Engineered bacteria in disease diagnosis and therapy: A synthetic biology perspective

Abstract

Synthetic biology is an interdisciplinary field that integrates knowledge and techniques from modern biology and many other disciplines to design and construct novel biological systems or to modify existing life forms. Its core technologies include gene editing (e.g., CRISPR/Cas9), DNA assembly, in vivo directed evolution, and integration with artificial intelligence. The development of these technologies has greatly advanced the application of synthetic biology in medicine. In disease diagnosis, engineered bacteria have shown considerable promise. They can be designed to sense disease-specific signals and produce detectable reporter outputs, thereby establishing new paradigms for early diagnosis and real-time disease monitoring. For example, bacteria engineered via synthetic biology have been developed as "living sensors" to detect disease biomarkers. In therapeutic applications, synthetic biology offers a fresh perspective on using microorganisms to treat diseases. Researchers can design and construct microorganisms with tailored functions for targeted drug delivery, immunotherapy, and microbiome modulation. These applications not only improve the precision and efficacy of treatments but also offer innovative solutions to overcome the limitations of conventional therapeutic approaches. However, despite their considerable potential, the clinical translation of engineered bacteria still faces numerous challenges, such as ensuring stable in vivo colonization, controlling immunogenicity, standardizing large-scale production, and establishing robust regulatory and ethical frameworks. This review summarizes engineering strategies aimed at enhancing the safety and efficacy of bacterial therapies, with the goal of optimizing bacterial functions and expanding their potential in diagnostics and precision medicine.

Graphical abstract

1. Introduction

Synthetic biology is an emerging interdisciplinary field that has arisen from the integration of modern biology with various other disciplines, including molecular biology, systems biology, engineering, computer science, and materials science [1]. It focuses on designing and constructing novel biological systems and redesigning existing ones using engineering principles [[2], [3], [4], [5], [6]]. Following the discovery of the DNA double helix and the completion of the Human Genome Project, synthetic biology is now considered the third biotechnological revolution [7,8]. As early as the 1980s, Barbara Hobom coined the term "synthetic biology" to describe genetically modified bacteria created using recombinant DNA technology [9]. Since then, researchers have increasingly focused on exploring the potential of synthetic biology to engineer bacteria for disease diagnosis and treatment [[10], [11], [12]]. A landmark achievement in this field occurred in 1978, when Genentech announced the successful production of recombinant human insulin using Escherichia coli [13,14]. Four years later, the U.S. Food and Drug Administration (FDA) approved its market release, demonstrating the therapeutic potential of engineered microbes.

While early investigators exploited the natural properties of bacteria for disease intervention, their applications were constrained by the lack of advanced gene-editing tools. In recent years, especially with the advent of precise gene-editing tools such as CRISPR-Cas9, the design of artificial gene circuits and the reconstruction of complex metabolic pathways have advanced considerably [[15], [16], [17]]. Consequently, synthetic biology has made groundbreaking contributions to medicine.

Microbes engineered through synthetic biology have become a central focus of current research in microbial diagnostics and therapeutics. Their key advantage lies in programmability: by precisely rewriting the genome, researchers can direct cells to produce therapeutic molecules or entire metabolic pathways on demand [12]. For instance, Sanmarco et al. developed a synthetic probiotic that produces lactate, which activates the HIF-1α-NDUFA4L2 signaling pathway in dendritic cells, thereby suppressing autoimmune responses in the central nervous system (CNS) [18]. In addition, researchers have combined magnetic nanomaterials with engineered bacteria, enabling spatiotemporally controlled bacterial lysis and drug release using magnetic fields. This strategy activates antitumor immune responses and achieves synergistic treatment of both primary and distant colorectal cancer lesions [19].

Engineered bacteria not only serve therapeutic roles but are also being equipped with diagnostic functions [20]. For example, researchers have developed an "intelligent" whole-cell engineered bacterium named i-ROBOT, which integrates three modules fluorescent reporting, base editing, and drug secretion to simultaneously diagnose, record information, and ameliorate inflammatory bowel disease (IBD) in vivo [20]. Thus, using synthetic biology to engineer microorganisms for disease diagnosis and treatment has opened new avenues in medicine.

Despite the immense potential of engineered bacteria, their clinical application still faces challenges, such as stable in vivo colonization, control of immunogenicity, and standardization of large-scale production. Moreover, regulatory policies and ethical oversight must be developed in tandem to balance innovation with safety. This review summarizes engineering strategies aimed at enhancing the safety and efficacy of engineered bacterial therapies, with the goal of optimizing bacterial functions and expanding their potential in diagnostics and precision medicine.

2. Core technologies in synthetic biology

Synthetic biology is an interdisciplinary field that integrates knowledge from biology, engineering, computer science, and chemistry [4,[21], [22], [23]]. Its core mission is to design and construct novel biological systems with specific functions. The following sections discuss the core technologies commonly employed in this field and their applications in the study of engineered bacteria. The studies of engineered bacteria that are the focus of this review exemplify this concept.

Below, we examine several key technologies within this domain and illustrate their exemplary use in constructing engineered bacteria.

