CONTENTS:

INTRODUCTION
GENERAL PROPERTIES OF THE FLUORESCENT AIDA-DYES
SPECTROSCOPIC FEATURES OF AIDA-DYES

GENERAL SYNTHESIS

SPECIFIC EXAMPLE: AIDA-1

SYNTHESIS OF AIDA-LABELLED LIBRARIES

CONCLUSION

EXPERIMENTAL PROCEDURE

ACKNOWLEDGEMENTS

REFERENCES
 
 

INTRODUCTION

High throughput screening has become a discipline assimilating synthetic chemistry, biochemistry, biophysics and molecular biology combined with miniaturized detection/liquid handling technologies and automation processes. Additionally, fluorescence spectroscopy has evolved to a routinely used ultra-sensitive detection technology on modern screening platforms. An integration of combinatorial chemistry and fluorescence spectroscopy creates a powerful tool to probe biological targets.

The integration can be done on the chemical synthesis level by providing generically labelled compound libraries. Especially, labelled target molecules of unknown function (orphan targets) can be tested for their binding affinity. This aspect forms the clamp between combinatorial chemistry and functional genomics.

We report on the chemistry and optical spectroscopy of new fluorophores for application in combinatorial chemistry and high throughput screening in solution and on the solid support. The chemistry of these fluorophores is generally described as AIDA-chemistry (An Indazole based Discovery and Analysis Tool).¹

GENERAL PROPERTIES OF THE FLUORESCENT AIDA-DYES

 
 AIDA can accept fluorescence resonance energy from natural protein fluorescence

AIDA can donate fluorescence resonance energy to labelled targets

AIDA belongs to a class of chemically very stable fluorophores

AIDA can be immobilised on solid support

AIDA on solid support can serve as starting point for combinatorial chemistry

AIDA is a generic fluorescence label of compound libraries

SPECTROSCOPIC FEATURES OF AIDA-DYES

To prove the spectroscopic responses of AIDA-conjugates, we have synthesised AIDA-biotin conjugates in solution as well as on solid support. The biotin was attached to the core of the chromophor by different spacers and on different positions of the phenyl substituents.

The excitation spectra of derivatives 1 and 2 revealed a nearly perfect overlap of their absorption bands with the emission of tryptophan. This feature of AIDA establishes the physical basis to act as acceptor for natural protein fluorescence (figures 1a-b). The corresponding emission spectra of AIDA-conjugates 1 and 2 revealed overlap integrals with the absorption spectra of commercial fluorescence labels (e.g. BODIPY-FL™, figures 1c-d) establishing the donor properties of AIDA.

Figure 1.

(1a)-(1d): Excitation and emission spectra of AIDA-biotin conjugates 1 and 2 (in the graphical representations assigned as BLI and IPB, respectively), and avidin-tryptophan (1a-b) and avidin-BODIPY-FL™ (1c-d), showing the spectral overlap between tryptophan, BODIPY-FL™ emission and excitation, respectively.

Donor and Acceptor Properties of AIDA:

The avidin - biotin high affinity interaction was used to probe the donor and acceptor qualities of AIDA-conjugate 1 (BLI) (figures 2 and 3).

(a) AIDA as an acceptor of natural protein fluorescence (Figure 2)

Figure 2.

Figure 2 shows the fluorescence excitation spectra of 50 nM BLI (curve 3) and complexed to 1 µM unlabeled avidin (curve 4). An emission wavelength at 480 nm was chosen because the tryptophan fluorescence of avidin had almost reached the baseline level at this emission energy, resulting in only minor avidin fluorescence contribution between 250 and 370 nm (curve 1). The difference spectrum: [I(BLI (free) )+ I(Avidin (free) )] - I(BLI-Avidin complex ), shown in black, indicates that the BLI fluorescence is quenched by about 25% by complexation in the avidin binding site. This quenching effect can be attributed to all radiative and non-radiative processes occurring to the BLI in the vicinity of the tryptophans in the binding cleft. From 250 to 305 nm the enhancement of the tryptophan and tyrosine fluorescence intensity is 70% based on the sum of BLI and avidin (curve 5).

