Evaluation of Biophotonic Imaging To Estimate Bacterial Burden in Mice Infected with Highly Virulent Compared to Less Virulent Streptococcus pneumoniae Serotypes

Abstract

Bioluminescence imaging is an innovative, noninvasive tool to analyze infectious disease progression under real-life conditions in small laboratory animals. However, the relevance of bioluminescence imaging to monitor invasive compared to noninvasive bacterial infections of the lung has not been examined so far. In the current study, we systematically evaluated the importance of bioluminescence imaging to monitor pneumococcal disease progression by correlating biophotonic signals with lung bacterial loads in two mouse strains (BALB/c, C57BL/6) infected with either self-glowing, bioluminescent serotype 19 Streptococcus pneumoniae causing focal pneumonia or serotype 2 S. pneumoniae causing invasive pneumococcal disease. The best correlations between bioluminescence signals and lung CFU counts were observed in BALB/c mice compared to C57BL/6 mice just on day 3 after infection with invasive serotype 2 S. pneumoniae, while excellent correlations between photon counts and bacterial loads were observed in isolated lungs of BALB/c and C57BL/6 mice, irrespective of the employed pneumococcal serotype. Moreover, good correlations between biophotonic signals and CFU counts were also observed in mice upon infection with serotype 19 S. pneumoniae causing focal pneumonia in mice, again with best correlation values obtained for BALB/c mice at day 3 postinfection. Collectively, we show that the relevance of biophotonic imaging to monitor S. pneumoniae-induced lung infections in mice is largely influenced by the disease model under investigation. The provided data may be important for studies of infectious diseases.

The Gram-positive bacterium Streptococcus pneumoniae is the most prevalent pathogen in community-acquired pneumonia (CAP). The worldwide increase in antibiotic resistance of S. pneumoniae against frequently used antibiotics and the rapid global spread of multidrug-resistant clones require novel immunization strategies and antibiotic substances (17). Temporal information about novel drug or immunization efficacies to control disease progression is conventionally characterized by repeated, time-stacked animal sacrifice, e.g., for determination of bacterial loads (4). However, more cost-efficient and time-saving methods are needed for therapeutic and preventive screening approaches in laboratory animals (1, 7, 17). In this regard, bioluminescence imaging (BLI) technology is emerging as a powerful tool for monitoring real-time progression of infectious diseases, making it the most commonly used imaging technique to study drug efficacy profiles in infectious disease models (2, 3, 9-12, 16). However, to the best of our knowledge, no studies so far have systematically evaluated the reliability of bioluminescence imaging to accurately reflect infectious disease progression in different mouse lines and different infection models.

In the current study, BALB/c and C57BL/6 mice were infected either with self-glowing, bioluminescent capsular serotype 19 S. pneumoniae, known to trigger focal pneumonia in mice, or with self-glowing capsular serotype 2 S. pneumoniae, known to rapidly cause invasive pneumococcal disease, followed by correlation analysis of biophotonic signals with respective bacterial CFU for up to 3 days postinfection.

MATERIALS AND METHODS

Mice.

Female C57BL/6 and BALB/c mice were purchased from Charles River (Sulzfeld, Germany) and were kept under conventional conditions with free access to food and water. Mice were used in all experiments at 8 to 12 weeks of age in accordance with the guidelines of our Institutional Animal Care and Use Committee. Animal experiments were approved by our local government authorities. The main reasons for choosing these two mouse strains were that they are the most commonly employed inbred mouse strains in disease models and that their contrasting fur colors (black and white) are ideally suited to study the impact of fur color on bioluminescence imaging approaches.

Culture and quantification of S. pneumoniae.

In the current study, we employed two different bioluminescent, “self-glowing” strains of S. pneumoniae, i.e., a highly virulent serotype 2 S. pneumoniae strain (D39 lux), which has been shown to rapidly cause invasive disease progression in mice (9, 14), and a less virulent capsular group 19 S. pneumoniae strain (EF3030 lux), which is known to primarily cause focal pneumonia in mice (9, 20). Both strains of S. pneumoniae were transformed with essentially the same plasmid expressing the luxABCDE operon of Photorhabdus luminescens and an erythromycin resistance cassette for positive selection purposes and were generated as recently described (2, 3). Importantly, constitutive bioluminescence emission of the employed pneumococcal strains did not require application of a substrate (“self-glowing” strains).

