The development of antibiotic resistance has been an inevitable problem leading to an increased demand for novel antibacterial drugs. To address this need, we initiated a structure-based drug design program wherein desmethyl analogues (i.e., Methyl to Hydrogen) of the 3rd-generation macrolide antibiotic telithromycin were prepared via chemical synthesis. Our approach will determine the biological functions of the methyl groups present at the C-4, C-8 and C-10 position of the ketolide. These structural modifications were proposed based on the structural data interpreted by Steitz and co-workers after obtaining crystal structures of macrolides erythromycin and telithromycin bound to the 50S ribosomal subunits of H. marismortui. Steitz argued that in bacteria, A2058G mutations confer resistance due to a steric clash of the amino group of guanine 2058 with the C-4 methyl group. In turn, we hypothesize that our desmethyl analogs are predicted to address antibiotic resistance arising from this mutation by relieving the steric clash.

To readily access the analogs, we proposed to synthesize, 4,8,10-tridesmethyl telithromycin, 4,10-didesmethyl telithromycin, 4,8-didesmethyl telithromycin and 4-desmethyl telithromycin as four targeted desmethyl analogs of telithromycin. This thesis includes the total synthesis and biological evaluation of 4,8,10-tridesmethyl telithromycin and 4,10-didesmethyl telithromycin analogs and the progress towards the total synthesis of 4-desmethyl telithromycin analog.

We employed Nozaki-Hiyama-Kishi (NHK) and ring closing metathesis (RCM) reactions as the two macrocyclization methods towards the total synthesis of these analogs. The RCM was superior compared to the NHK macrocyclization where in grams of these macrocycles were accessible. An optimized method for installing the desosamine sugar onto the C-5 alcohol using the Woodward's thiopyrimidine donor was developed. Baker's one-pot carbamoylation/intramolecular aza-Michael method was utilized to install the oxazolidinone side chain of telithromycin. The total synthesis of 4,8,10-tridesmethyl telithromycin required 42 steps overall (31 steps in the longest linear sequence). The analog 4,10-didesmethyl telithromycin was synthesized in 44 steps overall (32 steps in the longest linear sequence). These analogs were able to inhibit bacterial growth, presumably by targeting the bacterial ribosome. In addition, 4,8,10-tridesmethyl telithromycin analog was more potent than telithromycin against an A2058T mutant and 4,10-didesmethyl telithromycin analog was more potent than 4,8,10-tridesmethyl telithromycin against an A2058G mutant.

Also, a concise synthesis of D-desosamine was accomplished in five steps and in 15% overall yield from commercial methyl &agr;-D-glucopyranoside. Other efforts involved the contribution of key intermediates towards the total synthesis of 4,8-didesmethyl telithromycin are described.