Elsevier

Current Opinion in Biotechnology

Volume 35, December 2015, Pages 94-102
Current Opinion in Biotechnology

Engineered bacteria as therapeutic agents

https://doi.org/10.1016/j.copbio.2015.05.004Get rights and content

Highlights

  • Engineered bacteria provide a continuous supply of therapeutic molecules.

  • Bacterial therapies target inflammation, immunity, cancer, and metabolic disorders.

  • Novel strategies deliver transcription factors and enzymes to reprogram host cells.

  • Synthetic biology allows the design of safe bacteria for therapy.

Although bacteria are generally regarded as the causative agents of infectious diseases, most bacteria inhabiting the human body are non-pathogenic and some of them can be turned, after proper engineering, into ‘smart’ living therapeutics of defined properties for the treatment of different illnesses. This review focuses on recent developments to engineer bacteria for the treatment of diverse human pathologies, including inflammatory bowel diseases, autoimmune disorders, cancer, metabolic diseases and obesity, as well as to combat bacterial and viral infections. We discuss significant advances provided by synthetic biology to fully reprogram bacteria as human therapeutics, including novel measures for strict biocontainment.

Introduction

Bacteria are key elements for human health. Two evidences support this notion, the existence of a stably population of microbes, termed microbiota, in healthy individuals and the number of health disorders associated to axenic (germ-free) animals [1, 2]. In addition, bacteria have been actively administered in patients suffering from different illnesses for over a century. These intentional administrations are in general carried out with natural isolates obtained from the microbiota of healthy individuals and are referred to as probiotics. In most cases they belong to lactic acid bacteria (LAB) and in a lesser extent to Escherichia coli strains [3]. The development of efficient DNA technologies for manipulation of microbial genomes and the increasing knowledge of the molecular basis of diseases are allowing the engineering of tailored bacteria for the treatment of human disorders. Bacteria can be altered to produce a continuous and inexpensive supply of heterologous molecules of biomedical interest, such as human hormones, interleukins (ILs) and antibodies (Abs) within specific organs or tissues. We have restricted this review to bacterial engineering for human therapies, but similar concepts can be applied for the development of live bacterial vaccines [4].

Inflammatory bowel diseases (IBDs) have been prototypical targets of probiotic therapies, given the immunomodulatory effects that certain strains are able to exert in the gastrointestinal tract (GIT). Patients suffering from IBDs frequently have alterations in pattern recognition receptors and pro-inflammatory genes, which elicit an abnormal activation of the immune system in the gut and lead to chronic intestinal inflammation and to an increased rate of oxidative stress [5]. Different bacteria, mostly LAB, have been engineered to express a wide variety of compounds in situ (i.e. anti-inflammatory cytokines, anti-oxidant enzymes) to prevent the appearance of these symptoms [6]. These strategies are summarized in Figure 1. At least three of these engineered microorganisms have been tested in clinical trials. A Lactococcus lactis strain, engineered to secrete the anti-inflammatory cytokine IL-10 [7] showed a clinical benefit in 8 out of 10 patients tested with Crohn's disease [8], probably by inducing regulatory T cells (Tregs) through activation of dendritic cells (DCs) [9]. Another approach involved the use of engineered L. lactis strain secreting single domain antibody fragments (nanobodies) [10] against the pro-inflammatory cytokine Tumor Necrosis Factor alpha (TNF-α. This study showed promising results in murine models of IBD [11] and the bacterial strain has been tested in a phase 1 clinical trial by Actogenix (http://www.actogenix.com/). Lastly, L. lactis expressing human Trefoil Factor 1 (hTFF1) — a cytopeptide involved in epithelial wound healing - has been formulated as a mouthwash for the treatment of oral mucositis [12], which is a common complication found in patients that are subjected to chemo- and radiotherapies. This therapy has already passed through a phase 1 pharmacokinetic study and a phase 1b clinical trial involving cancer patients receiving chemotherapy, which showed alleviation of ulcerative oral mucositis symptoms in 30% of cases [13].