Conclusion:

When BLI is used as acceptor and protein fluorescence in avidin as donor, both an energy transfer to, and a quenching effect on BLI can be used as a signal for the complex formation.

(b) AIDA as a donor to BODIPY-FL™ labelled avidin (Figure 3):

Figure 3.

Figure 3 shows the fluorescence emission spectra of 50 nM BLI (1, curve 1), 1 µM avidin-BODIPY™ (curve 3), the complex (curve 4) - and the relevant difference spectra (curve 5). In these experiments BLI was used as donor and BODIPY™ as acceptor. Two excitation wavelengths, 338 nm (maximum) and 325 nm were used to test for BODIPY™ excitation at the blue edge of the BLI (1) absorption spectrum. At both excitation wavelengths, the BLI emission between 360 nm and 542 nm is strongly quenched by 78% in the complexed form. 8% of the emission energy of BLI (1) is transferred to BODIPY-FL™.

Conclusion:

The energy transfer from AIDA-biotin conjugate 1 (BLI) to avidin-BODIPY™ is weak. But the quenching effect of almost 80% on BLI between 360 nm and 516 nm is an extremely efficient signal for biotin interaction with avidin monitored by BLI at wavelengths between 440 nm and 500 nm.

GENERAL SYNTHESIS:

Among the known possibilities to synthesise the indazole heterocycle² and more specifically 1,3-diaryl-1H-indazoles, the methods described by Gladstone et al.³ enable an efficient access to the desired compounds.

The synthesis starts from substituted benzophenones 3 which are first converted to arylhydrazones 4 by reaction with substituted arylhydrazines in methanol. The corresponding benzophenone arylhydrazones 4 are transformed to 1-(arylazo)-1,1-diarylmethyl acetic esters 5 by oxidation with lead tetraacetate, following a procedure first described by Iffland.⁴

The arylazo-acetates 5 finally undergo ring-closure to indazoles 6 upon treatment with Lewis acids (e.g. boron trifluoride ethyl etherate) and proceeds via an intermediate 1-aza-2-azoniaallene.⁵ The reaction is highly regioselective for the geminal arylsubstituents of the arylazo-acetates 5. Cyclisation does not occur into aromatic rings bearing strong electron withdrawing substituents, such as carboxyl-, ester-, amide-, or nitro groups, respectively.

The synthesis of such derivatives (see table for selected examples) can be carried out without purification by chromatography and gives the desired compounds in moderate to good yields. The final products are chemically very stable under strong acidic and basic conditions.

Additional derivatives are accessible by aromatic electrophilic substitution (e.g. nitration) on the position 5 of the indazole core or by further modification of an already existing functional group (e.g. reduction of a nitro group).

Functionalized 1,3-diarylsubstituted indazoles can also be synthesised on solid support starting with immobilised arylhydrazones of appropriate benzophenone derivatives.⁶

SPECIFIC EXAMPLE: AIDA-1

For loading of solid supports with AIDA-labels and further transformation to the labelled conjugates, bifunctional indazoles are converted to protected amino acids. Compound 6a is synthesised from benzoyl benzoic acid methylester and 4-hydrazino benzoic acid according to the general procedure. Subsequent treatment with diaminopropane followed by a Fmoc-protection gives compound 8 (AIDA-1) which can then be coupled to an amino resin (preferred: TentaGel) bearing an acid labile or a photocleavable linker.

The diaminopropane moiety acts as a spacer providing a protected functional group to start the synthesis of combinatorial compound libraries.

SYNTHESIS OF AIDA-LABELLED LIBRARIES

For the synthesis of AIDA-labelled on-bead-libraries, we follow the split-and-mix strategy⁷ which is particularly adapted for the production of very large numbers of compounds.

The synthesized conjugates (one-bead-one-compound) have the following general structure:

CONCLUSION

A generic fluoresence labelling tool (AIDA) for combinatorial compound libraries has been designed. A versatile synthesis of a labelling reagent to prepare resins for library synthesis has been created (AIDA-1 as a Fmoc-protected amino acid derivative).

Up to now, several AIDA-1-labelled libraries have been synthesized. The libraries were screened on-bead against various targets. Subsequently, the spectroscopic features of the fluorescent label on the compound were used to validate the primary on-bead binding of target molecules in solution. The publication of the results is under preparation.

