Real-time visualization of Y. ruckeri infection
One of the most important applications of bioluminescent reporter systems is their use to monitor bacterial or viral infections in animals. To analyse the pathogen–host interaction in a fish model and monitor the bacterial dissemination throughout the course of infection, two groups of 30 rainbow trout weighing from 8 to 10 g were infected with Y. ruckeri 150 harbouring the pCS26-Pac plasmid (Bjarnason et al., 2003). According to previous LD50 determination experiments, one group of fish was infected by intraperitoneal injection [104 cfu (0.1 ml)−1 per fish] and the other one by bath immersion (107 cfu ml−1) for 1 h. At 24 h intervals over a 3 day period, five fish from each challenge were euthanized, dissected and analysed with an IVIS® Imaging System (Xenogen) to monitor the bacterial progression inside the fish. For both infection systems it was verified that the modified strain retained the ability to cause disease in rainbow trout. The plasmid stability during the experiments was confirmed by plating cells on nutrient agar medium (NA) and NA with kanamycin (50 μg ml−1) and visualizing the resulting colonies with the IVIS Lumina. When infection was carried out by intraperitoneal injection, more than 90% of the fish analysed presented large numbers of lux-positive bacteria in the swim bladder at 24 h post infection (Fig. 2A). The high oxygen concentration existing in this organ may favour the growth of Y. ruckeri. At 48–72 h post injection, approximately in the 60% of fish analysed the infection was extended from the swim bladder to the rest of the fish, including gills and intestine tissues (Fig. 2B and C). The colonization of these organs was previously described by Fernández and colleagues (2003) who used an inducible β-galactosidase expression system to monitor bacterial tissue distribution in rainbow trout. This bacterial distribution pattern is also consistent with former studies, in which Avci and Birincioglu (2005) described hyperaemia and haemorrhages in the swim bladder after intraperitoneal injection with Y. ruckeri.
Figure 2. Bioluminescent tracking of Y. ruckeri pCS26-Pac infection in rainbow trout. Groups of 30 fish weighting from 8 to 10 g were challenged with Y. ruckeri pCS26-Pac. At 24 h intervals over a 3 day period, five fish from each experiment were euthanized and analysed with the IVIS® Imaging System (Xenogen). One side of the fish was removed to enhance visualization of bacteria in the tissues. Bioluminescence production was captured for 3 min and images were processed with a binning of 2. Colour standards represent ‘RLU max’. This measure expresses the highest number of counts in a pixel inside the region analysed. The experiment was made in triplicate.
A–C. Fish infected by intraperitoneal injection with 104 cfu per fish.
D–F. Fish infected by bath immersion with 107 cfu ml−1 for 1 h. Note that ‘RLU max’ is higher in E (where bacteria are concentrated in a specific region of the gut) than in F (where they are disseminated along the gut). Total RLU was similar (∼ 106 RLU) in both images.
Download figure to PowerPoint
Immersion bath challenge reflected much better the natural infection of Y. ruckeri in rainbow trout. In this case, the results showed that at early times post infection (24–72 h) bacteria colonized progressively the intestinal tract, this being the main organ where the bioluminescence signal was localized (Fig. 2D–F). This bacterial distribution is different from the one found for Listonella anguillarum where the fish skin is the first tissue to be affected by the bacterium indicating that this protection layer could be the first target and the portal of entry for the pathogen (Weber et al., 2010).
Although at the first stages of infection the presence of Y. ruckeri in the intestine is clear suggesting this organ as the initial target of the bacterium, the entry for the pathogen could occur through other sites, such as the gills or the skin wounds as was previously indicated by Tobback and colleagues (2009). Considering the detection limit of the IVIS® Imaging System, the number of bacteria during the first 12 h post infection may be insufficient for a visible luminescence and sites of entry in which bacterial replication is limited could appear as non-luminescent.