Bioluminescent serotype 2 S. pneumoniae D39 lux was grown in Todd-Hewitt broth (THB) (Oxoid, Basingstoke, United Kingdom) supplemented with 0.5% yeast extract (Difco/BD Biosciences, Sparks, MD) in the presence of 0.2 μg/ml erythromycin (Sigma-Aldrich, Steinheim, Germany). Serotype 19 S. pneumoniae EF3030 lux was grown in Todd-Hewitt broth supplemented with 10% fetal calf serum (FCS) in the presence of 0.2 μg/ml erythromycin. Aliquots were snap-frozen in liquid nitrogen and stored at −80°C. S. pneumoniae stocks were quantified by plating serial dilutions of the bacteria on sheep blood agar plates (BD Biosciences, Heidelberg, Germany), followed by incubation of the plates at 37°C and 5% CO₂ for 18 h and subsequent determination of CFU.

Infection of mice with S. pneumoniae.

Mice were anesthetized with tetrazoline hydrochloride (5 mg/kg) and ketamine (75 mg/kg). Subsequently, mice were orotracheally intubated with a 29-gauge Abbocath (Abbott, Wiesbaden, Germany), which was inserted into the trachea under visual control with transillumination of the neck region. Lung infection was induced by orotracheal instillation of either 1.5 × 10⁷ CFU of serotype 2 S. pneumoniae (D39 lux) or serotype 19 S. pneumoniae (EF3030 lux) in a volume of 50 μl THB to trigger either invasive pneumococcal disease or focal pneumonia in mice. Subsequently, mice were brought back to their cages with free access to food and water.

Bioluminescence imaging.

Groups of infected mice were subjected to bioluminescence analysis on days 1, 2, or 3 of infection using a Xenogen bioluminescence analyzer (Xenogen, Alameda, CA), essentially as described recently (9). Regions of interest (ROI) were manually selected, and the results were quantified as average radiance of photons emitted per second, area, and steradian (p/s/cm²/sr) by using the IgorPro image analysis software package (Fig. 1). Regions of interest generated for in vivo analysis were identical between mice, and respective ROIs generated for ex vivo bioluminescence analysis of lungs were adjusted to the size and shape of individual lung lobes, overall requiring only minimal adjustments between lung lobes of different mice. All correlation studies performed were done with the average radiance of photons after subtraction of background signals obtained from lungs or lung lobes of uninfected control mice. Under the chosen experimental conditions, in vivo background photon counts of lungs of untreated C57BL/6 mice ranged between 80 and 300 p/s/cm²/sr and 400 and 1,000 p/s/cm²/sr for lungs of untreated BALB/c mice. For further details, see the supplemental material.

FIG. 1.

Gating strategy for in vivo and ex vivo quantification of bioluminescence signals from lungs of S. pneumoniae-infected mice. Mice infected with self-glowing serotype 19 or 2 S. pneumoniae (1.5 × 10⁷ CFU/mouse) were subjected to analysis of bioluminescence emission using a Xenogen bioluminescence analyzer. (A) Pseudocolor image overlay of a grayscale photograph of anesthetized BALB/c (left) and C57BL/6 (right) mice infected with S. pneumoniae with manually selected regions of interest (ROIs) (in vivo gating). (B) Manually selected ROIs of pseudocolor images of lung lobes of a mouse infected with S. pneumoniae (ex vivo gating).

To evaluate the contribution of bioluminescent bacteria located within the lung capillary system, relative to the bronchoalveolar space, to overall lung bioluminescence emission of mice challenged with invasive serotype 2 S. pneumoniae, in selected experiments, lungs of serotype 2 S. pneumoniae (D39 lux)-infected mice either were left untreated or were treated by perfusion in situ with sterile Hanks' balanced salt solution (HBSS) (5 ml) or bronchoalveolar lavage (BAL) prior to their ex vivo biophotonic imaging analysis. Whole-lung washes were performed as recently described, with instillation of 300-μl aliquots of ice-cold sterile PBS followed by careful aspiration, until a total BAL fluid volume of 6 ml was collected (13, 15).

Determination of bacterial loads.

