Reporter systems for in vivo tracking of lactic acid bacteria in animal model studies

Abstract

Bioluminescence (BLI) and fluorescence imaging (FI) allow for non-invasive detection of viable microorganisms from within living tissue and are thus ideally suited for in vivo probiotic studies. Highly sensitive optical imaging techniques detect signals from the excitation of fluorescent proteins, or luciferase-catalyzed oxidation reactions. The excellent relation between microbial numbers and photon emission allow for quantification of tagged bacteria in vivo with extreme accuracy. More information is gained over a shorter period compared to traditional pre-clinical animal studies. The review summarizes the latest advances in in vivo bioluminescence and fluorescence imaging and points out the advantages and limitations of different techniques. The practical application of BLI and FI in the tracking of lactic acid bacteria in animal models is addressed.

Keywords:

• lactic acid bacteria
• reporter systems

Abbreviations

BLI	=	bioluminescence
CCD	=	charge-coupled device
FI	=	fluorescence imaging
FMNH2	=	flavin mononucleotide
GFP	=	green fluorescent protein
GIT	=	gastro-intestinal tract
IBS	=	irritable bowel syndrome
IVIS	=	in vivo imaging system
LAB	=	lactic acid bacteria
MRI	=	magnetic resonance imaging
NIR	=	near-infrared
RFP	=	red fluorescent proteins
Zn-DPA	=	zinc(II) dipicolyamine.

Introduction

Lactic acid bacteria (LAB) have several health benefits and a number of claims have been made relating to probiotic properties, e.g. the alleviation of constipation, treatment and prevention of atopic dermatitis, treatment of irritable bowel syndrome (IBS), prevention of cardiovascular diseases and, more recently, the treatment of cancerous tumors.^1,2 The chance of finding a single strain that adheres to all of these properties is remote, hence the reason for so many strain combinations in probiotic preparations. Many of the strains are selected based on in vitro-attained phenotypic properties that adhere to the general definition of a probiotic and to a lesser extent the ability of cells to colonize the gastro-intestinal tract (GIT). It is thus not surprising that the probiotic market is dominated by a few well-studied species of the genera Lactobacillus and Bifidobacterium.² It is only in recent years that reports have been published on the probiotic properties of LAB from genera such as Streptococcus, Leuconostoc, Lactococcus and Enterococcus.^3-5 Enterococci are usually not included in probiotic preparations, due to their association with abnormal physiological conditions, underlying disease and immunosuppression.⁶ Despite the few negative reports, clinical trials have shown that Enterococcus faecium SF68 is effective in the prevention and treatment of diarrhea⁷ and lowers the cholesterol levels in serum.⁸ Enterococcus faecium CRL 183 lowered cholesterol levels by 43% in in vitro tests⁹ and E. faecium Fargo 688→ alleviated symptoms of IBS.^10,11 Enterococcus mundtii ST4SA colonised the GIT of mice, competed against Clostridium sporogenes and Enterococcus faecalis for adherence to epithelial cells and mucus,¹² produces a broad-spectrum bacteriocin^3,13,14 and prevented symptoms of Salmonella infection in Wistar rats.¹⁵ The more we learn about the survival of other genera of LAB in the GIT, the more probiotic properties may be discovered.

Survival of probiotic bacteria is usually evaluated based on the ability of the cells to adhere to the GIT. Most of these studies are performed in vitro on human cell lines such as Caco-2, HT-29 and HT29-MTX.¹⁶ In vitro studies in GIT-models that simulate the human intestine² is a step closer to real conditions in the intestine, but remains an in vitro approach. Although in vitro studies provide valuable insight into the metabolism, adherence and antimicrobial activity of probiotic cells, the interaction between microbial and epithelial cells, and interaction among microorganisms in the GIT can only be studied in vivo. Cell interactions are best studied with labeled cells, i.e. cells that express genes encoding easily detectable reporter proteins. Tracking of probiotic strains through the GIT is possible, provided that the proteins are expressed constitutively and, preferably, without being influenced by changes in physiological conditions.¹⁷

Many reporter proteins are unstable when expressed in the GIT or mature slowly once expressed.¹⁸ The challenge is thus not so much to express a reporter protein in vivo, but to select a protein with the longest possible half-life. Overexpression of the reporter protein is also not a prerequisite, since low levels can be detected with the latest development in imaging technology. One of the most used detection systems is the Caliper in vivo image system (IVIS), equipped with a highly sensitive charge-coupled device (CCD) camera (Caliper Life Sciences, Hopkinton, MA, USA).