In order to know if the lack of luminescent signals in some fish organs such as spleen or liver in the early stages of the infection (Fig. 2A, B, D and E) was due to light quenching, consequence of their non-translucent character, the luminescence and the cfu of individual organs were calculated. For this, 10 fish were injected with 0.1 ml of a suspension containing 105 cfu ml−1 of Y. ruckeri harbouring the pCS26-Pac plasmid. After 24 h post infection five fish were euthanized and gills, swim bladder, gut, spleen and kidney of each fish were screened independently with the IVIS® Imaging System and RLU were calculated. Afterwards, each organ was suspended in 200 μl of phosphate-buffered saline (PBS) and vigorously shaken and cfu were determined by plating each suspension in NA plates containing kanamycin. The results indicated that in the early stages of the infection after intraperitoneal injection, the spleen (∼ 10 cfu per organ), the gills (∼ 102 cfu per organ), the liver (∼ 102 cfu per organ) and the gut (∼ 103 cfu per organ) did not emit luminescence (Fig. 2A and B) because the number of cfu of Y. ruckeri 150 pCS26-Pac per organ was under the detection limit of the IVIS® Imaging System. Only the swim bladder contained more than 104 cfu. In the later stages of the infection (Fig. 2C) when a large number of bacteria were disseminated throughout the fish, all the organs emitted luminescence.
Bioluminescence imaging revealed that two different models of infection can be postulated depending on how bacteria are administered. However, both systems usually finish with a systemic dissemination of the pathogen which involves the fish death. This appeared at 48–72 h and 5 days after intraperitoneal injection and immersion challenge respectively. Bath infection procedure is considered as a more natural route of bacterial administration but the intraperitoneal injection has the advantage of enabling the standardization of the bacterial dose administered to each fish. For this reason, in vivo studies which involve a high number of individuals or require statistical analysis are usually carried out with this kind of infection. However, it should be considered the loss of information that may occur during the study of some factors, especially those involved in the first stages of the infection process. Thus, the interpretation of the results is always subject to the kind of infection carried out.
Quantification by BLI of in vitro and in vivo expression of Y. ruckeri genes
As a direct approach, the pCS26-Pac plasmid containing the P. luminescens luxCDABE operon could not be used for analysing the expression of different promoters of Y. ruckeri during the infection process since it itself yielded highly levels of bioluminescence (Fig. 3). In order to avoid readthrough from promoters upstream of the luxCDABE operon, a strong transcriptional terminator from the Y. ruckeri cdsAB operon (Méndez et al., 2011) was amplified by PCR and cloned at the 5′ end of the lux operon giving as a result the plasmid pCS26-PacTER. As could be observed in Fig. 3A, the presence of this transcriptional terminator led to the complete extinction of bioluminescence when the bacterium was grown on NA showing that no transcription from the lux operon occurred under these conditions. The lack of light in bacteria harbouring the pCS26-PacTER plasmid was also confirmed in vivo. No light at all was detected in fish challenged with Y. ruckeri 150R harbouring the pCS26-PacTER in comparison to the high levels of bioluminescence found when the plasmid lacked the terminator (Fig. 3B). Therefore, the constructed plasmid can also be used in in vivo experiments to evaluate the level of transcription of defined promoters. Bioluminescence-tagged vectors were used previously to monitor the activity of certain promoters in real time during bacterial growth in culture (Yan et al., 2009). In Listeria monocytogenes, the analysis of lux promoter fusions demonstrated significant induction of bacterial virulence gene expression after infection of the insect Galleria mellonella at both 30°C and 37°C (Joyce and Gahan, 2010).
Figure 3. Bioluminescence imaging of Y. ruckeri harbouring pCS26-Pac and pCS26-PacTER plasmids. The effect of the terminator on light emission by Y. ruckeri strains harbouring both constructions were assessed both in vitro and in vivo by luminescence quantification. Colour bars on the right represent the scale for light quantification.
A. Strains were grown overnight on NA at 18°C.
B. Rainbow trout fry injected intraperitoneally with 105 cfu per fish of each strain were sacrificed at 48 h post infection.
