Living Engineered Bacterial Therapeutics: Emerging Affordable Precision Interventions

ABSTRACT

Live biotherapeutic products (LBPs), including engineered bacteria, are rapidly emerging as potential therapeutic interventions. These innovative therapies can serve as live in situ drug delivery platforms for the direct deposition of therapeutic payloads, including complex biologics, at sites of disease. This approach offers a platform likely to enhance therapeutic efficacy and decrease off‐target side effects. LBPs also can likely be distributed at a relatively low price point, as their production can be economically scaled up. LBPs represent an exciting new means for ensuring healthy lives and promoting well‐being for all ages, aligning with the World Health Organization's sustainable development goal 3.

Smart microbes are rapidly emerging as potential therapeutic interventions. This review is focused on discussing the potential of engineered bacteria for ensuring healthy lives and promoting well‐being for all ages, aligning with the World Health Organization's sustainable development goal 3.

1. Introduction

Various synthetic biology‐based approaches are currently used to engineer Gram‐positive and Gram‐negative bacteria as therapeutics. The bacteria are primarily designed to target diseases in the gut, tumours and skin regions that support the growth of a complex microbiota. For Gram‐positive bacteria, the chassis of choice includes Lactobacillus and Lactococcus species, while Nissle 1917 E. coli is favoured for Gram‐negative bacteria. Each of these species has GRAS (generally regarded as safe) status.

Gram‐positive bacteria have the advantage that protein payloads of interest can be displayed on their surface when fused to the exposed domains of cell wall proteins or secreted into their surroundings via native secretion systems. The complex outer envelope of Gram‐negative bacteria makes similar adaptations more challenging, as two lipid bilayers envelop their cell wall. Nevertheless, the genetic tractability of Gram‐negative bacteria has enabled the development of innovative strategies to display and secrete therapeutic payloads.

Here, we focus on the potential of engineered smart microbes in the treatment of communicable and some inflammatory and metabolic non‐communicable diseases. Information regarding similar efforts in engineering bacteria for the treatment of cancer can be found in several recently published comprehensive reviews (Lynch, Goers, and Lesser 2022; Gurbatri, Arpaia, and Danino 2022; Raman et al. 2023).

2. Engineered Bacteria for the Treatment and Prevention of Infectious Diseases

A major goal of sustainable development goal 3 is to develop interventions that address water‐borne diseases. Diarrhoea is the second leading global cause of death in children under five, with cases most prevalent in low‐ and middle‐income countries (Perin et al. 2022). Most cases are due to infections with enteric bacteria and viruses that are primarily transmitted via contaminated water. Smart microbes with the potential to tackle this challenge are continuing to emerge (Figure 1). While still in the early stages, these engineered bacteria hold great potential to serve as innovative antibiotic‐free interventions for treating infections and suppressing the carriage of multidrug‐resistant (MDR) bacterial pathogens.

FIGURE 1.

Schematic summary of engineered live bacteria for treating and preventing infectious diseases. (A) Bacteria have been engineered to produce therapeutic payloads constitutively or upon sensing environmental signals, that is, quorum sensing molecules or tetrathionate. Payloads include bacteriocins, nanobodies and enzymes, which are released via bacterial secretion systems, transporters and cell lysis systems or displayed on the outer bacterial surface. (B) Engineered bacteria have been outfitted to target specific pathogens, including Pseudomonas aeruginosa , Salmonella, EHEC and Shigella, and to secrete beta‐lactamases that degrade antibiotics.

2.1. Bacteriocin‐Based Interventions

Bacteria are armed with various means to kill competitors, including the production and secretion of bacteriocins. Bacteriocins are ribosomally synthesised secreted bioactive antimicrobial proteins that inhibit the growth of related species (for review, see Heilbronner et al. (2021)). Bacteriocin gene clusters (BGCs) encode the antimicrobial peptides, proteins that enable their biosynthesis and export and immunity proteins that protect the originating strain. In Gram‐negative bacteria, bacteriocins are secreted through the combined action of dedicated ABC transporters located in the inner membrane and porins in the outer membrane. The activity of these expression cassettes is often tightly regulated due to the metabolic burdens associated with their production.

EcN has inherent antimicrobial properties due, in part, to its secretion of microcins, small bacteriocins (Sassone‐Corsi et al. 2016). EcN's microcins are primed for expression and secretion in iron‐limiting conditions like the gut. Several groups have genetically engineered EcN to expand its repertoire and/or means of controlling its expression and secretion of microcins.