One of the best examples illustrating the value of a reporter system was the study on daptomycin and its effect against infections caused by Staphylococcus aureus.¹⁹ The authors used a luciferase-labeled strain of S. aureus in a murine model. The same technology was used to study the spread of Listeria monocytogenes infection in a murine model²⁰ and later by many other research groups to study the migration of bacteria in the GIT.^21-23 In this review, the advantages and disadvantages of different in vivo reporter systems are discussed, with emphasis on bioluminescent and fluorescent proteins.

Reporter Systems

Genes are selected as reporters when the characteristics of the expressed protein(s) allow easy detection of the cells that carry the reporter in a complex microbial environment.²⁴ A number of reporter molecules with improved activity have been developed over the past 2 decades.²⁵ Light emitted by these molecules is detected using advanced photon detectors in CCD cameras mounted within light-tight specimen chambers.²⁶ Macroscopic in vivo biophonic imaging is a highly sensitive, non-invasive, non-toxic and quantitative optical technique that is based on the detection of light produced by either fluorescent proteins or by enzyme-catalyzed oxidation reactions (bioluminescence). As opposed to traditional pre-clinical animal studies, in vivo optical imaging techniques such as BLI and FI enable researchers to tag bacteria and determine their cell numbers in vivo without taking biopsies or killing the animals.²⁷

Examples of useful imaging systems that are commercially available, including the various features of each, are listed in Table 1. By using sensitive light detection imaging systems such as the IVIS, the production of light from within living tissue is detected and quantified as it passes through different tissue types, including the skin, muscle and bone. In vivo bioluminescence and fluorescence imaging are usually performed using a planar imaging technique.²⁷ Firstly, a reference image is acquired under weak illumination. This is followed by illuminating the subject with light passing through a filter that is set to the excitation wavelength of a specified fluorophore. The emitted light signal is then captured in complete darkness using particular bandwidth emission filters. Bioluminescence imaging is performed in the same manner, without the need for a second illumination step. The exposure times required can vary from a second to several minutes, depending on the strength and location of the biophotonic signal within the animal. The resulting images are displayed in pseudocolour superimposed on the reference photographic images.

Table 1. Commercially available imaging systems

Manufacturer	Instrument	Features
Caliper Life Sciences (http://www.caliperls.com)	IVIS (several models)	Bioluminescence Fluorescence Digital X-ray (Lumina XR) Image freely moving animals (kinetic)
Berthold Technologies (http://www.berthholdtech.com)	NightOwl (2 camera options)	Bioluminescence Fluorescence
Biospace Lab (http://www.biospacelab.com)	PhotonImager	Bioluminescence Fluorescence Macrolens to convert to bioluminescence microscope Image freely moving animals (In Actio^R)
Cambridge Research and Instrumentation (CRi) (http://www.cri-inc.com)	Maestro	Fluorescence Spectral scanning
Carestream Health (http://www.carestreamhealth.com)	Kodak In Vivo Imaging systems (several models)	Bioluminescence Fluorescence Digital X-ray (FX/FX-Pro)
Li-Cor Biosciences (http://www.licor.com)	Pearl (1 mouse) Odyssey^R Imager+ Mousepod^™ (3 mice)	Near infra-red fluorescence
VisEn (http://www.visenmedical.com)	FMT 2500 Imaging system	Near infrared fluorescence Two modes: Reflectance Imaging and Quantitative Tomography Multimodality adaptors for CT/MR/PET

Before transforming a gene construct encoding bioluminescence or fluorescence into a microbial cell, it is important to access the suitability of the reporter gene for the specific experiment. If a bioluminescent phenotype is required, the choice of luciferase system usually depends on the microorganism of interest. An inherent advantage of the bacterial luciferase system is the ability to produce a bioluminescent signal without the need to add a substrate. The disadvantage is that lux genes are generally not optimally expressed in Gram-positive bacteria.^28,29 To circumvent the problem, Gram-positive ribosomal binding sites were cloned into the lux operon.³⁰

