Synthesis of the acromelic analogue

Our early work on kainoid synthesis was conducted in the racemic series while we were still establishing a reliable method for carrying out the asymmetric cyclisation, and we made both 2 and kainic acid 1 in racemic form first, later developing an asymmetric version of the entire synthesis of 1 and of the synthesis of a key intermediate for 2. 

There were several reasons for setting our sights on arylkainoid 254 as our first target. Firstly, being both highly active as a neurotransmitter analogue but also an unnatural product, it is a particularly desirable 

The final steps of the sequence are all downhill the unfortunately low-yielding two-stage reduction of the pyrrolidinone to a pyrrolidine, and deprotection to yield 2 in racemic form. 

Attempts to induce asymmetry in the cyclisation of 57 using the chiral lithium amides, which had worked well with simple benzamides, were frustrated by the inherent chirality, at low temperature, of the naphthamide itself. This feature is common to all 2-substituted tertiary aromatic amides, which may become atropisomeric at low temperature due to slow rotation about the Ar-CO bond.48 We therefore sought to 

---