Abstract
Chlamydia trachomatis is a bacterial sexually transmitted disease with over 1.3 million cases reported to the CDC in 2010. While Chlamydia infection is easily treated with antibiotics, up to 70% of infections are asymptomatic and go untreated. The current mouse model relies on invasive upper genital tract gross pathology readouts at ~60–80 days postinfection. High throughput optical imaging through the use of biomarkers has been successfully used to quickly evaluate several disease processes. Here we evaluate Neutrophil Elastase 680 (Elastase680) for its ability to measure Chlamydia muridarum associated inflammation in live mice using fluorescence molecular tomography (FMT) and In Vivo Imaging System (IVIS). Optical imaging was able to distinguish with statistical significance between vaccinated and nonvaccinated mice as well as mock-challenged and challenged mice 2 weeks after challenge which was 9 weeks sooner than typical gross pathological assessment. Immunohistochemistry confirmed the presence of neutrophils and correlated well with both in vivo and ex vivo imaging. In this report we demonstrate that Elastase680 can be used as a molecular imaging biomarker for inflammation associated with chlamydial infection in a mouse model and that these biomarkers can significantly decrease the time for pathology evaluation and thus increase the rate of therapeutics discovery.
1. Introduction
Chlamydia trachomatis is the most common notifiable sexually transmitted infection with approximately 1.4 million cases reported to the Centers for Disease Control and Prevention in 2011 [1]. While Chlamydia urogenital infection can be effectively treated with antibiotic therapy, a majority of infections remain undetected and therefore go untreated. Chlamydia urogenital infection can lead to serious chronic disease consequences including urethritis/cervicitis at the mucosal site of infection, ascending infection resulting in pelvic inflammatory disease (PID), ectopic pregnancy, and tubal factor infertility (TFI) [2, 3].
Development of a vaccine for Chlamydia is widely considered the most effective approach for the prevention of infection and the associated disease sequelae. Well-characterized tools to support the effective evaluation of therapeutic intervention are prerequisite to the development of a Chlamydia vaccine and this includes tools for the evaluation of immune responses induced by vaccination and their impact on the course of infection. Equally important is an effective evaluation of the impact these experimental interventions have on the development of genital tract pathology.
Mouse models of reproductive tract infection with Chlamydia muridarum have proven useful for the study of immune mediators associated with bacterial clearance, development of disease pathogenesis, and the evaluation of vaccination efficacy [4–10]. Infection with C. muridarum in mice results in short-lived immunity that has been demonstrated to be dependent upon T-cell associated immune mechanisms for clearance, it leads to injury of the oviducts and reduction in fertility similar to the human condition, and it has been shown that repeated infection can elicit even more severe disease sequelae such as pelvic inflammatory disease (PID) and tubal factor infertility (TFI) like disease [10–13].
The exact mechanisms leading to tissue damage are not well understood. A cellular pathogenesis paradigm by which ineffective bacterial clearance (Chlamydia persistence) leads to release of proinflammatory cytokines from infected cells and resultant tissue damage is one prominent theory [14]. Alternatively, there is an immunological pathogenesis paradigm where repeated infection increases severity of the inflammatory response resulting in tissue damage at the focal sites of infection [14, 15]. It may be that both of these paradigms play some role in Chlamydial reproductive tract pathogenesis.
One hallmark of acute chlamydial genital tract infection in the mouse model is a large neutrophilic infiltrate, with both recruitment and longevity of this influx linked to severity of oviduct pathology [16, 17]. This parallels the observation in humans of elevated cervical sample neutrophil defensin levels in patients with Chlamydia-associated endometritis compared with uninfected subjects [18]. Further, primary infection with either wild type or plasmid-deficient C. muridarum strains followed by secondary challenge with a wild type strain resulted in a reduced frequency of neutrophil accumulation in the cervix and oviducts compared with primary wild type strain challenge [19]. Sterilizing immunity was not achieved by the prior challenge, yet there was a measured reduction in the inflammatory response and resultant tissue pathology. Taken together, these findings suggest that measurements of neutrophil infiltration following challenge would correlate with Chlamydia-related pathology in the mouse model, and measurement at key time points could be used to support the evaluation of vaccines or other novel therapeutic interventions.