A vast array of reporter systems is available (Tables 2 and 3). The most popular include bioluminescence with bacterial lux genes (lux), firefly and click beetle luciferase (luc), and fluorescence recorded in the case of green and red proteins.^22,31,32 Other well established imaging systems include magnetic resonance imaging (MRI) and positron emission tomography.^25,33

Table 2. Properties of selected fluorescent reporter proteins

Fluorescent protein variant	Origin	Excitation max. (nm)	Emission max. (nm)
DsRed	Discosoma sp.	558	583
mCherry	DsRed	587	610
tdTomato	DsRed	554	581
mRasberry	DsRed	598	610
mPlum	DsRed	590	649
TurboRFP	eqFP578	553	574
mKate	eqFP578	588	635
Katushka	eqFP578	588	635
TagRFP	eqFP578	553	574
RFP693	eqFP611	588	639
tdRFP611	eqFP611	558	609
AQ143	aeCP597	595	655
IFP1.4	DrBphP	684	708

Table 3. Imaging systems utilizing bioluminescent bacteria

Species	Reporter system
Bacillus anthracis (85)	lux
Citrobacter rodentium (23)	lux
Brucella melitensis (86)	lux
Bifidobacterium breve (87)	lux
Escherichia coli (21)	lux, luc
Listeria monocytogenes (20)	lux
Mycobacterium tuberculosis (88)	lux, luc
Pseudomonas aeruginosa (89)	lux
Pseudomonas fluorescence (90)	lux
Salmonella enterica Typhimurium (75)	lux
Staphylococcus aureus (19, 29)	lux
Streptococcus pneumonia (30)	lux

In vivo Fluorescence

Fluorescence is defined as the emission of light from a chemical substrate when excited with a specific wavelength. Thus, to record fluorescence, the excitation and emission wavelengths have to be carefully selected, but also the conditions that support photostability and maturation of the protein²⁷ A number of fluorescent proteins (FPs) and probes have been developed for biological and molecular research.

Fluorescent proteins

Fluorescent proteins are genetically encoded in vivo reporters that are easily imaged^34,35 and have been used to tag other proteins, eukaryotic and prokaryotic cells.^36,37 A comparison of 3 of the most common and widely used reporter systems that confer identifiable characteristics is listed in Table 4. For a FP to be effectively used as a reporter molecule in in vivo optical imaging, it has to maintain emission levels when constantly illuminated and has to reach peak intensity within a specific time, a concept referred to as maturation time of the chromophore.³⁸

Table 4. Comparison of Lux, GFP and mCherry

Characteristic	Lux	GFP	mCherry
Color of light	Bluish-white	Green but variants are yellow, blue and cyan	Red, shade of purple in visible light
External requirements for light emission	None; except must view in dark room with dark adapted eyes	Requires excitation by UV light	Short wavelengths of light – excitation maximum of 587nm
Intrinsic requirements for light emission	5 additional genes plus luxR and luxI	None	None
Duration/stability of light emission	Lasts 5–9 days, stops after nutrient depletion	Last months as long as plates do not dry out	Shade of purple in visible light, resistant to photo bleaching
O2 dependency for light emission	Requires O2	None; some O2 needed for chromophore formation	Functions under hypoxic conditions

Green fluorescent protein (GFP) is encoded by a gene originally isolated from the jellyfish Aequoria victoria.³⁹ Osamu Shimomura, Roger Tsien and Martin Chalfie were awarded the Nobel Prize in Chemistry for the impact GFP technology had on research in life sciences. The first GFP gene was cloned and expressed in other organisms, including E. coli, during the early 1990s.⁴⁰ Since then, numerous derivatives of GFP with enhanced fluorescence and the potential to be expressed in different organisms were constructed. Cells expressing GFP, and derivatives thereof, emit light in the blue, green and yellow range of the spectrum. The protein emits light when excited by long-wavelength UV or blue light, without the requirement for an exogenous substrate or complex nutrients.^31,41 Green fluorescent proteins have been expressed in a number of LAB, including Lactobacillus sakei, Lactobacillus fructosus, Lactobacillus delbrueckii subsp. lactis and E. faecalis.^42,43 GFP was also be used to detect Lactobacillus spp. in the GIT of chickens.⁴² Recently, Vibrio cholera cells expressing GFP were used to visualize gastro-intestinal infection in zebrafish larvae.⁴⁴ Similarly, Rieu and co-workers⁴⁵ infected zebrafish larvae with cells of Lactobacillus casei expressing GFP to study GIT colonisation.