Subsequent to bioluminescence analysis performed on days 1, 2, and 3 postinfection, bacterial loads were determined in lungs of S. pneumoniae-infected mice. For determination of bacteremia, peripheral blood was collected from the inferior vena cava followed by plating of serial dilutions on sheep blood agar plates. For determination of CFU in lungs, organs were dissected and homogenized in Hanks' balanced salt solution without supplements using a tissue homogenizer (IKA, Staufen, Gemany). Resulting homogenates were filtered through a 100-μm cell strainer (BD Falcon), and aliquots of each sample were then plated in 10-fold serial dilutions on sheep blood agar plates followed by incubation at 37°C and 5% CO₂ for determination of bacterial loads, as recently described in detail (14, 19).

Statistics.

Correlation analysis of in vivo or ex vivo collected bioluminescence signals and CFU counts was performed by Spearman rank correlation test using GraphPadPrism 5. Statistically significant differences were assumed when P values (two-tailed) were at least ≤0.05.

RESULTS

Evaluation of bioluminescence imaging of pneumococcal pneumonia in mice.

We first analyzed the correlation of bioluminescence signals with bacterial loads in a model of focal pneumococcal pneumonia by infection of C57BL/6 mice with less virulent serotype 19 S. pneumoniae (1.5 × 10⁷ CFU/mouse). As shown in Fig. 2 and Fig. S1 in the supplemental material, more than 50% of lungs of mice analyzed on days 1 and 2 postinfection were found to emit either very low (photon counts of <5 × 10² p/s/cm²/sr) or no bioluminescence signals (Fig. 2A and B; see Fig. S1A and B in the supplemental material), resulting in correlation coefficients of 0.77 and 0.84, respectively, between photon counts and bacterial loads (Fig. 2A and B). When analyzed at later time points postinfection (day 3), in vivo bioluminescence signals emanating from lungs of mice infected with serotype 19 S. pneumoniae correlated even more strongly with CFU counts (r = 0.92), with only small differences relative to correlation coefficients of ex vivo-analyzed lung lobes from the same mice (r = 0.92 versus r = 0.87) (Fig. 2C; see Fig. S1C in the supplemental material). In contrast, isolated lung lobes of the same mice were found to exhibit strong bioluminescence signals, with bacterial counts ranging between 10⁴ and 10⁷ CFU/mouse lung (Fig. 2A and B; see Fig. S1A and B in the supplemental material), overall resulting in strong correlations between photon emission and CFU counts (Fig. 2A and B). Overall, we observed that ex vivo-analyzed mouse lungs (days 1 and 2) exhibited approximately 10³- to 10⁴-fold increased photon counts relative to in vivo-analyzed mouse lungs (Fig. 2A to C).

FIG. 2.

Biophotonic analysis of developing focal pneumococcal pneumonia. Mice were infected with self-glowing serotype 19 S. pneumoniae (1.5 × 10⁷ CFU/mouse) and subjected to bioluminescence analysis on day 1, 2, or 3 postinfection (in vivo analysis), using a Xenogen bioluminescence analyzer. Immediately after in vivo biophotonic imaging, mice were sacrificed and individual lung lobes of each mouse were subjected to bioluminescence analysis (ex vivo analysis) on day 1, 2, or 3 postinfection, followed by quantification of the respective bacterial loads. (A to C) Correlation of bioluminescence signals and CFU counts of the same lungs of C57BL/6 mice on day 1 (A), day 2 (B), and day 3 (C) after infection with serotype 19 S. pneumoniae. (D to F) Correlation of bioluminescence signals and CFU counts of the same lungs of BALB/c mice on day 1 (D), day 2 (E), and day 3 (F) after infection with serotype 19 S. pneumoniae. r, Spearman correlation coefficient; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

In the next set of experiments, we performed a bioluminescence analysis of BALB/c mice infected with self-glowing, less virulent serotype 19 S. pneumoniae (1.5 × 10⁷ CFU/mouse). As shown in Fig. 2D and Fig. S1D in the supplemental material, virtually all S. pneumoniae-infected BALB/c mice showed detectable bioluminescence signals emanating from lungs as early as 24 h postinfection, together with a significant correlation (r = 0.82) of in vivo bioluminescence signals with CFU counts (Fig. 2D; see Fig. S1D in the supplemental material). At the same time, ex vivo bioluminescence signals of isolated lungs infected for 24 h with S. pneumoniae also resulted in slightly superior correlations (r = 0.9) with CFU counts, compared to in vivo correlation analysis (Fig. 2D). At day 2 postinfection, ex vivo photon counts still correlated well with CFU of lungs (r = 0.93), whereas in vivo bioluminescence signal and respective CFU analysis gave a correlation coefficient of 0.5 only at this time of infection (Fig. 2E). As shown in Fig. 2F (also see Fig. S1F in the supplemental material), strong correlations between in vivo and ex vivo bioluminescence signals and bacterial loads were observed in mice on day 3 postinfection, albeit with a strong interindividual variability between mice, indicating differences between mice to cope with infection. As in our observations of C57BL/6 mice, we observed that ex vivo bioluminescence analysis of lungs of BALB/c mice revealed approximately 100- to 1,000-fold higher photon counts relative to in vivo bioluminescence counts of the same mice (Fig. 2D to F).

