Development of an Engineered Therapeutic E. coli Nissle for the Treatment of Phenylketonuria

Phenylketonuria is a monogenic disease characterized by the inability to metabolize phenylalanine (Phe), which can result in neurotoxicity. Standard treatment for PKU comprises severe dietary restriction of protein to control plasma Phe levels. There is an unmet medical need for an orally administered therapeutic that can metabolize dietary Phe in the gastrointestinal tract and reduce plasma Phe. To address this problem, we engineered a strain of Escherichia coli Nissle 1917, termed SYNB1618, to produce two distinct Phe metabolizing enzymes: phenylalanine ammonia lyase (PAL) and L-amino acid deaminase (LAAD). We also developed an in vitro simulation (IVS) platform that recapitulates physiological parameters of the human upper gastrointestinal tract. Under IVS conditions, SYNB1618 consumed Phe over a period of several hours, consistent with human upper gastrointestinal transit. The product of PAL, trans-cinnamic acid (TCA) is quantitatively converted to hippuric acid (HA) in vivo and excreted in the urine, enabling the use of HA as a predictive biomarker of SYNB1618 activity. In non-human primates, administration of SYNB1618 resulted in dose-dependent increases in serum TCA and urinary HA. Next, we constructed a mechanistic in silico model considering human upper gastrointestinal physiology as well as IVS characterization of SYNB1618. Monte Carlo simulations were used to model SYNB1618 dose escalation in humans, corresponding to a dose-dependent increase in predicted Phe consumption. A Phase 1 clinical study revealed that orally-administered SYNB1618 was safe and generally well tolerated by healthy volunteers and PKU patients. In addition, dose-dependent recovery of TCA and HA was observed with SYNB1618 administration in plasma and urine, respectively. Notably, urinary HA recovery in humans closely mirrored predictions made by our in silico model. These studies indicate that SYNB1618 can impact Phe metabolism in a dose-dependent manner in preclinical models and humans, supporting the potential for the treatment of metabolic disorders with engineered bacteria.