A variety of fluorescent proteins have been discovered and developed since the first application of GFP as a marker of gene expression in the nematode Caenorhabditis elegans.⁴⁶ This includes the discovery of the DsRed FP from the coral Discosoma sp and the development of a range of far-red FPs termed the ‘mFruits’.³⁴ The use of the DsRed fluorescent protein in live-cell imaging was hampered due to its slow maturation time (caused by its tetrameric form) and low photo stability.⁴⁷ The ‘mFruit’ FPs, developed in the laboratory of Roger Tsien are based on molecular-directed evolution of the DsRed FP.³⁴ Characteristics that render FPs most suitable for optical imaging are listed in Table 2.

To improve maturation time, brightness and photo-stability, optimized variants of red fluorescent proteins (RFP) were developed, of which mCherry, encoded by the mCherry gene, is the best known derivative.^34,48,49 The protein has a monomeric structure which matures from a nascent polypeptide into a folded fluorescent form within 15 min. Maximal fluorescence is reached in less than 2 h in E. coli and the protein matures 4 times faster than the DsRed protein it was derived from.³⁴

The monomeric structure of the mCherry fluorescent protein makes it ideal for protein fusions, as monomeric proteins tend to be the least disruptive to the function of the protein to which they are fused. The mCherry protein is non-toxic, and it can be expressed at high levels without adding any unwanted physiological stresses to host organisms.³² The mCherry fluorophore is also highly resistant to photo bleaching, which means that the fluorophore will not lose its ability to fluoresce during continuous illumination.⁵⁰ The mCherry gene has been codon-optimized for optimal expression in various organisms,^{32,34,35,51,52} including probiotic lactic acid bacteria (Fig. 1). Van Zyl and coworkers⁵² used the mCherry gene to study the colonization of Lactobacillus plantarum and Enterococcus mundtii strains in the intestinal tract of mice.

Figure 1. Lactobacillus plantarum 423 dam::cat-mCherry cells expressing the red fluorescent mCherry gene in the GIT of mice ex vivo. The color bar indicates the relative signal intensity measured as photons/sec/cm^-2/sr. Image modified from van Zyl.⁵²

Alternatives to fluorescent proteins

Genetic engineering of microorganisms is often time consuming and requires a considerable level of expertise in molecular biology. Although genetically encoded markers are used in biological sciences, they are not always easy to apply. A second strategy employs an NIR molecular probe that consists of a fluorescent reporter group.⁵³ The authors used a synthetic zinc(II) dipicolyamine (Zn-DPA) coordination complex that binds to anionic surfaces of bacterial cells. Carbocyanine dye, acting as injectable near-infrared (NIR) probe, attaches selectively to an affinity group with 2 Zn-DPA units and stains the whole cell. The probe has an excitation wavelength of 794 nm and emission wavelength of 810 nm. By using in vitro fluorescence microscopy, the authors have shown that the probe can be used to stain the periphery of S. aureus cells. In addition, they also demonstrated that the probe could be used in vivo to target S. aureus in a mouse leg infection model.⁵³

Limitations and side effects of fluorescent proteins

In recent years, bacterial imaging has become an emerging technology with many applications in biological science. Despite this, there are numerous limitations and challenges that affect the application of molecular markers.³⁸ Optical imaging methods such as radio-imaging and MRI are much more developed than in vivo imaging of bacteria.⁵³ A clear drawback of optical imaging within live animals is the penetration of light through tissue. Although higher wavelength bacterial imaging systems such as NIR probes can penetrate readily through animal tissue, the imaging of bacterial cells in deeper tissues with higher background fluorescence is often problematic.²⁷ The deeper the tissue penetration, the more bacterial cells are required to produce a detectable fluorescent signal. Typically, for each centimeter of tissue depth, there is an approximate 10-fold loss in photon intensity.⁵⁴ Endogenous chromophores such as hemoglobin, keratin and collagen absorb light within tissues in the visible spectrum (400–760 nm).⁵⁵ Generally, the absorption of wavelengths longer than 600 nm (red part of the spectrum) by endogenously produced fluorophores is reduced, allowing the propagation of red light through several centimeters of tissue.^56,57