Bioluminescence imaging of invasive pneumococcal disease in mice.

To characterize the impact of bioluminescence imaging for studying different clinical courses of pneumococcal lung infection, we next infected C57BL/6 mice with highly invasive serotype 2 S. pneumoniae, which is known to rapidly cause invasive pneumococcal disease in mice. Similar to our observation with serotype 19 pneumococcal pneumonia, C57BL/6 mice infected with serotype 2 S. pneumoniae (1.5 × 10⁷ CFU/mouse), though exhibiting bacterial loads of 10⁵ to 10⁷ CFU/lung, demonstrated a very weak bioluminescence emission on day 1 postchallenge (Fig. 3A; see Fig. S2A in the supplemental material), nevertheless yielding significant correlation coefficients at this time point of investigation. In contrast, ex vivo bioluminescence emission of individual lung lobes of the same mice resulted in strongly increased bioluminescence profiles and significant correlations with respective CFU counts (Fig. 3A). At days 2 and 3 of infection, we found superior correlations between photon emission and CFU counts under both in vivo and ex vivo conditions of bioluminescence analysis (Fig. 3B and C). At the same time, differences in bioluminescence signals (ranging up to 10³- to 10⁴-fold) between ex vivo- and in vivo-analyzed lungs of mice declined over time (Fig. 3A to C). Notably, ex vivo bioluminescence imaging allowed detection of bacterial loads as low as ∼10⁴ CFU/lung, whereas <10⁵ CFU/lung were not adequately reflected by in vivo bioluminescence emission (Fig. 3A to C; see Fig. S2 in the supplemental material).

FIG. 3.

Biophotonic analysis of invasive pneumococcal disease. Mice were infected with self-glowing serotype 2 S. pneumoniae (1.5 × 10⁷ CFU/mouse) followed by bioluminescence imaging of disease progression on days 1, 2, and 3 postchallenge. Subsequent to analysis of whole mouse lungs (in vivo analysis), mice were euthanized and individual lung lobes were subjected to bioluminescence analysis (ex vivo analysis). In addition, lungs of the respective mice were further processed for determination of bacterial loads. (A to F) Correlation of bioluminescence signals with CFU counts of the respective lungs of C57BL/6 (A to C) and BALB/c (D to F) mice on day 1 (A and D), day 2 (B and E), and day 3 (C and F) after infection with serotype 2 S. pneumoniae. r, Spearman correlation coefficient; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Next, we evaluated bioluminescence imaging of lungs of BALB/c mice infected with highly invasive serotype 2 S. pneumoniae. Interestingly, 10³ to 10⁵ p/s/cm²/sr were detected in mice in vivo as early as 1 day postinfection, along with CFU counts ranging from 5 × 10⁶ to 5 × 10⁷ CFU/lung, though not resulting in a significant correlation (Fig. 3D; see Fig. S2D in the supplemental material). Again, strong differences between in vivo and ex vivo bioluminescence signal intensities were noted in BALB/c mice after serotype 2 S. pneumoniae challenge at day 1 postinfection (Fig. 3D), a result which, however, was found to rapidly disappear over time (Fig. 3E and F). Excellent correlations between ex vivo bioluminescence signals and CFU counts were obtained on day 2 and particularly on day 3 postinfection (r = 0.98 for day 2, r = 1.0 for day 3). More importantly, correlation coefficients between in vivo bioluminescence signals and CFU counts also increased over time, resulting in perfect correlations for mice infected for 3 days with invasive serotype 2 S. pneumoniae (Fig. 3D to F), thereby illustrating the power of bioluminescence imaging for monitoring invasive pneumococcal disease models.

Effect of bacteremia on lung bioluminescence signal intensity in BALB/c mice infected with invasive serotype 2 S. pneumoniae.

