cate (see Supp. Info. for assay protocol and full stilbene structures).
V. fischeri mutant strain comparable to some of the most
potent known native and non-native LuxR antagonists (e.g.,
N-(3-oxo-dodecanoyl)-l-homoserine lactone (OdDHL);
Figure 2).[19] Indeed, these stilbenes were capable of inhibit-
ing LuxR by ꢂ 80% at a 1:1 ratio against OHHL. Addi-
tional experiments are required to elucidate the mechanism
by which these stilbenes elicit their LuxR inhibitory effects.
Nevertheless, these screening data suggest that stilbenes
with structures beyond the resveratrol (1) scaffold are capa-
ble of modulating QS responses. Few compounds with struc-
tures dissimilar to native QS signals have been shown to in-
terfere with QS;[18,20] therefore, the discovery of these active
ligands from such a small library of compounds (100-
member) is noteworthy.
Table 1. Representative purity and conversion data for the stilbene mac-
roarray products 11a,b.
Stilbene R1
R2
R3
R4
R5
Conversion
[%][a]
Purity
[%][b]
11a-4
H
H
H
H
H
H
NO2
H
NO2
CN
H
H
F
F
CF3
H
NO2
H
CN
H
H
F
F
NO2
CF3
CH2CN
H
CN
H
H
CH3
H
H
CH3
H
CH2CN
H
CN
H
CH3
H
H
CH3
H
H
H
NO2
H
H
H
H
H
H
H
99
95
98
99
92
94
97
98
98
96
93
88
87
88
92
94
99
97
96
93
88
92
93
94
11a-15
11a-22
11a-26
11a-27
11a-38 NO2
11a-41 CH3
11a-44
11a-48
11a-50
11b-4
11b-15
11b-22
11b-26
11b-27 NO2
11b-38 CH3
11b-41
11b-44
11b-48
11b-50
CH3 96
H
H
H
H
H
H
H
H
H
99
97
95
91
88
>99
99
NO2
NO2
H
H
H
NO2
H
H
H
H
H
H
H
H
H
In summary, we have demonstrated that palladium-medi-
>99
ꢀ
CH3 98
ated, carbon carbon bond forming reactions are compatible
H
H
H
H
H
H
H
>99
99
99
93
>99
93
with the small molecule macroarray approach. As a test
case, a 100-member macroarray of unique stilbenes was con-
structed via spatially addressed Heck reactions in high con-
versions and purities. Subsequent biological testing of the
stilbene library (11a,b) in a bacterial QS assay revealed sev-
eral new and potent stilbene-derived inhibitors of a QS re-
ceptor in the Gram-negative bacterium, V. fischeri. Overall,
this work serves to highlight the emerging potential of the
small molecule macroarray platform for chemical biology
research. Ongoing work is directed at further extending the
scope of metal-mediated couplings on planar cellulose sup-
ports, as well as a thorough biological evaluation of the stil-
bene library, and will be reported in due course.
NO2
NO2
H
H
H
H
H
H
H
[a] Calculated by comparing initial styrene loadings and the quantity of
styrene remaining after the Heck reaction using UV calibration curves;
see the Supporting Information. [b] HPLC UV detection at 278 nm (for
stilbenes 11a) and 254 nm (for stilbenes 11b).
We next sought to evaluate the activity of the stilbene li-
brary (11a,b) as QS inhibitors, and performed a competitive
antagonism screen against the LuxR receptor in the bacteri-
al symbiont V. fischeri.[19] V. fischeri contains the canonical
QS system in Gram-negative bacteria based on LuxR-type
receptors, and thus represented an excellent model organism
for this initial study.[18,20] We utilized a V. fischeri mutant
strain that reported LuxR inhibition via luminescence, and
examined 5 mm stilbene against 5 mm of V. fischeriꢀs native
QS ligand, N-(3-oxo-hexanoyl)-l-homoserine lactone
(OHHL; Figure 2).
Experimental Section
Representative Heck reaction protocol on styrene-derivatized cellulose
support 8a,b: Spotting solutions were prepared by dissolving aryl iodides
(9, 0.25 mmol), PdACHTNUTRGENNUG(OAc)2 (50 mL of a 0.166m PdACHTUNTGERN(NUGN OAc)2 solution in
NMP), and Na2CO3 (31.8 mg, 0.30 mmol) in NMP (50 mL). Aliquots
(3.0 mL) of this solution were pipetted onto the 100 spots of styrene-func-
tionalized support 8a,b. The support was heated at 708C for 12 min on a
sand bath and then washed by immersion and swirling in 150 mL portions
Preliminary assays revealed five stilbenes (11a-10, 11a-12,
11a-20, 11a-37, and 11b-9) with inhibitory activities in the
2694
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2692 – 2695

ꢀ
Carbon Carbon Bond-Forming Reactions
COMMUNICATION
of DMF (2ꢀ), H2O (2ꢀ), MeOH (2ꢀ), hexanes (2ꢀ) and CH2Cl2 (2ꢀ).