Besides the treatment of IBDs, engineered LAB secreting specific peptide antigens have been delivered orally to induce mucosal immune tolerance to antigens involved in food-allergies (e.g. ovalbumin) [14] or to gliadin antigens in the celiac disease [15]. Interestingly, this approach has been also applied to treat type 1 diabetes (T1D), an autoimmune disorder that arises from insufficient tolerance to self-antigens of pancreatic β-cells. An engineered L. lactis strain secreting the auto-antigen proinsulin and the anti-inflammatory cytokine IL-10, combined with subtherapeutic doses of systemic anti-CD3 Abs, was administered to non-obese diabetic (NOD) mice [16]. The immunosuppressive effect of anti-CD3 Abs (targeting the T-cell receptor) hampers immune system activation, whereas IL-10 avoids inflammation in the gastrointestinal mucosa and induces the development of Tregs, providing altogether an ideal scenario for the induction of tolerance. After 6 weeks of daily oral administration of the engineered strain, approximately 60% of the mice stably reverted diabetes, probably as a consequence of the infiltration of Tregs into pancreatic islets. Responding mice showed normal glucose levels in blood, and remained responsive to other antigens, suggesting an antigen-specific immunosuppression. A similar combination therapy based on anti-CD3 Abs and the oral delivery of engineered L. lactis secreting IL-10 and the diabetes-related auto-antigen glutamic acid decarboxylase (GAD)-65 has been reported [17]. This combination was effective in the treatment of advanced stages of T1D with severe hyperglycemia. Induction of tolerance against heat shock proteins (HSPs) has been also achieved using LAB. HSPs participate in the control of autoimmunity and in the cellular response to stress and are overexpressed in diseases involving inflammation and autoimmunity mechanisms, such as atherosclerosis and encephalomyelitis [18]. Oral administration of L. lactis strains secreting different forms of HSP-65 exerted a protective effect against endothelial damage and against the formation of atherosclerotic lesions in a low-density lipoprotein receptor-deficient (LDL-RD) mouse model [19]. This approach also prevented the development of encephalomyelitis in mice, showing reduced inflammatory cell infiltrate in the spinal cord [20]. These effects were associated with increased production of IL-10 and reduced levels of IL-17 and interferon gamma (IFN-γ).

Firsts reports of bacterial treatments against solid tumors date from the end of the 19th century. Over the years many genera of (facultative and strict) anaerobic bacteria (e.g. Salmonella, E. coli, Clostridium, Bifidobacterium) were reported to proliferate preferentially within solid tumors [21] due to a combination of mechanisms. These include enhanced bacterial entry and retention in tumors caused by their chaotic and leaky vasculature [22, 23] and unimpeded bacterial replication in the anoxic and immune-deficient tumor microenvironment, which lacks macrophage and neutrophil clearance mechanisms [24]. Chemotaxis and active motility of bacteria has also been shown to have a positive influence in the colonization of tumors [25, 26]. Intra-vesicle administration of Bacillus Calmétte-Guerin (BCG) has been used in the clinic during the last decades as the standard treatment for high-grade non-muscle invasive bladder cancer [27], representing a good example of a current anti-cancer therapy in which bacteria is employed to stimulate the immune system in order to promote the killing of cancer cells by a mechanism that is not yet fully understood.