Download figure to PowerPoint
To assess the validity of the pCS26-PacTER in the study of gene expression in Y. ruckeri, promoters of four well-characterized operons of Y. ruckeri 150R such as yrp1 (Fernández et al., 2003), yhlBA (Fernández et al., 2007b), yctCBA (Navais et al., 2011) and cdsAB (Méndez et al., 2011), as well as the constitutive promoter PgyrB from the gyrase B gene were PCR amplified and cloned into the pCS26-PacTER plasmid. After introducing the corresponding constructions into Y. ruckeri 150R, the resulting strains were analysed by bioluminescence quantification both on NA at 18°C and during bacterial infection of fry rainbow trout. Interestingly, the promoter of the yhlAB operon (Pyhl) which is involved in the production of an iron- and temperature-regulated haemolysin (Fernández et al., 2007b) and the promoter of cdsAB operon (Pcds) involved in the uptake and later degradation of l-cysteine (Méndez et al., 2011), both described as in vivo-induced (Fernández et al., 2004), exhibited poor luminescence on NA (< 100 RLU) (Fig. 4A). This was significantly lower than those values recorded for the constitutive promoter of gyrase B gene of Y. ruckeri (PgyrB). These results are consistent with previous studies in which induction of Pcds promoter required the presence of cysteine in the culture medium (Méndez et al., 2011) and high concentrations of iron inhibited Pyhl promoter (Fernández et al., 2007b). On the contrary, the promoter of yctCBA operon (Pyct) involved in in vivo citrate uptake (Navais et al., 2011) and the protease Yrp1 (Pyrp1) (Fernández et al., 2003) were the promoters with the highest bioluminescence level on NA (Fig. 4A). In this case the presence in the medium of citrate and peptides that induce Pcyt and Pyrp1, respectively, could explain the results observed. In fact, when these promoters were analysed in minimal medium M9 (Romalde et al., 1991) supplemented with 0.5% glucose, 0.2% casamino acids and 1.5% agar (M9GC) both activities decreased considerably (286 RLU for Pyct and 2407 RLU for pYrp1).
Figure 4. In vitro and in vivo expression of specific Y. ruckeri promoters. Transcriptional fusions between the luxCDABE operon and promoters corresponding to gyrB (PgyrB), yhlBA (Pyhl), yctCBA (Pyct), cdsAB (Pcds) and yrp1 (Pyrp1) genes were constructed. The corresponding plasmids were introduced in Y. ruckeri 150R and the resulting strains named Y. ruckeri 150RPyhl, Y. ruckeri 150RPyrp1, Y. ruckeri 150RPyct, Y. ruckeri 150RPcds and Y. ruckeri 150RPgyrB were analysed for bioluminescence emission.
A. In vitro expression of overnight cultures on NA at 18°C. A Y. ruckeri strain containing the plasmid pCS26-PacTER was inoculated on the right of each plate. Colour bars represent the scale for promoter activity quantification.
B. Quantification of in vivo expression of specific Y. ruckeri promoters. The swim bladder from 10 fish inoculated with each tagged strain was removed and imaged with the IVIS® Imaging System. ‘Total RLU’ was obtained. Then, the swim bladder was suspended in 200 μl of PBS and 0.1 ml aliquot was used to make decimal dilutions. From the 10−5 dilution, 0.1 ml was taken out and used to inoculate NA plates with kanamycin. The number of cfu per organ was determined. Data represent the mean and deviations of the RLU per 1000 cfu for the swim bladders analysed. The experiment was made in triplicate.
Download figure to PowerPoint
To be consistent with previous experiments in which in vivo-induced promoters were isolated by administering bacteria by intraperitoneal injection (Fernández et al., 2004) and taking advantage of the standardized manipulation provided by this route, groups of 25 fish were infected by intraperitoneal injection with 105 cfu of each bioluminescence-tagged strain (25 fish per strain). At the end of the infection process (48 h post infection) approximately 50% of fish inoculated had died. Among these, 10 fish from each of strains were collected and the swim bladders, as the reference organs, were analysed. In these organs, the relation between bioluminescence (‘total RLU’) and cfu was estimated. The results indicated that all the promoters showed activity inside the fish (Fig. 4B). The high expression of Pyhl, Pcds and Pyct promoters relative to the PgyrB constitutive promoter was expected since the three systems were isolated as in vivo-induced genes by in vivo expression technology (IVET; Fernández et al., 2004). Besides, two isogenic mutant strains in yhlA and cdsA genes encoding a haemolysin and cysteine transporter, respectively (Fernández et al., 2007b; Méndez et al., 2011), showed a considerable decrease in virulence indicating that they should be expressed for full bacterial virulence. Pyrp1 showed also higher bioluminescence in vivo. This is consistent with the results obtained with the reporter gene lacZ (Fernández et al., 2003) and the important attenuation displayed by the isogenic yrp1 mutant strain (Fernández et al., 2002). Finally, Y. ruckeri pCS26-PacTER used as negative control showed no luminescence inside the fish. The results confirmed that the four promoters Pcds, Pyhl, Pyct and Pyrp1 were induced inside the fish. The coherence of the results showed the effectiveness of this method for analysing gene virulence promoters in a fish model by intraperitoneal injection. All of the animal experiments were conducted under the European legislation governing animal welfare and were authorized and supervised by the Animal Experimentation Ethics Committee of Oviedo University.