Bacteria are workhorses of recombinant protein technology and were the production vessel for the first ever FDA-approved biopharmaceutical, Humulin, a recombinant human insulin. Synthetic biology is currently putting bacteria to work in a new, if seemly contradictory, direction—as the therapeutic agent itself. Although intentionally infecting patients with bacteria seems counter-intuitive, the realization that the host microbiome can have profound effects on human health and disease has stimulated interest in leveraging the native microbiome to manage diseases.
Now, synthetic biology is taking this approach a step further by introducing genetically engineered bacteria into the host to treat a wide spectrum of illnesses, from inflammatory disorders and obesity, to infectious diseases, and even cancer. Presently, many studies have demonstrated efficacy in preclinical animal models. For instance, an Escherichia coli Nissle 1917 strain was genetically modified to secrete N-acylphosphatidylethanolamines (NAPEs), which are N-acylethanolamide (NAE) precursors. NAEs are a family of lipids that are synthesized in the small intestine after feeding; they reduce food intake and therefore obesity. Mice that were administered NAPE-producing E. coli in their drinking water ate far less and displayed decreased adiposity and lower insulin resistance in comparison to mice treated with control microbes. Although still in the development phase, NAPE-producing E. coli could eventually help obese people lose weight.
In another study, the same E. coli Nissle 1917 strain was genetically manipulated to sense and kill Pseudomonas aeruginosa, a multidrug resistant human pathogen. The ‘Sense-Kill’ E. coli detects P. aeruginosa by quorum sensing of N-acyl homoserine lactone, a specific P. aeruginosa product, and responds by producing pyocin S5, a P. aeruginosa-killing toxin, and lysin E7, a molecule that elicits E. coli lysis. The lysed E. coli release the pyocin S5, killing P. aeruginosa; essentially, the ‘Sense-Kill’ self-destruct as they deliver their payload. The modified E. coli also expressed dispersin B (DspB), an enzyme that destabilizes mature biofilms. When administered to a mouse model of chronic P. aeruginosa infection, the ‘Sense-Kill’ E. coli significantly lowered the P. aeruginosa load. The ‘Sense-Kill’ E. coli could be envisioned as a method to treat drug-resistant P. aeruginosa infections in human patients, and the method is conceptually extendable to fight other human pathogens such as C. difficile.
These examples demonstrate the potential of genetically modified microorganisms to treat diseases. Although most are still in the preclinical stages of development, several have made it into early phase clinical trials to treat a range of conditions, including metabolic disorders, autoimmune disorders, oral mucositis, and bacterial vaginosis.
Hereditary phenylketonuria (PKU) is a metabolic disease caused by autosomal recessive inactivating mutations to phenylalanine hydroxylase (PAH), the enzyme responsible for hydroxylation of phenylalanine to tyrosine. This causes a buildup of dietary phenylalanine to potentially toxic levels, which elicits irreversible mental disability if untreated. Synlogic, a company based in Cambridge, Massachusetts, has designed SYNB1618, a probiotic microorganism that produces a phenylalanine-degrading enzyme, which clears phenylalanine and allows PKU patients to consume more dietary protein. SYNB1618 is currently in phase I/IIa in healthy volunteers and PKU patients. Another agent, SYNB1020, has been engineered to convert ammonia into arginine to treat hyperammonemia, a metabolic condition caused by genetic mutations to the urea cycle or liver damage. SYNB1020 has cleared a phase I trial in healthy volunteers and is presently in phase Ib/IIa trial in cirrhosis patients with hepatic insufficiency and hyperammonemia.
Type I diabetes (T1D) is a type of diabetes mellitus in which the pancreas secretes very little or no insulin at all and is thought to arise from autoimmune destruction of the pancreatic insulin-producing b-cells. Intrexon, a biotechnology company headquartered in Germantown, Maryland, has developed AG019, a Lactococcus lactis bacterium designed to stimulate immune tolerance to T1D. AG019 secretes human proinsulin and interleukin 10 (IL-10), an anti-inflammatory cytokine, into the mucosal lining of gastro-intestinal tissues. AG019 is in phase Ib/IIa trials in recently diagnosed T1D patients with/without teplizumab, a humanized anti-CD3 monoclonal antibody designed to treat or prevent T1D.
Intrexon, in collaboration with Oragenics, is also developing a AG013 mouth wash to treat oral mucositis, a painful inflammation and ulceration of mucosal membranes lining the oral cavity, throat, and esophagus. It is among the most frequent adverse reactions of chemo- and radiotherapy for cancer. AG013 is an L. lactis strain bioengineered to secrete human trefoil factor 1 (TFF1), a protein that stabilizes and protects the mucosa from injury and stimulates healing. AG013 passed a phase I trial of oral mucositis in patients administered induction chemotherapy for head and neck squamous cell carcinomas (HNSCC) and is passing into phase II trial in HNSCC patients receiving concomitant chemoradiation.
Bacterial vaginosis (BV) is characterized by a change from a predominantly Lactobacillus to a polymicrobial vaginal flora. BV is linked to several health issues, such as increased pelvic inflammation, urinary tract infection (UTI), vulnerability to HIV infection, and preterm births. Osel, located in Mountain View, California, has engineered LACTIN-V, a Lactobacillus crispatus strain, that produces molecules that sustain a healthy microbiome. LACTIN-V has completed phase I trial in healthy volunteers, and has progressed to phase IIa and phase IIb trials with antibiotic metronidazole for BV and phase II trial in patients with recurrent UTI.
Despite the potential of engineered bacteria in preclinical studies and early phase clinical trials, there are a number of potential difficulties that could arise. Bacteria are capable of horizontal gene transfer, so the bioengineered microorganism could transfer its synthetic or human genes to native bacteria, or vice versa, gaining genetic material from native microbes. It may not always be possible to foresee the consequences of this process. Furthermore, the genetically manipulated bacterium could colonize the host or be transmitted to other individuals, for example by gaining virulence factors from natural microbes. One strategy to circumvent this and keep the engineered bacterium under control is to incapacitate it by introducing a kill-switch—a vulnerability that prevents it from growing outside the host or without an exogenous agent. However, kill-switches can fail if the engineered organism gains the necessary genes from native bacteria for autonomous survival.
Progress is being made with genetically modified bacteria for treating diseases, with demonstrated short-term safety in early phase trials. Enthusiasm should be tempered with possible complications arising from horizontal gene transfer, host colonization, and transmissibility of the bioengineered bacterium to other people. However, it is anticipated that carefully designed biological containment strategies into the microbe could help mitigate concerns over these potential problems.