A 0097 ] trans-( 9 , 10 )-Dihydro-11-Aminoethanoanthracene-12-Carboxylic Acid ( AMEAC ) , A New Synthon For ß-Peptides

ß-Amino acids and their peptides have most recently found intense interest in the research community [1].Oligomers ("foldamers") of unnatural amino acids have a wide range of potential applications. They may adopt compact, specific conformations. They could be used to form helices, turns and sheets and develop new types of tertiary structures. Certainly short peptides of this type have potential pharmaceutical applications.

Starting from anthracene and maleic anhydride, Diels Alder cycloaddition , hydrolysis and inversion quantitatively leads to the well known [3] trans -9,10-dihydro-9,10-ethanoanthracene-11,12-dicarboxylic acid 1 (EADC, Scheme 1).This C2-symmetric acid can be easily separated into the enantiomers [3,6]. 1 has been used previously to prepare the 11,12-diamino analogue, which has found use as catalyst [4] for asymmetric allylic alkylations and as benzoylated variation for an enantioselective synthesis of carbanucleosides [5].It has also been used by us to prepare a Pirkle-Type chiral stationary phase which especially well separated enantiomers of aryl substituted lactones and cyclic carbamates [6] From racemic 1 we have prepared the racemic title compound in three steps without major problems (Scheme 2).Synthesis of the pure (R,R) and (S,S)-enantiomers proved to be more difficult than expected since partly racemisation occured in the step of alkaline hydrolysis to monoester 3 and also after or during Curtius degradation to the desired amino acid 4. (see Chapter HPLC-analysis below).

HPLC Analysis of Enantiomers
As mentioned above, starting from enantiomer S,S -1, we found significant racemization (45ee) after HPLC analysis of the monomethylester 3. From previous work we used an optimized chiral stationary phase (CSP) which is based on a ULMO analogue, but contains a short urea linker to silica [6].Conditions: 25°C; flow 1.0 ml/min; mobile phase nheptane/2-propanol= 98/2; 0.1% TFA.Under those conditions the separation factor of diacid 1 (resp.monoester3) was 1.42 (1.17), k´2 2.08 (1.27) and resolution 3.96 (1.90).The enantiomers of the bisester appeared as singlet(k´0.74),but that was well separated from the other four signals.Despite the loss of enantiomeric purity, it is important that one can easily deduct that the first eluting enantiomer is the (S,S)-monoester, exactly as it has been observed with the acid.Analysis of the enantiomers of AMEAC could not be directly achieved on our Pirkle columns.However, derivatization with dinitrobenzoylchloride (DNB-Cl) proved to be easy and separation of DNB-AMEAC was achieved using a deaza ULMO CSP published by us as CSP IV in [7].This CSP has previously been shown by us to separate DNBderivatives [8].Using this method, we found that also the Curtius degradation lead to a further loss of enantiomerical purity.The analysis showed only ee 16.This was somewhat disappointing, but, as mentioned above, the absulute configuration of AMEAC could now be easily deduced from the chromatogram since the starting material had (S,S)-configuration and certainly that remained the major component.A semi-preparative HPLC separation loading 0.25 mg DNB-derivative onto the analytical column yielded 0.1 mg enantiopure products with 99.5 for the first and 99ee for the second peak , DNB-(R,R)-AMEAC).The chromatogram below shows the racemate and the pure (R,R)-derivative.

Fmoc-and Boc derivatives
In order to provide building blocks for peptide synthesis, Fmoc and Boc derivatives of AMEAC were prepared using standard conditions.Both crystalline products could be isolated in good yield and the enantiomers separated on chiral stationary phases.Chiral recognition of the Fmoc-derivative on a commercial preparative p-basic Pirkle naphthyl-CSP (250x20) was sufficient to separate 100 mg of the substance in three runs.

Conclusion
We have demonstrated, that Diels-Alder products of anthracene and maleic acid (and fumaric acid derivatives [10]) can be rearranged to give chiral ß-amino acids having a rigid aromatic backbone.The synthesis is streightforward and the yield is acceptable.Separation of enantiomers of AMEAC or the Boc-and Fmoc derivatives is an alternative to the direct synthesis from a enantiopure precursor since partly racemization is there at least a problem which would require careful optimization.As analytical tools a set of Pirkle type chiral stationary phases were found to be optimal, to analyse directly the enantiomers of the precursors as well as the acyl-and carbamoyl derivatives.

Experimental
All compounds and solvents were commercially available and used without further purification.HPLC runs were performed at 25 °C and usually monitored at 254 nm.Solvents used for mobile phases were of HPLC grade (MERCK, Darmstadt, Germany) Instrumentation HPLC measurements were performed using a Hewlett-Packard series HP1050 instrument (consisting of a pumping system, a multiple wavelength detector and an autosampler) and the HPChemstation software.NMR experiments were done on a Bruker 360 MHz instrument in CDCl 3 as the solvent.Semipreparative HPLC-separation of AMEAC as Fmoc-derivative was performed with a Knauer Pump (Germany) WellChrom Maxi Star K-1001 equipped with 50ml pump-head, a Hewlett-Packard 1050 variable wavelength detector with preparative cell.Column : L-Naphthylalanin (REGIS, Morton Grove,Ill,USA); dimension 250 x 21.1mm; 5 micron silica.

Acknowledgement
This paper was designed according to a mask created by Prof. Wolfgang STADLBAUER; some syntheses were performed by Pedro TRAAR and Harald MANG