For instance, investigators engineered EcN to secrete Microcin 47 (Mcc47) upon sensing tetrathionate, a metabolite produced in the setting of inflammation, a condition associated with gastroenteritis (Palmer et al. 2018). In another case, EcN has been outfitted to constitutively express and secrete high levels of Microcin J25 (McJ25) (Forkus et al. 2017) or Microcin I47 (McI47) (Mortzfeld et al. 2022). While Mcc47 and McJ25 specifically target pathogenic E. coli and closely related species such as Salmonella, McI47 has a broader range, targeting members of the Enterobacteriaceae family, including extended‐spectrum beta‐lactamase (ESBL) E. coli and carbapenem‐resistant Klebsiella pneumoniae . While all three engineered strains can be used to treat or prevent infections with enteric pathogens, the McI47‐secreting variant can potentially eradicate MDR bacteria from the gut microbiota.

Several groups have developed variants of EcN that secrete bacteriocins targeting more distantly related species, including two difficult‐to‐treat clinically relevant organisms, Gram‐positive vancomycin‐resistant Enterococci and Pseudomonas aeruginosa , a non‐enteric Gram‐negative bacterium. While these are not gut pathogens, patients colonised with them are at increased risk of developing difficult‐to‐treat infections at other sites. To outfit EcN to secrete heterologously expressed Gram‐positive bacteriocins, the antimicrobial peptides were modified with an E. coli microcin secretion system and introduced into a variant of EcN engineered to express the corresponding secretion system (Geldart et al. 2018). In another study, EcN was engineered to lyse and release an antimicrobial peptide upon sensing the presence of a P. aeruginosa autoinducer, a quorum‐sensing molecule released when the bacteria establish a replicative niche within the intestines (Hwang et al. 2017). Both platforms have demonstrated promise in partially abolishing gut colonisation in animal models, and each platform can be modified to target a variety of other bacterial pathogens.

2.2. Antibody‐Based Interventions

In parallel with efforts to equip EcN to expand its repertoire of secreted antimicrobial peptides, others are working on equipping EcN and other bacteria with new capabilities, including their ability to display or secrete single domain antibodies (aka nanobodies (Nbs)) that target and inactivate essential virulence determinants. Single‐domain antibodies, commonly referred to as nanobodies (Nbs), are emerging as a new therapeutic modality. They are the variable domains of heavy chain‐only antibodies found in members of the camelid family and sharks. These domains are small proteins (~15 kDa) that often bind with high affinity to their target proteins. While they have historically been derived from immunised members of the camelid family, synthetic Nb libraries are rapidly emerging.

For Gram‐positive bacteria, surface display is achieved by fusing proteins of interest to the exposed extracellular domains of cell wall proteins. One area in which this approach has been demonstrated to be particularly fruitful is in developing variants of Lactobacillus paracasei that display viral‐neutralising Nbs on their outer surface (Pant et al. 2006; Yuki et al. 2022). One of these strains has effectively reduced disease severity and viral load in infected mice (Pant et al. 2006). These observations suggest that such strains have the potential to be developed into an intervention for the prevention and treatment of infections caused by Rotavirus, the leading cause of acute diarrhoea among children < 5 in developing countries for which there are no current treatment options other than supportive care.

For Gram‐negative bacteria, two strategies for Nb surface display are being pursued. In the first, Nbs are fused to the outer membrane anchoring β‐barrel motif of Intimin, a bacterial adhesin. These synthetic adhesins can enhance bacterial interactions with targeted host cells that express the corresponding antigen, particularly when introduced into E. coli engineered to lack its natural adhesins (Piñero‐Lambea et al. 2015).

The second approach is to fuse Nbs to CsgA, the structural unit that forms curli filaments on the surface of E. coli and other related bacteria. Curli are amyloid fibres that generally function to promote biofilm formation. The modified curli filaments create a multivalent pathogen‐binding matrix on the surface of EcN. This platform has not yet been tested in an animal model. However, using cell culture models, the investigators have demonstrated that Curli‐Nb fibres expressed on the surface of EcN bind to and inactivate virulence determinants of three gastrointestinal pathogens, including Shigella, EHEC (Enterohemorrhagic E. coli ) and Cryptosporidium (Gelfat et al. 2022).