2.1. Gene editing technology

The advent of CRISPR-Cas9 has enabled precise and programmable modifications of bacterial genomes [15,[24], [25], [26], [27]]. Renowned for its simplicity, efficiency, and low off-target rate, it has become the preferred tool for constructing stable therapeutic strains [24,27]. Researchers use this technology for gene knockouts or integrations in various probiotic chassis. For example, Garrido et al. used CRISPR technology to streamline the genome of Mycoplasma pneumoniae by deleting virulence factor genes and engineering the bacterium to secrete antimicrobial peptides and dispersin B. This approach effectively eliminated Staphylococcus aureus biofilms in vivo [28]. Furthermore, Fang et al. integrated a Type I-E CRISPR-Cas system targeting multiple antibiotic resistance genes into the chromosome of Escherichia coli Nissle 1917 (EcN), achieving >99% efficiency in blocking horizontal gene transfer of resistance [29]. More recently, Brodel et al. developed a phage assisted platform that enables scarless integration of up to 100 kb of DNA into bacterial chromosomes. They used this system to install erythromycin biosynthesis pathways and multi-input genetic circuits into E. coli complex programs previously reliant on unstable plasmids thereby expanding the therapeutic potential of engineered bacteria for producing complex natural products and executing multi-stage therapies [30].

2.2. DNA assembly technology

Advances in DNA assembly technology have been crucial for the progress of synthetic biology [31]. Methods such as Gibson Assembly and Golden Gate cloning enable the construction of complex metabolic pathways and synthetic gene circuits composed of multiple genes. For instance, Yan et al. used these DNA assembly techniques to precisely integrate the entire 3-hydroxybutyrate synthesis pathway into the genome of EcN. This yielded a therapeutic strain capable of continuously producing 3-hydroxybutyrate in the intestine, thereby improving the gut microenvironment and alleviating colitis symptoms in mice [32]. This stable genomic integration strategy overcomes the instability and loss commonly associated with plasmid-based systems in vivo.

2.3. In vivo directed evolution technology

Directed evolution accelerates the selection of mutants with desired phenotypes by recapitulating natural selection under laboratory conditions. Unlike traditional directed evolution, which is constrained by library size and the efficiency of screening through multiple rounds of mutagenesis, emerging in vivo continuous evolution technologies (see Table 1) have dramatically accelerated this process. Among these, the OrthoRep system was initially developed in yeast by the Chang Liu team at UC Irvine, and its bacterial version has recently been established by the Jason Chin team at the MRC Laboratory of Molecular Biology [33,34]. For example, researchers have used phage-assisted continuous evolution platforms to engineer bacterial sensor kinases, enabling them to respond to previously unrecognized non-native small molecule signals. This paves the way for constructing engineered bacterial diagnostic sensors capable of detecting intestinal metabolites in real time in vivo [35,36]. This strategy opens new possibilities for developing in vivo diagnostic sensors.

Table 1.

Examples of In Vivo directed evolution technologies.

Technology Name	Team	Method	Advantages	Ref.
PACE Technology	Harvard University's David R. Liu Team	Couples' laboratory evolution of proteins with bacteriophage life cycle, enabling rapid continuous evolution through iterative phage replication.	About 100-fold faster than traditional methods; fully automated; eliminates manual intervention.	37
RAGE Technology	Zhao Huimin Research Group, University of Illinois	RNA interference-based genome evolution in yeast; identifies beneficial gene knockouts through iterative screening.	Accelerates screening of key genes by combining knockout with enhanced phenotypic selection.	38
Orthogonal Replication Systems	University of California, Irvine (Chang Liu lab)	OrthoRep system utilizes orthogonal DNA polymerase–replicon pair to continuously mutate user-defined genes in vivo without altering global mutation rate.	User-defined genes in OrthoRep continuously and rapidly evolve through serial passaging, a highly straightforward and scalable process.	34
	MRC Laboratory of Molecular Biology (Jason Chin lab)	Establishes a stable orthogonal replication system in E. coli using a bacteriophage PRD1-derived replicon selectively copied by an orthogonal DNA polymerase (O-DNAP) that does not copy the host genome.	Enables massively accelerated continuous evolution of diverse cargos (≥16.5 kb) in E. coli via an orthogonal DNA polymerase that does not copy the host genome.	33
CRISPR-guided Directed Evolution	Various (Methods in Enzymology)	Combines CRISPR base editors with plate-based screening for targeted in vivo mutagenesis in bacteria.	Balances ease of use with engineering power; flexible application for evolving CRISPR effectors.	39

2.4. Integration with artificial intelligence

The integration of artificial intelligence is profoundly reshaping the synthetic biology research paradigm [[40], [41], [42], [43]]. AI applications in areas such as component engineering, genetic circuit design, and metabolic engineering have markedly boosted R&D efficiency and expanded the range of achievable outcomes. In component engineering, for example, deep learning models such as generative adversarial networks are now used for the de novo design of functional protein sequences and promoters, thereby accelerating the optimization of natural components [40]. In genetic circuit design, in silico simulations can predict the dynamic behavior of circuits before any laboratory work begins. For instance, Li et al. used artificial neural networks to redesign inter-bacterial communication logic, creating a "perceptron" capable of recognizing and classifying chemical signal patterns [44]. In metabolic engineering, AI-driven metabolic models can rapidly identify the most promising intervention targets from large datasets. HamediRad et al.'s robotic platform, BioAutomata, combined with an AI-guided design-build-test-learn cycle, successfully optimized the lycopene biosynthesis pathway without human intervention, demonstrating the immense potential of AI in automated strain construction [42].

3. Application in disease diagnosis

In recent years, engineered bacteria have emerged as a novel diagnostic platform with significant potential. Using synthetic biology, bacteria have been engineered as "living sensors" that can detect disease biomarkers and generate corresponding signals. These engineered bacteria can be used for early disease detection, monitoring disease progression, and tracking therapeutic responses, thereby offering new tools for non-invasive and real-time disease monitoring.