Bacteria can bypass problems associated with poor selectivity and limited tumor penetrability of conventional cancer chemotherapies, and can be finely engineered to sense and respond to the tumor microenvironment [28]. However, the antitumoral effect of bacterial growth within tumors is generally weak, and different strategies have been followed to improve their therapeutic potential against cancer. One of them is the direct destruction of tumor cells through the secretion of bacterial toxins in situ (e.g. Staphylococcus aureus alpha hemolysin) [29, 30] or the expression of pro-drug converting enzymes that locally convert non-toxic prodrugs into drugs, like E. coli cytosine deaminase (CD) that transforms non-toxic prodrug 5-Fluorocytosine into toxic 5-Fluorouracil, resulting in a bacterial-directed enzyme prodrug therapy (BDEPT) localized in tumor areas [31]. BDEPT provides an excellent tumor selectivity since the drug is produced in situ, however, and similarly to conventional chemotherapy, its efficiency is highly dependent on physico-chemical properties of the prodrug that will define its ability to reach deep areas of the tumor in which bacteria (and therefore the converting enzyme) are located. In another strategy, engineered bacteria compete with the mechanisms that foster tumor formation (i.e. angiogenesis, resistance to apoptosis, evasion from the immune system, etc.) through the in situ delivery of polypeptides with pro-apoptotic activity (e.g. TNF-related apoptosis-inducing ligand -TRAIL-, Fas ligand, Noxa), anti-angiogenic factors (e.g. Endostatin, Thrombospondin 1), and cytokines (e.g. IL-2, LIGHT) that induce the immune system against tumor cells [28, 32]. These strategies are shown in Figure 2. Interestingly, an additional level of specificity is provided by controlled gene expression with promoters that specifically respond to small chemical inducers (e.g. l-arabinose, anhydrotetracycline, salicylate) [33, 34], tumor environment [35], hypoxia [36], or γ-irradiation [37]. Remarkably, the expression of the desired therapeutic protein can be specifically triggered within tumors using bacterial promoters responding to quorum sensing molecules released by the high-density of bacteria in tumors [38••]. A similar strategy has also been employed to amplify the expression of the desired protein induced with small molecules (e.g. l-arabinose) that diffuse poorly within the tumor mass [39]. Bacteria have also been engineered to silence the expression of important genes related to tumor development through RNA interference. For instance, E. coli was engineered to transfer to host cells plasmids encoding short-hairpin RNAs (shRNAs) silencing catenin □beta-1, whose overexpression is involved in several types of cancer [40]. This therapy has been granted by the FDA as orphan drug for the treatment of familial adenomatous polyposis and Marina Biotech (http://www.marinabio.com/) is currently developing clinical trials to analyze the safety and tolerability of its oral administration.

First reported clinical trials with refractory cancer patients using systemic administration of engineered Salmonella strains highlighted the trade-off between safe and effective dose, since the highest tolerated dose was insufficient for effective tumor colonization [41]. Hence, improving bacterial tumor colonization at low bacterial doses is an area of great interest. A promising advance is the constitutive and non-toxic expression on the bacterial surface of synthetic adhesins, which contain nanobodies targeting antigens expressed on the surface of tumor cells [42••]. Expression of synthetic adhesins in E. coli allowed a significant reduction of two-orders of magnitude in the dose of bacteria needed for efficient colonization of target solid tumors in mice.

Given that the intestinal microbiota influences the susceptibility of an individual to develop metabolic disorders, it seems reasonable that the introduction of properly engineered bacteria could ameliorate their symptoms. In the case of obesity it has been demonstrated that the composition of the microbiota in lean mice is different from the one found in obese mice [43, 44]. Furthermore, experiments involving transplantation of the intestinal microbiota from human identical twins discordant for obesity to germ-free mice revealed that only those mice receiving microbiota from the obese human developed obesity [45]. Recently, an approach to control obesity has been developed based on an engineered E. coli Nissle 1917 (EcN) strain expressing an N-acyltransferase from Arabidopsis thaliana that synthesizes N-acylphosphatidylethanolamines (NAPEs) [46••]. NAPEs are precursors of the N-acylethanolamide (NAE) family of lipids, which are naturally synthesized in the small intestine in response to feeding and cause a reduction in food intake and obesity. Incorporation of NAPE-expressing EcN in drinking water diminished food intake in mice, reducing body fat and weight gain, with no signs of adverse effects due to bacterial administration. Treated mice also had lower levels in plasma of insulin and leptin. Interestingly, the protective effect of the engineered bacteria persisted for at least 4 weeks after their removal from drinking water, suggesting a non-transient colonization of the GIT.

Engineered bacteria have been also administered to lower the elevated blood glucose levels (hyperglycemia) in T1D. The gut hormone Glucagon-like peptide (GLP)-1 was known to induce insulin production in epithelial cells [47]. An engineered Lactobacillus gasseri strain secreting GLP-1 has been able to reprogram intestinal cells into insulin-secreting cells [48]. Reprogrammed rat intestinal epithelia expressed important markers of pancreatic β cells, and the insulin secretion kinetics were similar to those of healthy control rats, indicating that insulin production was glucose-responsive.