In many applications, the use of fluorescent proteins such as DsRed is hampered by a slow maturation time, photobleaching, oligomerization, brightness of the chromophore and cell toxicity.³⁴ According to Wiedemann and co-workers,³⁸ a critical characteristic of fluorescent proteins suitable for in vivo application is brightness.³⁸ Brightness is defined as the product of the ability of a chromophore to absorb excitation light (molar extinction coefficient) and re-emit photons (quantum yield of fluorescence).^17,27 Consequently, fluorescent proteins with a higher extinction coefficient and quantum yield have higher physical brightness. The physical brightness of a chromophore may also be affected by the translation efficiency of the marker protein. Brightness of the monomeric mRuby RFP was increased 5 – 8-fold by codon-optimization of the marker protein for expression in mammalian cells.^58,59 Furthermore, the usefulness of fluorescent proteins for imaging is significantly influenced by the total number of functional molecules expressed, the percentage of proteins that develop into mature chromophores, the turnover rate of nascent polypeptides into functional chromophores and the time required before photobleaching occurs.^34,38

Phototoxicity is one of the most adverse side effects FPs may have on living cells and tissue. Light of short and longer wavelengths induces a phototoxic effect in cells.^60,61 Many of the negative effects related to the exposure of living cells to UV and visible light can be avoided by the use of microscopic methods that use infrared light for imaging of FPs.⁶²

In vivo Bioluminescence

Bioluminescence, defined as the release of energy in the form of light due to a chemical reaction, has been described for many bacteria, marine crustaceans, fungi, fish and insects. Of these examples, a group of enzymes collectively named luciferases are the best studied and are used in the labeling of bacteria, eukaryotic cells and studies that involve gene expression.^26,54 Luciferase generates visible light that arises from the oxidation of a substrate in the presence of ATP.⁶³ The best studied bioluminescent systems used in in vivo imaging are those that originated from bacterium Photorhabdus luminescens, the firefly Photinus pyralis, click beetle Pyrophorus plagiopthalamus and the marine copepod Gaussia princeps.²⁷

Bacterial lux bioluminescence

Luminous bacteria are the most widely distributed light-emitting organisms, with the majority existing in seawater and the rest living in terrestrial or freshwater environments. The strains usually belong to one of 3 genera; Photobacterium, Vibrio or Photorhabdus. Marine species are usually planktonic and are members of the genera Photobacterium or Vibrio, whereas terrestrial bacteria are closely associated with fish, squid and nematodes and are grouped within the genus Photorhabdus.⁶⁴

A bioluminescent signal (light) forms in the presence of oxygen and when the reaction is catalyzed by luciferase. The expression of genes encoding bioluminescence is regulated by the lux operon.⁶⁵ The lux operon used in most molecular studies (Fig. 2) originates from a Photorhabdus sp.^30,54 Bacterial luciferase is encoded by the structural genes luxA and luxB. Luciferase is thus a heterodimer composed of 2 different polypeptides known as α and β, respectively. Substrates for the bacterial luciferase are reduced flavin mononucleotide (FMNH₂), O₂ and long chain fatty aldehyde produced by the fatty acid reductase complex that is encoded by the luxCDE genes (Fig. 2). Two regulatory genes, luxR and luxI, encoding the regulatory protein and auto-inducer synthetase, respectively, are also required for light emission.

Figure 2. Arrangement of the luxCDABE open reading frames in Photobacterium spp, Vibrio spp. and Photorhabdus spp The luxI and luxR genes are adjacent to one another and luxI is the first of the 7 genes in the operon. luxI and luxR are required for activation of the luminescence genes.