These engineered bacteria can be used for early disease detection, monitoring disease progression, and tracking therapeutic responses, thereby offering new tools for non-invasive and real-time disease monitoring [45]. This system provides a simple, on-demand alternative to complex sequencing methods for microbiome profiling. The platform integrates two key components: RNA toehold switch sensors, which can be designed to detect virtually any RNA sequence, and a freeze-dried, cell-free transcription-translation system that permits long-term storage and can execute genetic circuits upon rehydration. By incorporating an isothermal RNA amplification technique (NASBA), the platform's detection sensitivity was enhanced to the femtomolar range. This tool demonstrated clinical utility by distinguishing active Clostridioides difficile infections from asymptomatic carriage.

In another example, a team at MIT designed engineered bacteria that activate fluorescent reporter genes upon sensing specific gut metabolites or inflammatory factors (such as thiosulfate), enabling early diagnosis of Crohn's disease or ulcerative colitis through fecal sample analysis (Fig. 1) [20]. Such studies transform bacteria into "living diagnostic devices" through modular genetic circuit design, thereby advancing real-time monitoring of intestinal diseases.

Fig. 1.

Design of i-ROBOT and it's in vivo validation for colitis diagnosis. (A) Schematic diagram of the engineered bacterial strain i-ROBOT for the diagnosis, recording, and treatment of colitis. (a) Illustration of the individual components and signaling pathways of i-ROBOT. The strain integrates three modules: a fluorescent reporter for real-time detection, a base-editing system for molecular recording, and a therapeutic module secreting AvCystatin to mitigate inflammation. (b) Functional overview: upon sensing thiosulfate (an inflammation-associated metabolite), i-ROBOT generates real-time fluorescence and heritable signals (genomic molecular recording with colorimetric output) to detect and monitor colitis. Concurrently, it secretes AvCystatin, alleviating DSS-induced colitis in mice. (B) Experimental validation of i-ROBOT's ability to detect inflammation in a murine colitis model. (a) Experimental design. Mice received 3% DSS ad libitum for 0, 1, 3, or 5 days. On day 5, each mouse received 2 × 10⁹ CFU i-ROBOT by gavage. Faeces and colon contents were collected 6–12 h post-gavage. (b) Disease activity index (DAI) on days 0, 1, 3, and 5. DAI increased with DSS exposure duration, confirming successful colitis induction. Data are mean ± SEM (n = 5 per group). ∗P < 0.05, ∗∗P < 0.01 vs. day 0. (c) Fluorescence intensity of i-ROBOT recovered from faecal samples, measured by flow cytometry. A marked increase was observed in the 5-day DSS group, indicating bacterial activation in the inflamed gut. ∗∗∗P < 0.001 vs. day 0. (d) Fluorescence intensity of i-ROBOT from colon contents. Like faecal samples, fluorescence was significantly higher after 5 days of DSS treatment. ∗∗∗P < 0.001 vs. day 0. (e) Sequencing analysis of i-ROBOT in faecal samples. Relative abundance of the engineered strain (blue) was significantly elevated in the 5-day DSS group, demonstrating enhanced colonization/survival in the inflamed gut. Data are mean ± SD (n = 5). ∗P < 0.05 vs. day 0. (f) Sequencing analysis of i-ROBOT in colon contents. Abundance increased after 5 days of DSS treatment, consistent with faecal results. ∗P < 0.05 vs. day 0. (g) Plate-count quantification of i-ROBOT (CFU/g) in faecal samples. Bacterial load was significantly higher in the 5-day DSS group. ∗P < 0.05 vs. day 0. (h) Plate-count quantification of i-ROBOT (CFU/g) in colon contents. A similar increase was observed after 5 days of DSS treatment. ∗P < 0.05 vs. day 0. All panels in (B) are adapted from Zou et al. [20] with permission.

Furthermore, a study engineered Acinetobacter baylyi to specifically recognize mutated genes associated with colorectal cancer (CRC), such as KRAS^G12D (Fig. 2) [46]. This bacterium exploits its natural horizontal gene transfer (HGT) capability, combined with a CRISPR-Cas system, to distinguish single-base mutations in tumor DNA. In mouse models, the bacteria successfully integrated tumor-released DNA and reported the event via antibiotic resistance genes. This "CATCH" strategy does not require sample purification and can directly detect cell-free DNA in the gut with high sensitivity, offering a non-invasive option for early CRC screening.

Fig. 2.

The CATCH strategy: EngineeredAcinetobacter baylyileverages horizontal gene transfer (HGT) to detect CRC-associated mutations in organoid and mouse models. (A) Schematic of the experimental workflow. A. baylyi biosensors are administered rectally into mice bearing colorectal tumors. The bacteria acquire tumor-released DNA via natural competence; if the DNA contains mutant KRAS, homologous recombination restores an antibiotic resistance gene, enabling selective growth on kanamycin plates. (B) In vitro co-culture assay. Biosensors were co-cultured with BTRZI-KRAS-kanR (CRC donor) organoid lysates or live organoids. Restoration of kanamycin resistance (kanR) via HGT was assessed by plating on kanamycin medium. (C) Recombination with DNA from crude organoid lysates allows biosensor growth on kanamycin medium. Colony formation occurred only with mutant KRAS template, demonstrating specific detection. (D) Fluorescence microscopy after 24 h co-culture. GFP-labeled biosensors (green) clustered more densely around mutant organoids than controls, suggesting preferential association. (E) Co-culture with established mutant organoids enables biosensor growth on kanamycin medium. CFU quantification confirmed significant enrichment only in the presence of mutant DNA. (F) In vivo HGT assay. BTRZI-KRAS-kanR tumors were induced in mice by colonoscopic injection (H&E staining confirmed tumor pathology). After rectal biosensor administration, luminal contents were analyzed for kanamycin-resistant colonies, confirming in situ tumor-DNA capture. All panels adapted from Cooper et al. [45] with permission.