A complementary approach has been the development of the PROT₃EcT platform, a suite of commensal and probiotic E. coli engineered with a modified tipless type III secretion system (T3SS) (Lynch et al. 2023). T3SSs are complex nanomachines that many Gram‐negative bacterial pathogens use to transport virulence proteins directly into mammalian cells. By removing the tip complex protein, which keeps this machine in an off position until it contacts host cells, secreted proteins are instead released into the environment. This is a significant modification as Gram‐negative bacteria naturally secrete few proteins in their surroundings.

PROT₃EcT variants stably colonise the intestines of mice and can be engineered to secrete functional nanobodies modified with a type III secretion sequence. An E. coli HS‐based PROT₃EcT variant outfitted to secrete a Nb that blocks the activity of an essential EHEC virulence protein delayed the onset of infection in a mouse model (Srivastava et al. 2024). The PROT₃EcT platform is modular (González‐Prieto, Lynch, and Lesser 2023), allowing for the engineering of additional therapeutic payloads, including those that target additional gastrointestinal pathogens.

Yeast that display an Nb that neutralises Clostridioides difficile toxins on their surface have demonstrated efficacy in prophylactic and therapeutic mouse models of disease (Chen et al. 2020). Together, these studies strongly suggest that the site‐specific delivery of nanobodies to the gut that blocks essential virulent determinants has the potential to be developed as an affordable intervention for the treatment of intestinal infections.

2.3. Emerging Bacterial‐Based Strategies to Prevent and Eradicate Gut Colonisation With Bacterial Pathogens

A variety of additional technologies are evolving to combat gut colonisation with MDR bacteria. Multiple groups are working to develop CRISPR/Cas editing technologies that enable the precise clearance of targeted bacteria from the gut microbiota. Efforts in this area initially focused on engineering phage (Bikard et al. 2014; Citorik, Mimee, and Lu 2014). However, a variant of EcN with an optimised conjugation type IV secretion system was developed recently. The resulting strain, eB‐COP (evolved broth‐conjugative probiotic) equipped with a transferrable plasmid encoding a CRISPR/cas cassette that targets an antibiotic resistance cassette, was found to efficiently promote the clearance of pathogenic E. coli and C. rodentium from the mouse intestinal microbiota (Neil et al. 2021). This platform can likely be extended to develop EcN that can target various other related bacterial species.

A complementary approach to limit gut colonisation with antibiotic‐resistant organisms has also recently been developed. In this case, the strategy is to limit the disruption of the gut microbiota in response to systemically administered broad‐spectrum antibiotics by degrading the antibiotics that reach the gut. Lactococcus lactis was engineered to secrete a beta‐lactamase, an enzyme that inactivates many beta‐lactam antibiotics, including penicillins, cephalosporins and carbapenems (Cubillos‐Ruiz et al. 2022). To avoid issues with the transfer of the beta‐lactamase expression cassette to other bacteria and to ensure the Lactococcus itself does not develop resistance, the system is encoded on two genetically unlinked biosynthesis clusters, which only assemble post‐secretion. In a mouse model, this engineered strain demonstrated protection against the development of C. difficile colitis.

2.4. Advancements in Bacterial‐Based Strategies for Combatting Cholera

Cholera is an acute, severe diarrheal infection caused by Vibrio cholerae that is typically transmitted via contaminated food or water. It is endemic in many low‐income countries and spreads rapidly during conflicts or crises when access to clean water is restricted. Early efforts to combat cholera using live bacterial therapies included engineering EcN to express an autoinducer (CAI‐1), a signal mediator that inhibits virulence gene expression in V. cholerae . Pretreatment with engineered EcN prevented colonisation by V. cholerae and significantly improved survival in an infant mouse model (Duan and March 2010).

While cholera vaccines are currently available, they provide limited protection. Towards developing a new oral cholera vaccine, HaitiV, a strain lacking the majority of Vibrio cholerae's virulence factors, was developed. HaitiV produces a nontoxic subunit of cholera toxin that stimulates adaptive immune protection in mice with a single dose, contrasting with killed OCVs that typically require multiple doses (Fakoya, Sit, and Waldor 2020). Surprisingly, HaitiV colonisation was found to block rabbits from developing an infection when orally administered infectious doses of wild‐type V. cholera (Hubbard et al. 2018), demonstrating that it can act as a probiotic to limit infections. These observations suggest that the distribution of HaitiV in the setting of an impending cholera outbreak can block infection with V. cholerae while an antibody response is generated.