Intervention against viral and bacterial infections has been addressed with engineered bacteria, which have been recently reviewed [49]. A wide variety of approaches have been followed, including binding of toxins, interference with quorum sensing molecules and adhesion mechanisms, release of neutralizing antibodies and antimicrobial factors, which are summarized in Figure 3. For instance, an engineered Streptococcus mutans strain deficient for lactic acid production — therefore unable to produce dental caries — and secreting a bacteriocin capable of killing virtually all other streptococci strains, was shown to displace cariogenic S. mutans in a rat model [50]. More recently, strains of L. lactis [51] and E. coli [52] have been engineered to secrete different bacteriocines in response to multiresistant Enterococcus faecalis pheromones and Pseudomonas aeruginosa quorum sensing molecules, respectively, showing bactericidal activity in vitro. Furthermore, an engineered Lactobacillus casei strain secreting human lactoferrin was demonstrated as effective to reduce the load of pathogenic E. coli in the duodenal fluid of infected mice, thus improving their illness score compared with untreated mice [53]. Production of antiviral molecules by engineered bacteria has been also investigated. Lactobacillus jensenii, a microorganism that is part of normal flora in human vagina, was engineered to secrete cyanovirin-N, a cyanobacterial protein with antiviral activity against HIV. Notably, trials in Rhesus macaques treated with the engineered L. jensenii strain and later challenged with simian HIV, showed a 63% reduction in HIV infection [54]. Modified Lactobacillus strains have been also used to deliver nanobodies interfering with rotavirus infection [55]. Lastly, there is a recognized potential in the incorporation of engineered bacteria into the microbiota of vector insects that mediate the transmission of human pathogens such as Trypanosoma cruzi, the causative agent of Chagas disease [56].

Although new technologies for genome edition such as transcription activator-like effector nucleases (TALENs) and clustered regulatory interspaced short palindromic repeat (CRISPR)/Cas9 endonuclease, are minimizing the risk of undesired genome modifications [57], reprogrammed stem cells obtained through DNA transfer techniques have a limited clinical applicability due to the risk of insertional mutagenesis during the differentiation process. In addition, it has been reported that successfully reprogrammed induced pluripotent stem cells (iPSCs) show a pronounced silencing of the transgenes encoding for reprogramming factors, suggesting that these factors are required only transiently to mediate cell reprogramming [58]. Transient protein delivery represents a good alternative to mediate differentiation of stem cells, since completely avoids the risk of insertional mutagenesis and provides a measure for the temporal control of cell exposure to reprogramming factors. The Type III Secretion System (T3SS) found in different Gram-negative pathogenic strains (Salmonella, Shigella, Yersinia, E. coli) is a needle-like macromolecular complex on the bacterial cell envelope that directly translocates effector proteins into the cytoplasm of infected host cells [59]. Attenuated strains carrying a functional T3SS have been used to deliver vaccine polypeptides and proteins of therapeutic potential (e.g. antibodies) into mammalian cells [60, 61]. Bacteria with T3SS have also been employed to deliver enzymes and transcription factors that can edit the mammalian genome and reprogram cell differentiation. An attenuated strain of Pseudomonas aeruginosa harboring a functional T3SS has been shown to deliver Cre-recombinase into the nucleus of mouse embryonic stem cells (or iPS) [62] triggering loxP mediated excision of the nuclear reprogramming cassette carrying c-Myc, Klf4, Oct4 and Sox2 [63]. The T3SS of this bacterial strain has also been used to translocate TALENs into human cells, which are capable of genome edition [64]. The same group reported the differentiation of mouse embryonic fibroblasts into myocites through the T3SS-mediated injection of MyoD, a transcription factor that acts as master regulator of myogenesis [65].

Synthetic biology aims to rationally design bacteria for therapy and other applications through the development of computational tools and techniques for extreme genetic manipulation. By these means, designed biological modules, devices and regulatory circuits of predictable behavior can be integrated into a bacterial chassis genome with strict biocontainment measures [66]. In recent years there have been multiple reports describing modular parts and more complex devices that can be used to program important aspects of the designed bacteria: controlled expression of payload proteins [67, 68], programmable adhesion to target surfaces and cells [42••], or the incorporation of stable memory into the engineered bacteria [69, 70] that can be used to detect small molecules in the gut [71]. Programming of bacterial tropism has also been demonstrated by engineering E. coli chemotaxis toward areas in which a pathogen, such as P. aeruginosa, is present [72]. Recent advances in engineering chemoreceptors and chemoeffectors [73] may allow a programmable chemotaxis to other molecules of interest in the future.