These examples illustrate the versatility of engineered bacteria in disease diagnosis. From tumor DNA detection to inflammatory marker sensing, bacteria are endowed with "smart sensing" capabilities through genetic circuit design or surface engineering. Their advantages include non-invasiveness, high sensitivity, and the capacity for in situ, real-time monitoring. In the future, these bacterial sensors are expected to be integrated with microfluidics or wearable devices to advance personalized medicine.

4. Application in treatment

The rapid development of synthetic biology has provided new perspectives on the use of microorganisms in disease treatment. Through genetic engineering, researchers can construct microorganisms with tailored functions for targeted drug delivery, immunotherapy, and microbiome modulation. These applications enhance therapeutic precision and offer novel solutions to overcome the limitations of traditional approaches (Table 2).

Table 2.

Representative applications and advantages of engineered microorganisms in disease therapy.

Bacteria	Assay	Disease	Application	Advantage	Ref.
E. coli Nissle 1917	A fusion protein of Tum-5 and p53 was constructed using matrix metalloproteinase (MMP) cleavage site (PLGLWA) as a fusion gene connector	Solid tumors	Delivery of p53 and Tum-5	Deliver anti-cancer genes or anti-tumor drugs to the hypoxic areas of tumors	47
	The most effective sequence (GLP-1 GM) was used to design EcN strains to express GLP-1 GM	Obesity	Beneficial for obesity, blood sugar, and hepatic steatosis	Less side effects	48
	Synthetic biology methods and engineered microbial therapy	Tumer	Immunotherapies in cancer	Have a significant synergistic effect with PD-L1 blocking antibodies in tumor clearance	49
	based on a blue-light responsive module and upconversion nanoparticles	Tumor	Tumor ablation ability	Targeted sustained release of anti-tumor factors, biocompatibility	50
	Insertion of the genes encoding phenylalanine ammonia lyase and l-amino acid deaminase into the genome	Phenylketonuria (PKU)	Bacteria in the gastrointestinal tract consume phenylalanine (Phe)	Safety and tolerance	51
E. coli Nissle 1917	Integrating the 3HB synthesis pathway into different positions of the EcN genome using homologous recombination strategy	Colitis	3-hydroxybutyrate (3HB) is continuously produced in the intestinal tract	Improve the intestinal microenvironment and alleviate the symptoms of colitis	52
	Bacteria were streaked in a form of written words onto LB plates (36 °C, 12 h) and then cultured in a 42 °C incubator for 3 h.	Breast Cancer	Can control engineering bacteria to produce therapeutic protein tumor necrosis factor alpha (TNF-α) in tumors	Heat-sensitive engineered bacteria	53
	Self-assembly with liposomes	Ulcerative colitis (UC)	Reduce inflammation and histological damage, regulate epithelial barrier homeostasis in the intestine	Enhance anti-inflammatory ability	54
	Integrated the type I-E CRISPR-Cas3 system derived from E. coli BW25113 into EcN	Multidrug-resistant (MDR) bacteria	Not only provides a new strategy for restricting the transfer of ARGs while using probiotics but also enriches the genetic engineering toolbox of EcN	A lower growth burden	29
	After constructing the dapB-deletion mutant, dapB, the HlyBD secretion cassette and P15A ori were assembled into the pKT-5M2e-HlyABD plasmid.	Influenza A virus	Influenza A vaccine	Induce persistent humoral and mucosal reactions in the respiratory tract	55
S. boulardii	Introduce the gene encoding ABAB into Saccharomyces boulardii	C. difficile infection (CDI)	Preventing CDI risk and treating CDI patients	Neutralize TcdA and TcdB toxins	56
L. reuteri	The Listeria adhesion protein (LAP) gene was introduced into Lactobacillus casei to enable it to express LAP on the cell surface	Fatal infections caused by Listeria (Lm)	Competitively exclude Lm and improve intestinal barrier dysfunction induced by Lm	Enhance intestinal immune regulatory function and improve inherent health	57
probiotics	Using computational modeling combined with experimental validation of lysis circuit dynamics to determine the optimal genetic circuit parameters for maximum therapeutic effect	Local tumor	Effectively combining with the cytokine granulocyte macrophage colony-stimulating factor (GM-CSF) produced by probiotics, enhanced therapeutic effects were achieved in a syngeneic mouse model with poor immunogenicity	Targeted delivery	58
M. pneumoniae	Designed synthetic promoters and identified an endogenous peptide signal sequence	Staphylococcus aureus infection	Bacterial infections related to biofilms	Secrete antibacterial films and bactericidal enzymes	59
L. casei	Bioengineered Lactobacillus casei probiotic (BLP)	Listeria infection	Prevention of fetoplacental transmission of Lm by LAP-expressing BLP during pregnancy	Suppressed Listeria monocytogenes (Lm)-induced inflammatory response in mothers.	60
S. cerevisiae	Directed evolution and synthetic gene circuits	Inflammatory bowel disease (IBD)	Reduce intestinal inflammation and decrease fibrosis	Responding to specific physiological signals and providing therapeutic responses	61
SYNB8802	Mathematical modeling	Enteric hyperoxaluria (EH)	Metabolizes oxalate within the GI tract	This computational framework predicts a dose‐dependent lowering of UOx levels with daily oral administration of SYNB8802.	62
Oncolytic bacteria	Manganese dioxide coated polyformaldehyde immobilized bacteria	Tumor	Anti-tumor immunotherapy	Potent and safe antitumor Immunotherapeutics	63