One of the most significant advantages of the bacterial luciferase system is the ability of the bacteria to synthesize the substrate (long chain aldehydes and FMNH₂) required for light production. When eukaryotic bioluminescence systems such as the click beetle luciferase are expressed in bacteria, the substrate (luciferin) has to be added exogenously.²⁸ The optimum temperature of the P. luminescence lux operon lies within the same range as that of mammalian tissues, making it one of the most suitable bacterial luciferase systems for in vivo applications.⁶⁶

The bacterial luciferase system is predominantly used when bacterial infection models are studied.^30,67 An example of this is the cloning of the P. luminescence lux operon into infectious pathogens such as E. coli, Citrobacter rodentium, Listeria monocytogenes and Salmonella spp^20-23 Pathogenicity of the labeled bacteria remained unaffected and they were visualized throughout infection. Bacteria labeled with the lux genes were detected in the GIT of mice (Fig. 3).⁵² Integration of the lux genes into the chromosome of bacteria significantly increases the stability of the enzyme in terms of the light emitted and is generally preferred over the use of plasmid expression.

Figure 3. Monitoring of the colonization of Listeria monocytogenes EGDe in the intestinal tract of mice. Strain EGDe contains the luxABCDE operon of Photorhabdus luminescence. The bioluminescence signal was measured transcutaneously 30 min after infection, using an IVIS system (Caliper Life Sciences). The image was acquired using an integration time of 1 min. The color bar indicates the relative signal intensity measured as photons/sec/cm^-2/sr. Image modified from van Zyl.⁵²

Firefly bioluminescence

Other than bacterial luciferase, the most commonly used luciferases for in vivo bioluminescence imaging is the luciferase (luc) genes from the North American firefly, P. pyralis and the click beetle P. plagiopthalamus. Expression in bacteria requires codon-optimization of the luc genes.²⁸ The firefly luciferase enzyme is approximately 62 kDa in size and is encoded by a single gene, luc.⁶⁸ The luminescence reaction involves oxidation of the substrate D-luciferin (D-(-)-2-(6’-hydroxy-2’benzothiazolyl) thiazone-4-carboxylic acid) in the presence of Mg-ATP. This results in the production of oxyluciferin, CO₂ and light.^26,69

The P. pyralis luciferase enzyme produces light with an emission peak at 560 nm.⁶⁵ However, generation of light by firefly luciferase is affected by temperature, producing light at a peak of 610 nm at 37^oC, rendering it suitable for in vivo applications.²⁶ Researchers have developed a number of luc genes with improved expression in bacterial and eukaryotic cells, with shifted emission spectra, by genetically modifying the wild type luc genes.⁶⁵

The amount of light emitted in relation to energy consumed by the firefly luciferase represents the most efficient bioluminescent reaction known and is, therefore, the most attractive choice for expression in mammalian cells.^27,70 Researchers have expressed the firefly luciferase successfully in muscle tissue and the nasal airways of mice.^71,72 Daniel and coworkers²⁸ tagged lactic acid bacteria with the P. pyralis luciferase using a plasmid expression system, revealing valuable information about the spatio-temporal persistence of the bacteria in mice.

Limitations and side-effects of luciferase systems in marker applications

One of the most significant drawbacks of the bacterial and firefly luciferases is the relation of bioluminescent light production to the metabolic activity of the microorganisms expressing the luciferase enzymes.²⁶ Luciferases rely heavily on the availability of microbial metabolites for continuous light emission. Bacterial luciferase relies on the bio-availability of FMNH₂, whereas firefly luciferase requires ATP. This is important to note during in vivo infection studies, as a decrease or an absence of bioluminescence production by marked pathogens could be due to the bacteria being in stationary growth phase and not due to eradication of the infection.^20,29,30 Supporting this notion, an absence of bioluminescence production as a result of bacteria being in a phase of dormancy has been reported for lux-expressing Leishmania amazonesis⁷³ and Mycobacterium smegmatis.³²

Absorption of light by mammalian tissues is one of the main factors influencing the sensitivity and detection limits of optical imaging (Table 5). Hemoglobin serves as one of the most significant causes of light absorption in animals.⁵⁵ While hemoglobin absorbs light in the visible spectrum (400–760 nm), longer wavelengths of light (above 600 nm) propagate through mammalian tissue more readily.⁵⁶ Furthermore, melanin influences the absorption of light in animals with dark fur. Nude animals or animals with white fur are thus more suited for in vivo optical imaging studies.

