Dependence on petroleum for energy and petrochemical products has led to high energy costs, a polluted environment, the depletion of a finite resource (oil), and the reliance on hostile nations for our energy security. Biofuels promise to alleviate some or all of these issues but remain costly and inefficient to produce, and difficult to integrate into our existing transportation infrastructure. A renewable fuel molecule capable of replacing petroleum for our fuel and chemical industries is desired. Isobutene is an important commercial chemical used for the synthesis of butyl rubber, terephthalic acid, specialty chemicals, and a gasoline performance additive known as alkylate. Currently, isobutene is produced from petroleum and hence is nonrenewable. Here we report that the Saccharomyces cerevisiae mevalonate diphosphate decarboxylase (ScMDD) can convert 3-hydroxy-3-methylbutyrate (3-HMB) to isobutene. Whole cells of Escherichia coli producing ScMDD formed isobutene from 3-HMB at a rate of 154 pmol h^-1 g cells^-1. His₆-ScMDD was purified by nickel affinity chromatography and shown to produce isobutene from 3-HMB at a rate of 1.33 pmol min^-1 mg^-1 protein. In contrast, no isobutene was detected from control cells lacking ScMDD, and controls showed that both His₆-ScMDD and 3-HMB were required for detectable isobutene formation in enzyme assays. ScMDD was subjected to error-prone PCR and 2 improved variants were characterized, ScMDD1 (I145F) and ScMDD2 (R74H). Whole cells of E. coli producing ScMDD1 and ScMDD2 produced isobutene from 3-HMB at rates of 3000 and 5888 pmol h^-1 g cells^-1 which are 19- and 38-fold increases compared to cells producing His₆-ScMDD. Although ScMDD was shown to be amenable to manipulation in order to increase its activity on 3-HMB, we estimate that a 106 fold increase in activity is needed for commercial application. Two novel methods were designed to increase enzyme activity--a mevalonate selection and an isobutene biosensor.

The mevalonate selection yielded one variant with a 3.9-fold increase in isopentenol production from mevalonate over wild-type, to 186.6 nmol min^-1 g cells^-1, as well as a 51.6% increase in isobutene-formation. The isobutene biosensor was built and confirmed to respond to isobutene, toluene, and isoprene with induction ratios of 1.64, 3.58, and 1.3, respectively.