4.1. Drug delivery

Drug delivery systems represent an important application of synthetic biology in medicine. Using genetic engineering, microorganisms can be designed to release drugs under specific physiological conditions, particularly within the tumor microenvironment. For example, a clinically relevant bacterium was engineered for targeted tumor therapy by synchronizing cycles of bacterial lysis for in vivo delivery [64]. One strategy involves using engineered bacteria to target hypoxic tumor regions, for instance, EcN was engineered to express a fusion protein of Tum-5 and p53 linked by a matrix metalloproteinase (MMP) cleavage site, enabling targeted delivery of anti-cancer genes [47]. Researchers used microfluidic devices to characterize the engineered lytic strain and demonstrated its potential as a drug delivery platform in co-culture with human cancer cells. In a mouse model of hepatic colorectal metastases, the combination of this engineered bacterium with chemotherapy significantly reduced tumor activity and improved survival rates compared with either therapy alone. This study established a methodology for developing bacteria-based therapies that can target disease sites using synthetic biology tools.

Recently, engineered microbes have shown considerable promise for treating inflammatory bowel disease (IBD) and ulcerative colitis (UC) [61,54]. A dual-bacterial, dual-drug expression system was developed to treat IBD [65]. In that study, an engineered strain of Escherichia coli was constructed to neutralize pro-inflammatory factors and enhance anti-inflammatory pathways. Oral administration of these bacteria led to a decrease in colonic inflammatory cell infiltration and a reduction in pro-inflammatory cytokine levels. The treatment also had a regulatory effect on the gut microbiota without exacerbating intestinal fibrosis, providing a new approach for IBD intervention. Similarly, researchers have used directed evolution and synthetic gene circuits to engineer Saccharomyces cerevisiae (yeast) capable of responding to specific physiological signals to reduce intestinal inflammation and decrease fibrosis [61]. Furthermore, another study reported the self-assembly of engineered EcN with liposomes to enhance its anti-inflammatory capacity, thereby regulating intestinal epithelial barrier homeostasis and alleviating UC [54].

Moreover, synthetic microorganisms can incorporate complex genetic circuits to achieve dynamic control of drug release. For example, bacteria can be designed to sense internal cues, such as those of the tumor microenvironment, and release antitumor drugs only after detecting tumor-related signals, thereby achieving precision therapy [66]. Alternatively, release can be controlled by external stimuli. One study developed EcN carrying a blue-light responsive module and up conversion nanoparticles, enabling targeted and sustained release of anti-tumor factors for tumor ablation upon near infrared (NIR) light irradiation [50]. This strategy improves drug targeting and reduces damage to healthy tissues, illustrating the broad potential of synthetic microorganisms in drug delivery.

4.2. Synergistic interaction between immunotherapy and microbiota

Immunotherapy has emerged as a major advance in cancer treatment; it works by activating the patient's own immune system to attack cancer cells [[67], [68], [69]]. Concurrently, the application of synthetic biology in this area has garnered widespread attention [[70], [71], [72]]. By engineering microorganisms to synergize with immunotherapy, therapeutic efficacy can be enhanced. Research indicates that microorganisms can augment immunotherapy efficacy by modulating the host's immune response. For instance, by genetically engineering Enterococcus faecalis, researchers discovered that specific structural modifications of its cell wall peptidoglycan can modulate immune cell activity, thereby enhancing the immune response [73]. These findings provide a new perspective on microbe-immune interactions and offer potential targets for new immunotherapeutic strategies. Modulating the gut microbiota may improve the efficacy of immune checkpoint inhibitors (ICIs), thereby offering more therapeutic options for cancer patients.

The synergistic interaction is also reflected in the impact of microbial communities on the efficacy of ICIs, which block inhibitory receptors such as PD-1 and CTLA-4. Patient responses to these drugs are variable and may be related to gut microbiota composition. Studies have shown that the presence of certain microbes can increase the response rate to immunotherapy by modulating the metabolism and function of immune cells [74]. For example, engineered EcN has been shown to synergize with PD-L1 blocking antibodies in tumor clearance [49]. Other approaches use engineered probiotics as local delivery vehicles for immunotherapy. One study used probiotics equipped with an optimized lysis circuit to deliver checkpoint blockade nanobodies locally to the tumor, achieving enhanced therapeutic effects [58]. Another strategy involves using oncolytic mineralized bacteria, which can serve as potent and safe anti-tumor Immunotherapeutics when administered locally [63]. In another study, researchers engineered thermosensitive bacteria capable of responding to heat stimulation within 30 min (Fig. 3) [53]. These bacteria enabled precise functional control in the intestines of multiple model organisms, including C. elegans, bees, and mice. The study further demonstrated that the engineered bacteria could colonize the tumor microenvironment in mice. After three rounds of heat-induced expression of the therapeutic protein TNF-α, tumor growth was significantly inhibited. This work establishes "thermal control" as a novel strategy for precisely manipulating engineered bacteria in vivo for tumor therapy [53].

Fig. 3.