It was also recently discovered that colonisation with native unmodified Lactococcus lactis protects against infection with V. cholerae in an infant mouse model via its production of lactic acid. Capitalising on this observation, the investigators outfitted L. lactis with a gene cassette that triggers the expression of an enzymatic reporter upon detecting V. cholerae quorum‐sensing signals (Mao et al. 2018). This strain can potentially serve as an early warning system for detecting and delaying impending cholera outbreaks.

2.5. Smart Microbes for the Treatment of Non‐Healing Wounds

Chronic non‐healing wounds are a significant cause of morbidity and mortality, particularly in developing countries, where access to healthcare and resources can be limited. Smart microbes for their treatment are emerging, particularly bacteria engineered to secrete CXLC12. CXLC12 is a chemokine that promotes wound healing via the recruitment of macrophages and neutrophils, which favour healing while keeping invading microorganisms at bay. Lactobacilli engineered to secrete CXCL12 accelerated wound healing in a preclinical pig study (Öhnstedt et al. 2022) and a phase 1 clinical trial (Öhnstedt et al. 2023). In another study, healing was enhanced when wounds treated with a CXCL12‐secreting Lactococcus lactis were exposed to LED yellow light, an intervention that stimulates collagen synthesis (Zhao et al. 2021). To address potential issues with colonisation of the dry and nutrient‐poor epidermal environment, another group demonstrated that the application of a synthetic hydrogel encapsulating both CXCL12‐secreting L. lactis and photosynthetic cyanobacteria was also capable of wound healing. In this case, the CXLD12‐secreting L. lactis utilised the sucrose produced by the cyanobacteria, thus enabling their survival within the hydrogel (Li et al. 2023).

3. Engineered Bacteria for the Treatment and Prevention of Non‐Communicable Diseases

3.1. Engineered Bacteria for the Treatment of Inflammatory Bowel Diseases (IBDs)

The incidence of inflammatory bowel diseases, including ulcerative colitis and Crohn's disease, is rising in low‐ and middle‐income countries, where access to state‐of‐the‐art therapies is limited (Rajbhandari et al. 2023). EcN is used as an intervention for the treatment of IBD in Europe and Canada, given its inherent anti‐inflammatory properties. It is as efficacious as an oral agent, mesalamine, in preventing flares (Jia et al. 2018). Using synthetic biology approaches, several groups have developed variants of EcN with enhanced anti‐inflammatory properties (Figure 2).

FIGURE 2.

Schematic summary of engineered live bacteria for treatment of inflammatory bowel diseases (IBDs). (A) Bacteria have been engineered to produce therapeutic payloads constitutively or upon sensing markers of gut inflammation, that is, nitric oxide (NO) and tetrathionate (${\mathrm{S}}_4{\mathrm{O}}_6^{2-}$). Payloads include Trefoil factors (TFFs), anti‐TNFα nanobodies and AvCystatin, which are released via secretion systems or transporters or displayed on the outer bacterial surface. (B) In the inflamed GI environment, variants of these smart microbes are observed to suppress inflammation and/or stimulate epithelial repair.

In one case, EcN was engineered to display trefoil factors, proteins that promote mucosal healing, on its surface by fusing these factors to Curli (Praveschotinunt et al. 2019). In another, EcN was engineered to secrete Nbs that neutralise TNFα, a pro‐inflammatory cytokine, via the PROT₃EcT platform (Lynch et al. 2023). These strategies are adaptations of those described above for the display/secretion of Nbs targeting essential bacterial virulence determinants.

In a complementary approach, EcN was modified to encode heterologous metabolic pathways that enable the production of high levels of 3‐hydroxybutyrate (3HB), a ketone body with anti‐inflammatory properties (Yan et al. 2021). In another, EcN was engineered to secrete epidermal growth factor beta (EGF‐β), another protein that promotes mucosal healing. In this case, secretion was obtained by introducing the lipase ABC transporter from Erwinia chrysanthemi into the chromosome of EcN (Yu et al. 2019). These approaches have demonstrated varying degrees of efficacy in suppressing colitis in mouse preclinical models of colitis.

Efforts are also underway to develop variants of EcN that sense and respond to environmental conditions. In one case, a variant was engineered to secrete type III interferon, an anti‐inflammatory cytokine, upon sensing nitric oxide (NO), a marker of colorectal inflammation. This was accomplished by expressing type III interferon tagged with a YebF secretion sequence under the control of an NO‐responsive promoter. Despite secreting very low cytokine levels (ng/mL), this strain demonstrated promise in an organoid model of intestinal inflammation (Chua et al. 2023). In another, EcN that secretes Acanthocheilonema viteae Cystatin (AvCystatin), a helminth‐derived immunomodulator upon sensing tetrathionate, a marker of gut inflammation, was demonstrated to suppress inflammation in another mouse model of IBD. In this case, AvCystatin was engineered with a signal peptide, enabling EcN to secrete it via a heterologous haemolysin secretion system (Zou et al. 2023).