Ideal engineered bacteria for therapy should be sensitive to antibiotics and be free of mobile elements such as transposons and plasmids. Stable integration of the recombinant DNA in the chromosome is the simplest way to minimize gene flow, but there are other strategies available, like a mutually dependent host-plasmid platform based on conditional origins of replication, auxotrophies and toxin anti-toxin pairs [74]. In addition, engineered bacteria must have own containment strategies resistant to environmental supplementation, mutagenic drift and horizontal gene transfer. All these safeguards should avoid the spreading of these microorganisms into the environment and the proliferation of deleterious bacteria. Classically, biocontainment has been achieved through either engineered auxotrophies (e.g. strains deficient for thymidylate synthase) or induced lethality, which have been applied in clinical trials [8, 13•]. Synthetic biology is providing stricter biocontainment of the engineered bacteria. Minimal genomes encoding only the genes needed to sustain life might preclude unexpected evolution of engineered microbes. These minimal genomes could be generated through genome reduction techniques [75] and provide an excellent genetic chassis to implement the designed therapeutic gene devices (Figure 4). However, the definitive firewall for biocontainment might be the use of artificial genetic languages. Indeed, there are already available engineered E. coli strains that either incorporate a non-standard amino acid in the core of essential proteins [76••] or the synthetic thymine analog 5-chlorouracil instead of the natural thymine nucleotide in the DNA [77]. These strains exhibit strong resistance to evolutionary escape through mutagenesis or horizontal gene transfer, and cannot be supplemented with natural compounds.

Section snippets

Concluding remarks

Engineered bacteria represent an effective method to deliver therapeutic molecules in vivo allowing the development of novel treatments against infections and major human diseases such as inflammatory, autoimmune and metabolic disorders, and cancer. In addition, engineered bacteria could help to reprogram host cells by delivering transcription factors and enzymes for genome modification. Next-generation therapeutic applications of bacteria will be based on the conscious design of microorganisms

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

Work in the laboratory of LAF is supported by research grants from the Spanish Ministerio de Economía y Competitividad (MINECO) (BIO2014-60305R and BIO2011-26689), BACFITERed (SAF2014-56716-REDT), Comunidad Autónoma de Madrid (S2010-BMD-2312), La Caixa Foundation, and the European Research Council (ERC-2012-ADG_20120314). CPL was supported by a PhD FPI contract BES-2009-024051 from MINECO. DRG was supported by contract Apoyo a la Investigación from the Comunidad Autónoma de Madrid.

References (77)

  • A.T. St Jean et al.

    Bacterial delivery of Staphylococcus aureus alpha-hemolysin causes regression and necrosis in murine tumors

    Mol Ther

    (2014)
  • P. Lehouritis et al.

    Bacterial-directed enzyme prodrug therapy

    J Control Release

    (2013)
  • J.-H. Jeong et al.

    Anti-tumoral effect of the mitochondrial target domain of Noxa delivered by an engineered Salmonella typhimurium

    PLoS ONE

    (2014)
  • J.L. Royo et al.

    In vivo gene regulation in Salmonella spp. by a salicylate-dependent control circuit

    Nat Methods

    (2007)
  • A. Suzuki et al.

    Glucagon-like peptide 1 (1-37) converts intestinal epithelial cells into insulin-producing cells

    Proc Natl Acad Sci U S A

    (2003)
  • L.A. Lagenaur et al.

    Prevention of vaginal SHIV transmission in macaques by a live recombinant Lactobacillus

    Mucosal Immunol

    (2011)
  • J.E. Galán et al.

    Bacterial type III secretion systems: specialized nanomachines for protein delivery into target cells

    Annu Rev Microbiol

    (2014)
  • C. Bichsel et al.

    Bacterial delivery of nuclear proteins into pluripotent and differentiated cells

    PLoS ONE

    (2011)
  • C. Bichsel et al.

    Direct reprogramming of fibroblasts to myocytes via bacterial injection of MyoD protein

    Cell Reprogram

    (2013)
  • O. Wright et al.

    GeneGuard: a modular plasmid system designed for biosafety

    ACS Synth Biol

    (2015)
  • I. Sekirov et al.