Validation of engineered thermosensitive bacteria (HSB-T) for precise heat-inducible TNF-α secretion and inhibition of 4T1 tumors. (A) Conceptual overview: heat-induced TNF-α expression enables spatiotemporally controlled tumor therapy. (B) Schematic of the genetic construct. HSB-T carries a heat-shock promoter driving TNF-α; brief heat stimulation (42 °C, 30 min) triggers rapid TNF-α production and secretion. (C) Experimental timeline for in vivo therapy. 4T1 tumor-bearing mice received intratumoral injections of chitosan-encapsulated HSB-T, followed by three cycles of local heat application. (D) ELISA quantification of TNF-α secretion in vitro. Supernatants from HSB-T cultures were collected without heating (HSB-T) or after a 30-min heat shock (HSB-T + heat). Wild-type E. coli Nissle 1917 (WT) served as control. Heat induction increased TNF-α >10-fold (∗∗∗∗P < 0.0001). Data are mean ± SD (n = 3). (E) Cytotoxicity of HSB-T supernatants against 4T1 cells measured by CCK-8 assay. Only heat-induced supernatant significantly reduced cell viability (∗∗∗∗P < 0.0001), confirming functional TNF-α. (F) Live/dead staining of 4T1 cells after 24 h treatment. Viable cells appear green (calcein-AM), dead cells red (propidium iodide). Extensive red fluorescence was observed only in the HSB-T + heat group (scale bar = 200 μm). All panels adapted from Li et al. [63] with permission.

For instance, engineered EcN has been constructed as an influenza A vaccine, capable of inducing persistent humoral and mucosal immune responses in the respiratory tract of mice [55]. Therefore, modulating microbial communities may offer a new strategy for improving immunotherapy efficacy. Advances in synthetic biology enable more precise manipulation of microbial functions, opening new possibilities for cancer immunotherapy. Researchers can use gene editing and metabolic engineering to modify microorganisms to produce specific metabolites, secrete immunomodulatory factors, or directly target tumor cells. These engineered microorganisms can serve as adjuncts to immunotherapy or as independent therapeutic agents. For example, research teams are developing engineered bacteria that specifically target tumor cells and release antitumor drugs within the tumor microenvironment, thereby increasing local drug concentrations and reducing systemic toxicity. Future research will continue to explore how to optimize these synergistic effects to provide more effective treatment options for patients.

4.3. Metabolic engineering and microbiome repair

Beyond drug delivery and immune modulation, engineered bacteria show significant potential in metabolic engineering and microbiome modulation, treating diseases by regulating host metabolism. A prime example is the treatment of phenylketonuria (PKU): researchers inserted genes encoding phenylalanine ammonia lyase and l-amino acid deaminase into the EcN genome, enabling the bacteria to consume phenylalanine (Phe) within the gastrointestinal tract [51]. Similarly, for enteric hyperoxaluria (EH), researchers developed SYNB8802, an engineered bacterium capable of effectively metabolizing oxalate in the GI tract; its efficacy predicted by mathematical modeling [62]. These bacteria can also be modified to treat chronic metabolic disorders. For instance, EcN was engineered to express a GLP-1 analog, which demonstrated beneficial effects on obesity, blood glucose, and hepatic steatosis in obese mouse models [48]. Furthermore, by integrating the 3-hydroxybutyrate (3HB) synthesis pathway into the EcN genome, engineered probiotics can continuously produce 3HB in the intestine, thereby improving the gut microenvironment and alleviating colitis symptoms [52]. Another strategy for microbiome modulation involves immunomodulation. Cubilios Ruiz et al. engineered Lactococcus lactis to secrete IL-10 in the gut, promoting regulatory T-cell differentiation and suppressing inflammation. In a mouse model of colitis, oral administration of these bacteria reduced inflammatory scores and restored healthy microbial community structure [75]. This approach exemplifies systems-level therapy: engineered bacteria create a microenvironment that favor healthy ecosystem re-establishment, rather than directly targeting pathogens. Beyond exogenous probiotics, a paradigm-shifting approach is in situ microbiome editing. Gelsinger et al. developed a phage delivered CRISPR transposase system for site specific gene integration into resident gut bacteria. In a mouse model of phenylketonuria, they delivered a phenylalanine ammonia lyase gene into native E. coli, achieving a sustained therapeutic effect for over 6 months after a single treatment bypassing the engraftment challenges of exogenous probiotics [76].

4.4. Applications in anti-infective therapy

The emergence of antibiotic-resistant bacteria poses a major global public health challenge. Infections caused by antibiotic resistant bacteria are responsible for up to 700,000 deaths annually, a figure projected to reach 10 million by 2050 [77]. Pathogens such as methicillin-resistant Staphylococcus aureus (MRSA) and multidrug-resistant Escherichia coli (MDR E. coli) render traditional antibiotics ineffective, complicating treatment and increasing medical costs [78]. The overuse of antibiotics has accelerated the development of resistance through genetic mutation and horizontal gene transfer (Fig. 4) [79]. Therefore, the development of novel anti-infective strategies is urgently needed. The use of engineered bacteria for treating multidrug-resistant infections is an innovative approach. For example, an engineered bacteriophage was constructed to mitigate the development of antimicrobial resistance [80]. Researchers identified host-range-determining regions (HRDRs) in the tail fiber protein of the T3 bacteriophage and subjected them to site-directed mutagenesis to generate synthetic "phagebodies". This study demonstrated that mutations in HRDRs can alter the phage's host range, with certain phagebodies achieving long-term bacterial growth inhibition by preventing the emergence of resistance. This approach holds promise for developing next-generation antimicrobials.

Fig. 4.