3.2. Engineered Bacteria for the Treatment of Metabolic Disorders

Another emerging area of application of smart microbes is in treating metabolic disorders. This has primarily been accomplished by enhancing the metabolic properties of EcN and other bacteria to promote the consumption of toxic compounds in situ in the gut, minimising their uptake into circulation. In one case, variants of EcN were developed that promote the degradation of phenylalanine (Phe), a metabolite that accumulates in people with phenylketonuria (PKU). EcN was outfitted to express Phe metabolising enzymes in response to conditions in the anoxic gut. These strains demonstrated evidence of significantly reduced blood Phe concentration in mice and primate models of PKU (Isabella et al. 2018) and proof of efficacy in a phase 1 clinical trial (Puurunen et al. 2021). However, a phase 3 clinical trial conducted by Synlogic, the developer of these strains, was terminated prematurely, given concerns that the study was unlikely to meet the desired endpoint.

In other cases, variants of EcN were developed to promote the degradation of ammonia. This metabolite accumulates in patients with inborn mutations in ammonia‐metabolising enzymes or with significant liver disease, that is, cirrhosis (Kurtz et al. 2019) or methionine. This metabolite is found at elevated levels in the gut of patients with homocystinuria (HCU), another inborn disease of metabolism (Perreault et al. 2024). Similarly, variants that promote the degradation of oxalates that, when elevated, can lead to the development of kidney stones (Lubkowicz et al. 2022). Elevated urinary oxalate levels are found in patients with primary hyperoxaluria, a rare genetic disorder, as well as in patients with IBDs as well as those who have undergone gastric bypass surgery. Efforts are also underway to develop variants of EcN (Cao et al. 2021) and spores of Bacillus subtilis (Appala Naidu et al. 2019) that promote the degradation of alcohol in the gut as a strategy to decrease alcohol‐induced liver damage.

4. Hurdles to Tackle as Smart Microbes Move Towards the Clinics

It is early days in the development of smart microbes as therapeutics. While many have been generated, none have yet to be approved as a live biotherapeutic for medicinal use. There are no clear delineating guidelines regarding what is required for engineered microbes to advance to the clinics, as each engineered microbe has unique properties in terms of its colonisation and clearance properties, engineered microbial properties and mode of delivery. Each strain will need to be studied in well‐controlled clinical trials in intended patient populations to establish safety and efficacy.

The majority of engineered microbes we have discussed here focused on delivering therapeutics to the gut. Some have demonstrated efficacy in mouse models, yet many differences exist between the anatomy and physiology of the human and rodent gastrointestinal tracts. Potential future steps in analysing these strains before proceeding to human trials include investigating efficacy in larger animals, including pigs and non‐human primates. However, a limitation can be a lack of an appropriate disease model. Another option is to use organ‐on‐chip and computer models designed to capture complex aspects of human disease.

Large‐scale production and dosing strategies are additional hurdles that must be addressed. Will cold chain storage be possible, or will solid formulations be developed that can be stably maintained at room temperature? Similarly, biocontainment needs to be addressed. While it is possible to use auxotrophic strains that cannot replicate outside the lab, prolonged colonisation might be required in other cases to enable a clinical response. In this case, they will need to be engineered to allow for their clearance from patients, when necessary, and to prevent their spread when released into the environment.

5. Concluding Remarks

In summary, it is early and exciting times in the field of smart engineered microbes. There has been a recent explosion in the development of innovative smart microbes as human therapeutics. In parallel, synthetic biologists are developing novel ways to engineer strains to sense and respond to their surroundings and control their dosage and strategies for biocontainment. The field holds much promise, particularly in addressing some of the needs of developing countries. Yet, many hurdles exist before engineered microbes become a reality and achieve widespread acceptance.

Author Contributions

Rajkamal Srivastava: conceptualization; writing – original draft; writing – review and editing. Cammie F. Lesser: conceptualization; writing – original draft; writing – review and editing; funding acquistion.

Conflicts of Interest

The PROT₃EcT platform is the subject of US Patent no. 9,951,340 and US Patent 10,702,559, filed by Massachusetts General Hospital.

Data Availability Statement

The authors have nothing to report.