Table 5. Factors that can affect the sensitivity of imaging systems ²⁷

Reporter gene (fluorescent/bioluminescent) expression level

Emission light wavelength (referring to reporter gene chosen)

Excitation wavelength (fluorescence only)

Sensitivity of detection system

Background fluorescence/bioluminescence

Animal fur color or pigmentation

Signal location within the animal (type and depth of tissue)

Since luciferases are oxygenases, molecular oxygen is a basic requirement for all luciferase enzymes to produce light. The lack of oxygen may thus limit the application of luciferases as reporters in anaerobic environments such as the GIT.²⁶ However, oxygen enters the GIT via diffusion from oxygen-rich tissue surrounding the lumen. A low level of oxygen was sufficient for production of detectable levels of bioluminescence by lux-expressing Citrobacter rodentium cells that colonized the murine GIT.²³

Expression of Reporter Genes in Lactic Acid Bacteria

Tracking of LAB in a complex environment such as the GIT is a challenge and requires the constitutive expression of reporter genes. Added to the challenge are the differences in transformation efficiency obtained for strains of the same species, as demonstrated by using the same electroporation protocol on different strains of L. crispatus.^74,75

Reporter genes are usually cloned into plasmids or integrated into the genome. By integration into the genome, gene expression is much more stable and antibiotic selection is not required, which is a big advantage in animal studies.⁷⁶ Integration into the genome also provides the added advantage of stabilizing the exogenous DNA by irreversible incorporation.⁷⁷ Target sequences for the genomic integration of reporter genes should be genes or regions that are non-essential to the host to ensure that the physiology of the host remains unchanged.⁷⁸ It is also important to make sure that recombinant strains only differ from the wild type by the newly acquired characteristics (fluorescence or bioluminescence). Complex experiments to prove that the host cell physiology has not been negatively affected, can be avoided by disrupting a non-functional genetic locus on the host chromosome. Prophage genes that are generally considered to be transcriptionally silent in the lysogenic phase are good examples of genetic loci that are often used for homologous recombination of foreign genes in LAB. Several LAB species harbour phage genes on their chromosomes, and heterologous DNA can be inserted, for example into the attB sites present in many prophages.⁷⁶ A more refined integration strategy makes use of 2 recombination events for integration. Following the integration of the entire plasmid, a second recombination event occurs to remove the plasmid backbone, while the gene of interest is irreversibly left behind in the host chromosome.⁷⁹ The event in which recombination occurs followed by looping out of the vector is determined by how the homologous sequences are arranged on the vector. Double recombination integration has been used with many LAB species, e.g., Lactobacillus casei, Lactobacillus plantarum and Lactobacillus acidophilus.⁸⁰

Since very few chromosomal sites have been investigated that will allow high levels of reporter gene expression in LAB, cloning of reporter genes into multiple copy number plasmids remains an attractive alternative for constitutive expression,⁸¹ providing that the plasmids are maintained by the host.^28,82 Plasmid reporter gene expression levels are usually higher than those of some chromosomally integrated genes.⁸³

Translation efficiency of a protein in a heterologous host organism can be significantly increased by modifying the codon usage frequency, thus codon optimization.⁸⁴ To enhance the translational efficiency of a reporter gene, the gene can be chemically synthesized with an optimized codon usage for expression in specific LAB. Garcia-Cayuela and de Carinanos³⁵ constructed fluorescent protein vectors for analyzing the expression strength of different promoters in LAB and E. coli. The authors used a synthetic red fluorescent protein that had been codon-optimized for increased expression in L. plantarum and E. faecalis. Luciferase reporter genes from eukaryotes such as the firefly (P. pyralis) or click beetle (P. plagiopthalamus) have a codon usage that is not suitable for expression in bacteria. Nucleotide sequences have to be optimised to maximize expression of the luciferase reporter protein in LAB. Firefly luciferase expression in M. tuberculosis increased 30-fold in signal intensity after codon-optimization.²⁷

Conclusion

Several reporter systems are available to study the survival of lactic acid bacteria in the mammalian GIT. Many of the reporter genes are either not expressed in the GIT or the luminescence produced is too weak to be detected with optical imaging methods. The excitation and emission wavelengths have to be carefully selected. The advantage of bacterial luciferase is that it requires FMNH₂, and not ATP, as in the case of firefly luciferase. This is important in studying the presence of pathogens in the GIT, as the bioluminescence signal is not linked to the metabolic activity (ATP levels) of the pathogen. Furthermore, expression of the firefly luciferase in bacteria requires codon-optimization.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.