    Gut microbiota in health and disease

    Physiol Rev

    (2010)
  • J. Behnsen et al.

    Probiotics: properties, examples, and specific applications

    Cold Spring Harb Perspect Med

    (2013)
  • R.J. Xavier et al.

    Unravelling the pathogenesis of inflammatory bowel disease

    Nature

    (2007)
  • R. Martin et al.

    Role of commensal and probiotic bacteria in human health: a focus on inflammatory bowel disease

    Microb Cell Fact

    (2013)
  • L. Steidler et al.

    Treatment of murine colitis by Lactococcus lactis secreting interleukin-10

    Science

    (2000)
  • I.L. Huibregtse et al.

    Genetically Modified Lactococcus lactis for Delivery of Human Interleukin-10 to Dendritic Cells

    Gastroenterol Res Pract

    (2012)
  • S. Muyldermans

    Nanobodies: natural single-domain antibodies

    Annu Rev Biochem

    (2013)
  • S.A. Limaye et al.

    Phase 1b, multicenter, single blinded, placebo-controlled, sequential dose escalation study to assess the safety and tolerability of topically applied AG013 in subjects with locally advanced head and neck cancer receiving induction chemotherapy

    Cancer

    (2013)
  • I.L. Huibregtse et al.

    Induction of ovalbumin-specific tolerance by oral administration of Lactococcus lactis secreting ovalbumin

    Gastroenterology

    (2007)
  • T. Takiishi et al.

    Reversal of autoimmune diabetes by restoration of antigen-specific tolerance using genetically modified Lactococcus lactis in mice

    J Clin Invest

    (2012)
  • S. Robert et al.

    Oral delivery of glutamic acid decarboxylase (GAD)-65 and IL10 by Lactococcus lactis reverses diabetes in recent-onset NOD mice

    Diabetes

    (2014)
  • C.A. Colaco et al.

    Heat shock proteins: stimulators of innate and acquired immunity

    Biomed Res Int

    (2013)
  • H. Jing et al.

    Oral administration of Lactococcus lactis delivered heat shock protein 65 attenuates atherosclerosis in low-density lipoprotein receptor-deficient mice

    Vaccine

    (2011)
  • H. Maeda

    The link between infection and cancer: tumor vasculature, free radicals, and drug delivery to tumors via the EPR effect

    Cancer Sci

    (2013)
  • K. Westphal et al.

    Containment of tumor-colonizing bacteria by host neutrophils

    Cancer Res

    (2008)
  • B.J. Toley et al.

    Motility is critical for effective distribution and accumulation of bacteria in tumor tissue

    Integr Biol (Camb)

    (2012)
  • M. Zhang et al.

    Trg-deficient Salmonella colonize quiescent tumor regions by exclusively penetrating or proliferating

    J Control Release

    (2015)
  • N.S. Forbes

    Engineering the perfect (bacterial) cancer therapy

    Nat Rev Cancer

    (2010)
  • Cited by (78)

    • Rheumatic diseases: The microbiota-immunity axis in development and treatment

      2023, Translational Autoimmunity: Volume 6: Advances in Autoimmune Rheumatic Diseases
    • Age of information in molecular communication channels

      2022, Digital Signal Processing: A Review Journal
    • Thermogenetics: Applications come of age

      2022, Biotechnology Advances
      Citation Excerpt :

      In certain situations, molecular beacons can also be attached to the cell for in vivo imaging (Gilad and Shapiro, 2017) or serve as in vivo thermometers for visualising thermogenesis (Kiyonaka et al., 2013). All these advantages of using cells as a therapeutic platform, have opened possibilities for the development of genetically reconfigured organisms to detect diseases in previously hard to reach anatomical sites such as the gastrointestinal tract (Kim et al., 2018; Kotula et al., 2014; Riglar et al., 2017) and also as delivery agents for cancer, gene and immuno-based therapies (Chien et al., 2017; Chua et al., 2017; Danino et al., 2015; Ho et al., 2018; Pinero-Lambea et al., 2015). An important criterion for biomedical applications is the need to precisely and non-invasively activate the programmed cell at the diseased location with high spatial and temporal specificities (Fig. 3a).

    View all citing articles on Scopus
    View full text