A two-step cascade-metabolic engineeredEscherichia colistrategy for the treatment of hyperlysinemia. (A) Schematic of the probiotic-mixture approach. Two engineered E. coli strains act in tandem: EcNT(pTLS) converts lysine to α-aminoadipate semialdehyde; EcNT(pK25) further converts this intermediate to α-aminoadipic acid, bypassing the defective saccharopine pathway in hyperlysinemia. (B) Therapeutic efficacy in vivo. Hyperlysinemia mice received daily oral gavage of the probiotic mixture or vehicle. Plasma lysine levels were measured; treatment significantly reduced lysine accumulation versus controls (∗∗∗P < 0.001). Data are mean ± SEM (n = 6 per group). (C) Body-weight changes during treatment. Treated mice maintained normal weight gain, while controls progressively lost weight (∗∗∗P < 0.001 for treatment effect). (D) Differential abundance analysis of faecal microbiota. Principal coordinate analysis (PCoA) based on 16S rRNA sequencing revealed distinct clustering of microbial communities between treated and untreated mice, indicating probiotic-induced microbiota modulation. (E) Relative abundance of key bacterial phyla. Treated mice showed increased Firmicutes/Bacteroidetes ratio and enrichment of beneficial genera like Lactobacillus. (F) Heatmap of significantly altered OTUs. Several OTUs associated with short-chain fatty acid production were enriched in the probiotic-treated group, consistent with improved metabolic health. All panels adapted from Geng et al. [78] with permission.

In another study, a probiotic yeast-based immunotherapy was developed to combat Clostridioides difficile infection [81]. Researchers genetically engineered the probiotic yeast Saccharomyces boulardii to produce a tetra-specific antibody (ABAB) that neutralizes the two main toxins (TcdA and TcdB) produced by C. difficile [56]. In mouse models, this oral yeast immunotherapy showed potential for preventing and treating both primary and recurrent C. difficile infection and could be used concurrently with antibiotics, offering a new therapeutic avenue. Other strategies focus on competitive exclusion and barrier enhancement. For example, bioengineered Lactobacillus probiotics (BLP) expressing the Listeria adhesion protein (LAP) gene were shown to competitively exclude Listeria monocytogenes (Lm), ameliorate intestinal barrier dysfunction, and prevent fetoplacental transmission of Lm in pregnant models [57,60]. To combat biofilms, researchers engineered Mycoplasma pneumoniae to secrete antibacterial peptides and bactericidal enzymes, effectively eliminating Staphylococcus aureus biofilms in vivo [59]. Furthermore, a novel approach aims to curb the spread of resistance itself by integrating a type I-E CRISPR-Cas system into EcN, thereby blocking the transfer of multiple antibiotic resistance genes (ARGs) [29].

4.5. Engineered bacteria and regenerative medicine

Microorganisms also hold broad promise in regenerative medicine. Synthetic biology can be used to design therapeutic microbes that exploit their unique characteristics to promote tissue repair and regeneration. Certain engineered bacteria can secrete growth factors that facilitate the repair of damaged tissues, and their metabolic products have also shown beneficial effects on cell proliferation and tissue regeneration. For example, in bone tissue engineering, researchers have constructed engineered bacteria that target bone tissue and release bone growth factors, thereby promoting the proliferation and differentiation of osteoblasts. This approach not only enhances the efficiency of bone regeneration but also reduces the complications associated with traditional methods [82].

4.6. Challenges and limitations

Although synthetic biology holds significant potential for microbial diagnostics and therapeutics, it faces numerous challenges in practical application. These challenges are primarily related to safety and ethics, microbial stability and efficacy, and regulatory and market access hurdles.

4.7. Safety and ethical concerns

The genetic modification of microbial genomes raises important safety and ethical concerns. First, modified microorganisms may have unpredictable effects on human health and the environment. The survival and proliferation of engineered bacteria within the host could lead to unintended ecological consequences, such as disruption of the indigenous microbiota or the emergence of novel pathogens [83]. Additionally, certain applications may entail biosecurity and biosafety risks; for example, the misuse of gene-editing technologies could give rise to biological threats. To address these challenges, researchers are developing new biosecurity strategies, such as creating orthogonal biological systems to prevent genetic information exchange between synthetic and natural organisms.

Second, the application of synthetic biology raises ethical concerns, particularly regarding gene-editing technologies. Public acceptance of genetic modification varies widely, with many individuals expressing concerns about the safety and efficacy of engineered microorganisms and fearing their potential misuse [84]. Ethical issues also encompass respect for human dignity, fairness, and the inherent uncertainties of applying these technologies. These issues require thorough discussion and regulation to ensure the sustainable and responsible use of synthetic biology. Research must adhere to ethical standards, ensuring transparency and traceability. Ethical reviews at the research design stage is crucial to ensure that studies do not have adverse impacts on human health or the environment [85]. Therefore, establishing appropriate ethical frameworks is vital.

4.8. Microbial efficacy and stability

The efficacy and stability of microbes in therapeutic applications represent another major challenge. Although engineered microbes may perform well under laboratory conditions, their survival and functionality in complex in vivo environments can be affected by numerous factors. Environmental changes, such as temperature, pH, and nutrient availability, may lead to the failure or reduced functionality of engineered microbes [86]. Furthermore, within the complex host environment, interactions with the indigenous microbial community can compromise their therapeutic efficacy and safety.

To overcome these challenges, researchers need to develop more stable microbial carriers and optimize their in vivo survival conditions. For instance, constructing more complex gene regulatory networks using synthetic biology could allow engineered microbes to self-regulate their physiological functions under different conditions, thereby improving their clinical stability and efficacy [87]. Additionally, microbial therapeutic strategies need to be personalized for different disease types to ensure their safety and effectiveness in specific patient populations.