EXPERIMENTAL PROCEDURE

Preparation of AIDA-1 loaded resin (TentaGel Rink Amide) for library synthesis

4-{3-[4-(methoxycarbonyl)-phenyl]-1H-indazol-1yl}-benzoic Acid (6a): was synthesised according to the literature^3,4

4-{3-{4-[(3-Aminopropyl)-aminocarbonyl]-phenyl}-1H-indazol-1yl}-benzoic Acid (7): 1,3-diamino-propane (80 mL) is added to 6a (10.0 g, 26.85 mmol). The heterogeneous reaction mixture is heated to 50 °C under stirring for 4h. The 1,3-diamino-propane is distilled off under reduced pressure giving a viscous oil. The minimum amount of methanol is added to the crude product and the desired compound precipitates. The crystals are collected by vacuum filtration. The white crystals are washed with a portion of cold methanol and then with several portions of diethyl ether. The product is dried under reduced pressure, yielding 9.86 g (89 %) of 7.

(4-{3-{4-[N-3-[(9H-Fluoren-9-yl)-methyloxycarbonylamino]-propyl]-aminocarbonyl}-phenyl-1H-indazol-1-yl} benzoic Acid (8): Potassium carbonate (2.67 g; 15.45 mmol) is added to a suspension of the amino acid 7 (3.20 g; 7.73 mmol) in water (60 mL) and 1,4-dioxane (35 mL). The mixture is stirred for 5 min and cooled to 0 ^oC in an ice bath. Fmoc-chloride (2.20 g; 8.5 mmol) in 1,4-dioxane (35 mL) is added dropwise within 10 min. Cooling is removed and stirring continued at room temperature for 12 h. A bulky precipitate is formed. The aqueous suspension is extracted with diethyl ether. The pH is adjusted to 1 with diluted HCl and then further diluted with brine (100 mL). The precipitate is collected by vacuum filtration and washed extensively with diethyl ether. The product is dried under reduced pressure, yielding 4.14 g (84 %) of 8.

Coupling of the fluorescent dye reagent (8) onto the resin: Deprotected TentaGel Rink amide resin is acylated with a 0.1 M solution of 8 (2 eq), 1-hydroxybenzotriazole (2 eq.), N,N´-diisopropyl carbodiimide (4 eq.) in dimethylformamide overnight at room temperature. The resin is washed thoroughly with several portions of dimethylformamide, then methanol, and finally dichloromethane.

ACKNOWLEDGEMENTS

We thank the Andrea Steck's group (Novartis Vienna) for structure confirmations by NMR-spectroscopy and Francis Roll (Novartis Basel) for analysis by HRMS. We also thank Rocco Falchetto (Novartis Basel) for his great efforts in decoding of single beads by µHPLC/MS.

REFERENCES

1. Auer M. and Gstach H. Fluorescent dyes (AIDA) for solid phase and solution phase screening. PCT Int. Appl. (2000), WO 0037448. Priority: US 98-217795 19981221.

2. for a review see: Elderfield, R. C. in "Heterocyclic Compounds" ed. Elderfield, R.C., vol 5, John Wiley & Sons, Inc., New York, 1957, p. 162.

3. Gladstone, W. A. F. and Norman, R. O. C. J. Chem. Soc. (1965), 3048-52; Gladstone, W. A. F. and Norman, R. O. C. J. Chem. Soc. (1965), 5177-82; Gladstone, W. A. F. et al. J. Chem. Soc. (1966), 1781-4.

4. Iffland, D. C. et al. J. Am. Chem. Soc. (1961) 83, 747.

5. Wang, Q. et al. Synthesis (1992), 7, 710-8.

6. Yan, B. and Gstach, H. Tetrahedron Lett. (1996), 37(46), 8325-8.

7. Furka, Arpad. The "Portioning-Mixing" (split-mix) synthesis. Proc. ECSOC-1: First Int. Electron. Conf. Synth. Org. Chem.; Proc. ECSOC-2: Second Int. Electron. Conf. Synth. Org. Chem. (1999).