Invasive serotype 2 S. pneumoniae is known to rapidly evade local lung host defense to trigger bacteremia and sepsis in mice (14). Therefore, to estimate to what degree bacteremia would contribute to biophotonic imaging of lungs of mice infected with serotype 2 S. pneumoniae, lungs of mice were either left untreated, perfused, or subjected to bronchoalveolar lavage immediately prior to bioluminescence analysis. Notably, all of the mice included in this analysis were confirmed to be highly bacteremic without significant differences between groups (data not shown). Notably, as shown in Fig. 4, analysis of bioluminescence signals of untreated compared to perfused lungs of serotype 2 S. pneumoniae-infected mice revealed 3 × 10⁵ ± 1 × 10⁵ versus 2 × 10⁴ ± 3 × 10³ p/s/cm²/sr (mean ± standard error of the mean [SEM]; n = 5 mice), thus demonstrating that the major portion of photon counts of lungs of mice challenged with invasive serotype 2 S. pneumoniae are derived from bioluminescent pathogens within the lung capillary system but not the bronchoalveolar space. In line with this, bronchoalveolar lavage of lungs of mice infected with serotype 2 S. pneumoniae only weakly affected overall lung photon counts (2 × 10⁵ ± 7 × 10⁴ p/s/cm²/sr), thus demonstrating minimal contribution of bronchoalveolar pathogens to total bioluminescence emission of lungs infected with invasive S. pneumoniae.

FIG. 4.

Impact of bacteremia on lung bioluminescence imaging in invasive pneumococcal disease. BALB/c mice were infected with self-glowing serotype 2 S. pneumoniae (1.5 × 10⁷ CFU/mouse) for 3 days. Subsequently, mice were euthanized and lung lobes were either left untreated, subjected to bronchoalveolar lavage, or perfused with HBSS prior to analysis of biophotonic signals. The shown data represent the mean ± SEM of results for 5 mice per treatment regimen.

DISCUSSION

The current study aimed at evaluating the use of bioluminescence imaging for monitoring of pneumococcal disease in two mouse models of either pneumococcal pneumonia induced by less virulent serotype 19 S. pneumoniae or pneumococcal sepsis induced by highly virulent serotype 2 S. pneumoniae in C57BL/6 compared to those in BALB/c mice. In vivo bioluminescence imaging was found to equally reflect bacterial loads and disease progression in BALB/c mice and C57BL/6 mice on days 1 and 2 of infection with highly virulent serotype 2 S. pneumoniae, whereas best correlations between photon and CFU counts were observed in BALB/c mice on day 3 of infection with serotype 2 S. pneumoniae. Overall, detection of bioluminescence signals per se during the observation period of 3 days appeared more sensitive in BALB/c than in C57BL/6 mice. Weakest bioluminescence signal detection was obtained in C57BL/6 mice infected with focal pneumonia-inducing serotype 19 S. pneumoniae, which, however, was found to be still sufficient to allow a thorough correlation analysis of photon counts with CFU under these conditions, similar to corresponding analyses in BALB/c mice. Together, these results imply that the importance of bioluminescence imaging for estimating bacterial burdens in the lungs of mice primarily depends on the pneumococcal serotype and, albeit to a lesser extent, on the mouse background employed.

Various biophysical parameters and factors have been described to affect bioluminescence imaging in mice. One important factor is the availability of oxygen at the microanatomical site of infection, which has been shown to strongly impact bioluminescence imaging. Also, potential signal impedance can occur due to oxyhemoglobin- and deoxyhemoglobin-dependent light absorption (4, 5, 8, 11, 18). Moreover, recent reports suggest that absolute values of bioluminescence produced by a defined number of bacteria differ in various anatomical sites because of absorption and scattering of light by overlying tissue. Finally, sensitivity of bioluminescence imaging can also be decreased by skin and fur pigmentation (6, 11). However, the power of bioluminescence imaging to monitor (and differentiate between) invasive compared to noninvasive bacterial infections of the lung has not been evaluated so far.

