Scientists Engineer E. coli From Stool Samples To Develop New Drugs

Researchers at the University of California, San Diego, have genetically engineered Escherichia coli (E. coli) bacteria isolated from stool samples to express therapeutic genes of interest. This method provides opportunities to potentially transform bacteria into resident medicine-producing factories within the gut. The results are published in the journal Cell.

Using bacteria to treat disease

The gut microbiome is becoming an increasingly important topic in our understanding of health and disease. Over the last two decades, research has shown that the microbiome can greatly influence the metabolism and metabolic health of the human host.

Now, scientists in the field of synthetic biology are trying to engineer live bacterial therapeutics (LBTs) to treat disease, using native bacteria taken from the gut microbiome. LBTs consist of genetically engineered bacteria developed with the purpose of treating, curing or preventing disease by expressing selected beneficial genes. These biotherapeutics have been tested in vitro and in vivo for the treatment of chronic diseases, such as phenylketonuria – a genetic disease caused by an inability to metabolize the amino acid phenylalanine – by engineering E. coli that produce phenylalanine-metabolizing enzymes in the gut.

“Since bacteria have the ability to sense their environment – and based on what they sense, produce proteins – many labs such as ours are interested in developing recombinant LBTs,” explains Dr. Amir Zarrinpar, assistant professor at the University of California and senior author of the study. “These recombinant LBTs are an exciting area of biomedical research for a number of reasons, but most importantly because they can be used to treat difficult diseases, and in more extreme versions, to produce therapies that are essentially curative,” he continues.

However, progress in this field has been hampered as the bacterial strains used as a “chassis” for LBTs generally fail to survive for long periods in the harsh and competitive environment of the gut, ultimately requiring consistent administration to ensure that they can induce physiological effects.

In contrast, native bacteria from the microbiome already residing in the gut are highly adapted to their environment, having evolved to survive in the tough conditions over thousands of previous generations. As a result, the researchers in the current study set out to develop a technique to develop an LBT by engineering bacteria isolated from stool samples, hypothesizing that native bacteria would make a more effective chassis for LBTs. This essentially allows researchers to “knock-in” specific functions into the microbiome of the host.

Engineering a live bacterial therapeutic

In their study, the researchers first set out to find a suitable bacterial species to use as a chassis for LBT. Zarrinpar explains, “We used native E. coli since there are a lot of tools available to engineer them. Honestly, we weren’t sure it was going to work with E. coli since they are such a small part of the microbiome (<0.1%). But we think that E. coli likely live in especially important areas (i.e., niches) that allow it to interact with the host.”

First, two E. coli strains were engineered using bacteria extracted from a mouse stool sample. The strain EcAZ-2^BSH+ was engineered to express bile salt hydrolase (BSH), a bacterial enzyme that breaks down and deconjugates bile acids produced by the liver to alter host glucose and lipid metabolism. The second stain, EcAZ-2^IL-10+, expresses the mammalian anti-inflammatory cytokine IL-10. These genes of interest were selected due to the implications of BSH on metabolism and diabetes, as well as the anti-inflammatory activity of IL-10 for treating inflammatory bowel diseases such as colitis.

The researchers showed that EcAZ-2^BSH+ and EcAZ-2^IL-10+ were both able to successfully colonize and persist within the gut for the entire lifespan of the mouse – proving this technique to be more effective than the addition of non-native bacteria, which are instead rapidly lost from the gut. Additionally, experiments showed that these bacteria effectively expressed their functional genes of interest, producing BSH and IL-10 in the gut.

Next, researchers investigated whether gene expression from these bacteria could influence the host’s physiological functions. Analyzing BSH activity, they showed that levels of both total and conjugated bile acids were significantly reduced in the EcAZ-2^BSH+-colonized mice. Furthermore, these mice also had reduced blood glucose and insulin levels after a mixed meal challenge (i.e., feeding with a liquid diet). EcAZ-2^BSH+ also improved insulin sensitivity and glucose tolerance in a genetically modified mouse model of Type 2 diabetes, which taken together demonstrates the potential of EcAZ-2^BSH+ as a promising diabetes treatment.

Lastly, the researchers investigated the possibility of translating this technique into humans. After isolating E. coli from human volunteers undergoing endoscopies, Zarrinpar and colleagues showed that they could effectively perform the first step of their LBT engineering protocol using genetic modification. However, the authors state that further work is required to transfer the engineered bacteria back to patients to assess how effectively they might colonize the gut and their potential use as disease treatment.

Targeting other chronic diseases

Overall, the results of this study show that native E. coli can be engineered to effectively express a gene of interest and stably colonize the gut, inducing physiological changes in mice months after the initial administration. Additionally, this technique also has the potential to be effective in human hosts.

When asked about the challenges in developing this technique, Zarrinpar explained: “Engineering undomesticated, native E. coli is not straightforward. They are quite resistant to having their genomes modified. Since the initial experiments outlined in this paper, we have gotten better at improving our approach to engineering these bacteria. But the engineering of these native strains has not been as straightforward as we would like.”

In the future, Zarrinpar and colleagues plan to develop this technology further to target even more diseases. “Honestly, we think we can target a long list of chronic and genetic diseases!” Zarrinpar adds. “If you look at funded grants to our lab (both me and my trainees), you can see that we are using engineered bacteria to affect major diseases/ailments, such as colorectal cancer, inborn errors of metabolism, atherosclerosis, aging, non-alcoholic steatohepatitis (NASH) and, as we showed in the paper, Type 2 diabetes and inflammatory bowel disease. If I had more people in the lab, I would pursue even more diseases.”

Dr. Amir Zarrinpar was speaking to Sarah Whelan, Science Writer for Technology Networks.

Reference: Russell BJ, Brown SD, Siguenza N, et al. Intestinal transgene delivery with native E. coli chassis allows persistent physiological changes. Cell. 2022;0(0). doi: 10.1016/j.cell.2022.06.050