4.9. Difficulties in regulation and market access

The rapid development of synthetic biology also presents regulatory and market access challenges. Because this field involves genetic modification of microorganisms, relevant laws and regulations are still not well established in many regions, leading to regulatory ambiguity [88]. Many countries lack unified standards for the clinical application of engineered microorganisms, creating obstacles for companies seeking clinical trial approval and market access.

Difficulties in market access are also economic in nature. R&D in synthetic biology typically requires substantial financial investment. In the current medical market, investors are often risk-averse regarding emerging technologies, which can result in promising synthetic biology projects struggling to secure adequate funding [89]. Therefore, establishing rational market access mechanisms and incentive policies will be key to advancing this field.

In summary, the application of synthetic biology in microbial disease diagnosis and treatment faces multifaceted challenges. Only by strengthening research on safety and ethics, enhancing microbial stability, and improving regulatory and market access mechanisms can the full potential of synthetic biology be realized.

5. Limitations and prospects

There has been remarkable progress in the field of disease diagnosis and therapy using engineered bacteria. However, the field still faces multiple technical bottlenecks and translational hurdles that require systematic investigation in future research.

5.1. Technical challenges

First, the genetic stability of chassis microorganisms remains an unresolved issue. One of the core challenges in synthetic biology is ensuring the evolutionary stability of heterologous genetic constructs. The expression of non-essential proteins imposes a metabolic burden on the host bacterium, reducing its relative fitness and leading to the gradual loss or attenuation of exogenous construct function during passaging [90]. Even when using stabilization strategies such as chromosomal integration, mutational drift, or circuit functional decay can still occur during long-term cultivation. Kill switches are a typical example. Due to the strong selective pressure they impose, escape mutants arise readily, making such circuits among the most difficult to maintain stably [91]. Studies have found that cell-killing efficacy is negatively correlated with evolutionary stability and that genetic escape patterns are driven by circuit complexity, toxin activity, protective capacity against inactivation, and mutation-prone sequences within the circuit [92]. Second, the application of gene-editing tools to non-model strains is limited. Currently, most research on engineered bacteria focuses on a few model strains, such as EcN. However, efficient and specific genetic manipulation tools are lacking for commensal bacteria with unique ecological niche functions, such as Bacteroides species [93]. The complexity and diversity of the gut microbiome present significant challenges for the development of effective genome engineering tools for human gut microorganisms [94]. Developing gene-editing systems applicable to diverse commensal bacteria will be crucial for expanding the scope of engineered bacteria application [95]. Furthermore, although progress has been made in biocontainment strategies (e.g., kill switches and auxotrophic designs), their long-term reliability in complex in vivo environments remains to be fully validated. Future efforts should therefore focus on developing multi-layered, orthogonal biocontainment systems to minimize the risk of environmental escape [91].

5.2. The gap between models and clinical translation

Current research on engineered bacteria relies heavily on mouse models. However, there are significant differences between rodents and humans' intestinal physiology, immune systems, and microbial community composition. These differences may affect the colonization efficiency and drug release kinetics of engineered bacteria. Engineered bacteria-based cancer therapies face challenges, including off-target toxicity, poor penetration into deep tumor tissues, and the emergence of drug resistance [96]. Intratumoral heterogeneity within solid tumors also poses a challenge: the distribution of hypoxic and necrotic zones within a tumor is uneven, potentially restricting the targeted colonization of engineered bacteria to specific areas rather than covering the entire tumor. Furthermore, the composition of a patient's indigenous gut microbiota varies considerably among individuals. Consequently, the ability of the same engineered bacterial strain to colonize and its therapeutic efficacy may differ dramatically between individuals, raising new questions for the development of personalized treatment strategies. Integrating microbial systems with CAR-T cell therapy presents challenges related to biocontainment, an incomplete mechanistic understanding of tumor-homing specificity, and stringent safety validation requirements [97].

5.3. Standardization and regulatory needs

For engineered bacterial therapies to be widely used in clinical practice, the associated reagents and protocols must be improved and standardized. Currently, the construction and testing of engineered bacteria are largely performed manually, with each laboratory employing its own modified protocols. There are no standardized assays for determining the potency of engineered bacteria: should colony-forming units, circuit functional activity, or both be measured? How can plasmid stability and genetic circuit integrity be maintained during large-scale fermentation? How can lyophilized formulations ensure bacterial survival during gastric transit? These questions require systematic investigation. Critically, the regulatory framework is still evolving. Depending on their intended use, live biotherapeutic products (LBPs) may be classified as foods or drugs, with those bearing therapeutic claims being defined as medicinal products containing live bacteria or yeast intended for the prevention or treatment of disease [98]. There are significant disparities across global markets legislation, permissible claims, market value, and quality requirements, and different probiotic product categories incur varying costs at different stages of development [99]. Guidelines from regulatory agencies such as the U.S. Food and Drug Administration and the European Medicines Agency for genetically modified LBPs require further refinement. Balancing traditional safety pharmacology studies with the unique attributes of live medicines, including their potential for in vivo evolution, presents a novel challenge for regulators [98]. Ultimately, the derivation and construction protocols for engineered bacteria must be standardized and automated to enhance reproducibility and scalability, paving the way for these innovative therapies to transition from laboratory research to routine clinical application.

CRediT authorship contribution statement

Yan Shen: Writing – original draft. Si-Ming Lu: Writing – original draft, Funding acquisition. Long Yang: Formal analysis, Data curation. Yang Li: Methodology, Formal analysis, Data curation. Yu Zhang: Writing – original draft, Funding acquisition. Li-Guo Liang: Writing – review & editing, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.