We observed that under all experimental conditions examined, bacterial loads of lungs comparable to the initial infection dose of ∼10⁷ CFU/mouse were clearly detectable by bioluminescence imaging in mice throughout the observation period. In contrast, in most of the mice subjected to bioluminescence analysis immediately after intratracheal delivery of 1.5 × 10⁷ CFU S. pneumoniae, virtually no bioluminescence emission was detectable (data not shown). This discrepant observation may be due to the fact that upon intratracheal instillation, bacteria are distributed along the trachea and the bronchial tree toward the lower respiratory tract and distal air spaces, thereby causing a “dilution effect” of instilled pneumococci within the lungs, making them undetectable by bioluminescence imaging. In addition, considerable numbers of these bacteria will be removed from the lung by mucociliary clearance mechanisms, and just those bacteria outgrowing within lung distal air spaces to cause pneumonia or invasive disease may contribute to overall photon emission. Therefore, depending on the employed infection dose, pathogen, and model system, bioluminescence imaging of bacterial infections in mice should be initiated no earlier than 24 h postinfection.

It is currently poorly characterized how bioluminescence imaging of pneumonia is affected by the virulence profile of the underlying pathogen. In the current study, we observed that bioluminescence imaging of pneumococcal pneumonia was less effective than imaging of invasive pneumococcal disease. In this context, we recently observed that less invasive serotype 19 S. pneumoniae also employed in the current study caused a significant mortality in mice that was due to lung permeability and intra-alveolar cytokine storms but not due to bacteremia and sepsis (9). Lung permeability is known to cause respiratory failure and death in patients with acute respiratory distress syndrome (ARDS) and, due to lowered oxygen availability, may also limit the photon generation by self-glowing serotype 19 S. pneumoniae in mice. Moreover, since serotype 19 pneumococci trigger the formation of consolidated infiltrates within distal air spaces characteristic of focal pneumonia, such inflammatory infiltrates may biophysically hamper detection of bacterium-derived photon emission by the charge-coupled-device (CCD) camera, thereby additionally lowering the sensitivity of in vivo bioluminescence imaging of focal pneumonia in mice.

On the other hand, we recently observed that highly invasive serotype 2 S. pneumoniae did not cause lung permeability or intra-alveolar cytokine storms in mice but rather rapidly escaped local lung host defense to cause systemic cytokine storms, bacteremia, and sepsis in mice (9). However, life-threatening invasive bacterial infections are known to drastically accelerate catabolic lipid metabolism, resulting in a rapid loss of body weight in mice. Thus, bacterial shift toward the lung capillary system, underlying bacteremia, and sepsis, together with resulting cachexia-like phenomena, may, at least in part, make bioluminescence imaging approaches more effective and sensitive in invasive infectious disease models, although in the current study, we did not see differences in body weight between serotype 2 S. pneumoniae-infected C57BL/6 and BALB/c mice (data not shown). In addition, increasing piloerection developing concomitantly with septic disease progression may be another factor in bioluminescence imaging, because fur standing on end may quench photon emission to a lesser extent than fur lying flat, analogous to the effects of fur pigment.

In the current study, we observed that photon signals collected from the lungs of mice at day 1 of infection with serotype 2 or 19 S. pneumoniae differed from those collected ex vivo but approximated each other only on days 2 and 3 of serotype 2 S. pneumoniae challenge. In contrast, in serotype 19 S. pneumoniae-induced focal pneumonia not progressing toward sepsis, discrepancies between in vivo- and ex vivo-collected lung photon counts were maintained over time. These data strongly suggest that bioluminescence imaging of invasive lung disease performed at early time points postinfection (day 1) appears to reflect bacterial loads within the alveolar compartment but when performed at later time points postinfection (day 3) appears to largely represent bacterial loads of the lung capillary system rather than the lung itself. This concept is clearly supported by the finding that perfusion of lungs challenged with invasive serotype 2 S. pneumoniae led to ∼90% decreased ex vivo-collected lung photon counts relative to photon counts collected from nonperfused mouse lungs, whereas the effect of bronchoalveolar lavage on lung bioluminescence emission appeared negligible in the described model system.

In summary, this is the first report to address the importance of bioluminescence imaging for monitoring pneumococcal lung infection caused by different serotypes of S. pneumoniae in two different mouse lines. The best reflection of bacterial loads by bioluminescence signals with minimal differences between in vivo- and ex vivo-collected photon counts were observed in BALB/c mice challenged with highly invasive serotype 2 S. pneumoniae. We conclude that bioluminescence imaging offers an alternative method to classical approaches to estimate bacterial burdens in mice, particularly in sepsis-related research areas, characterized by bacteremic disease progression, with a major impact on ethical, time, and cost considerations, and will be particularly helpful for improved preclinical characterization of anti-infective agents and their effectiveness.