Thereafter, stilbene macroarray 10a,b was dried on a sand bath at 708C
for 6 min prior to TFA vapor cleavage (60 min, 258C) to yield cleaved
stilbenes 11a,b. See Supporting Information for full synthetic details.
[11] T. Ast, N. Heine, H. Wenschuh, Tetrahedron Lett. 1999, 40, 4317.
[12] L. Jobron, G. Hummel, Angew. Chem. 2000, 112, 1704; Angew.
Chem. Int. Ed. 2000, 39, 1621.
[13] D. Scharn, L. Germeroth, H. Wenschuh, J. Org. Chem. 2001, 66, 507.
[14] K. C. Nicolaou, P. G. Bulger, D. Sarlah, Angew. Chem. 2005, 117,
4516; Angew. Chem. Int. Ed. 2005, 44, 4442.
[15] A. De Meijere, F. E. Meyer, Angew. Chem. 1994, 106, 2524; Angew.
Chem. Int. Ed. Engl. 1994, 33, 2473.
[16] E. Negishi, Handbook of Organopalladium Chemistry for Organic
Synthesis, Wiley InterScience, New York, 2002.
[17] S. A. Testero, E. G. Mata, J. Comb. Chem. 2008, 10, 487.
[18] B. L. Bassler, R. Losick, Cell 2006, 125, 237.
[19] G. D. Geske, J. C. OꢂNeill, D. M. Miller, M. E. Mattmann, H. E.
Blackwell, J. Am. Chem. Soc. 2007, 129, 13613.
Acknowledgements
Financial support for this work was provided by the NIH (AI063326),
Burroughs Welcome Fund, Research Corporation, and Johnson & John-
son. H.E.B. is an Alfred P. Sloan Foundation Fellow. We acknowledge
Prof. Ned Ruby (UW-Madison) for the bacterial reporter strain, and Dr.
Andrew Palmer for technical assistance with assays.
[20] G. D. Geske, J. C. OꢂNeill, H. E. Blackwell, Chem. Soc. Rev. 2008,
37, 1432.
[21] W. B. Wang, H. C. Lai, P. R. Hsueh, R. Y. Chiou, B. L. Shwu, S. J.
Liaw, J. Med. Microbiol. 2006, 55, 1313.
[22] L. Fulghesu, C. Giallorenzo, D. Savoia, J. Chemother. 2007, 19, 388.
[23] G. Brackman, U. Hillaert, S. Van Calenbergh, H. J. Nelis, T. Coenye,
Res. Microbiol. 2009, 160, 144.
Keywords: combinatorial chemistry
quorum sensing · small molecule macroarray · stilbenes
· Heck reaction ·
[24] M. Jang, L. Cai, Science 1997, 275, 218.
[1] D. J. Payne, M. N. Gwynn, D. J. Holmes, D. L. Pompliano, Nat. Rev.
Drug. Discovery 2007, 6, 29.
[2] S. L. Schreiber, Nature 2009, 457, 153.
[3] D. Morton, S. Leach, C. Cordier, S. Warriner, A. Nelson, Angew.
Chem. 2009, 121, 110; Angew. Chem. Int. Ed. 2009, 48, 104.
[4] M. A. Koch, L.-O. Wittenberg, S. Basu, D. A. Jeyaraj, E. Gourzouli-
dou, K. Reinecke, A. Odermatt, H. Waldmann, Proc. Natl. Acad.
Sci. USA 2004, 101, 16721.
[5] M. D. Bowman, J. C. OꢂNeill, J. R. Stringer, H. E. Blackwell, Chem.
Biol. 2007, 14, 351.
[6] Q. Lin, H. E. Blackwell, Chem. Commun. 2006, 2884.
[7] M. D. Bowmann, M. M. Jacobsen, H. E. Blackwell, Org. Lett. 2006,
8, 1645.
[25] J. M. Denu, Curr. Opin. Chem. Biol. 2005, 9, 431.
[26] J. A. Baur, D. A. Sinclair, Nat. Rev. Drug. Discovery 2006, 5, 493.
[27] H. Rink, Tetrahedron Lett. 1987, 28, 3787.
[28] See the Supporting Information for stereochemical analysis of the
stilbene products (11a,b).
[29] G. T. Crisp, Chem. Soc. Rev. 1998, 27, 427.
[30] M. T. Reetz, E. Westermann, Angew. Chem. 2000, 112, 170; Angew.
Chem. Int. Ed. 2000, 39, 165.
[31] The major impurity in these samples was the cleaved, unreacted sty-
rene. These styrenes were subsequently found to be inactive in our
biological assays (see the Supporting Information).
[8] L. Hu, K. Maurer, K. D. Moeller, Org. Lett. 2009, 11, 1273.
[9] R. Frank, Tetrahedron 1992, 48, 9217.
Received: December 16, 2009
[10] H. E. Blackwell, Curr. Opin. Chem. Biol. 2006, 10, 203.
Published online: February 4, 2010
Chem. Eur. J. 2010, 16, 2692 – 2695
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2695