Neutrophil elastase is a serine protease produced by activated neutrophils during the inflammatory response and has been used as an effective biomarker target [20, 21]. In this study, we evaluated the near-infrared optical imaging probe, Neutrophil Elastase 680 FAST (Elastase680) for the longitudinal evaluation of acute inflammation following primary Chlamydia infection and with Chlamydia challenge following vaccination. Near-infrared molecular imaging biomarkers have been used in the detection of inflammation in various animal models [22]. Use of these agents in conjunction with fluorescence molecular tomography (FMT) and computer tomography (CT) was applied for in vivo measurement and quantitative evaluation of the levels of inflammation in mouse models of atherosclerosis [23, 24]. In this study, signal from the Elastase680 probe was evaluated using both ex vivo (imaging of isolated genital tract tissues) and noninvasive live animal imaging approaches in comparison with standard histological tissue evaluation and immunohistochemical (IHC) tissue staining.
2. Methods
2.1. Ethics Statement
All experiments were approved by the Merck Institutional Animal Care and Use Committee (IACUC). All animal studies were conducted according to The Guide for the Care and Use of Laboratory Animals [25] and veterinary care was given to any animals requiring medical attention.
2.2. Animal Immunizations and Challenge
Albino C57B6 mice were purchased from Charles River and placed on alfalfa free diet (Dyets Inc., PA). Chlamydiae were propagated as described previously [26]. Briefly, mice were immunized with live Chlamydia muridarum Elementary Bodies (EB) at 1 × 106 EB/mouse in sucrose-phosphate-glutamate (SPG) buffer by the intraperitoneal (ip) route. Immunizations were administered on days 0, 20, and 30. At approximately 2 weeks following the last immunization, progesterone (medroxyprogesterone acetate, Depo-Provera; Pfizer; New York, NY) was administered subcutaneously (2.5 mg/dose) at 10 and 3 days before challenge. Mice were challenged intravaginally (approximately 1 month following last immunization) by direct instillation of 10 μL of SPG containing 1 × 104 C. muridarum EBs. The vaginal vault and ectocervix were swabbed using a microfiber swab (Fisher, Pittsburgh, PA) on days 7, 11, 14, 18, and 21 after challenge or some combination of these time points.
Swabs were placed into a 1.5 mL tube containing two sterile glass beads (5 mm diameter) and 300 μL of chlamydia isolation medium (Trinity Biotech, Berkeley Heights, NJ) on ice. Bacteria were eluted from the swabs and separated from cells by vortexing for 60 s. Eluted bacteria (100 uL) were plated onto a processing cartridge containing 100 uL of PBS and stored at −70°C until DNA extraction using the MagNA Pure 96 DNA and Viral NA small volume kit on the Roche MagNA pure machine (Roche, Indianapolis, IN). Real-time quantitative polymerase chain reaction (RT-qPCR) was performed to confirm the presence of infection as described previously [26].
2.3. Optical Imaging
At indicated time points, mice were shaved and Nair (Church and Dwight, Princeton, NJ) was used to remove residual hair on the ventral and dorsal abdominal region. Mice were injected intravenously with 2 nmol of Neutrophil Elastase 680 FAST (Elastase680, Perkin Elmer, Waltham, MA). Six hrs after injection mice were anesthetized with 2-3% Isoflurane (Isothesia, Henry Schein Animal Health, Dublin, OH) in O2 at a rate of 1 L/min, placed in the FMT holder, and then imaged in the dorsal position using an FMT 2500 imaging system (Perkin Elmer). In some studies, after live animal in vivo imaging, the mice were euthanized according to the AVMA Guidelines for the Euthanasia of Animals: 2013 Edition [27] and reproductive tracts harvested and imaged with the IVIS Spectrum (Perkin Elmer) system using 680 nm excitation and 750 nm emission filter sets before being processed for histomorphologic and immunohistochemical evaluation. Analysis was performed using Truequant (FMT) and Living Image (IVIS) software (Perkin Elmer, Waltham, MA). FMT data are presented as the amount of probe in pmol present in regions of interest (ROI). IVIS data are presented as Radiant Efficiency (photons/s)/(μW/cm2) which is a measure of photon flux from a source, normalized by exposure time, area of emission, and solid angle of the detector. Statistical analysis was performed using either one-way or two-way ANOVA on SigmaPlot software (Systat, San Jose, CA).
2.4. Micro-CT Imaging
Mice were imaged with micro-CT after FMT imaging. Mice were maintained under anesthesia and moved to the micro-CT scanner while in the FMT holder to maintain position. Micro-CT images were acquired using the eXplore Locus Ultra (TriFoil Imaging, Chatsworth, CA) with X-ray settings of 80 kV and 50 mA using the 16 sec anatomical protocol, which collects 1000 projections in a single 360° gantry rotation. The projection data for each scan was reconstructed using the AxRecon reconstruction system (Scanco, Switzerland). The reconstructed image slices consisted of 100 μm isotropic voxels scaled to Hounsfield units (HU). The FMT holder contains fiducial markers visible in both FMT and micro-CT datasets. FMT data were then exported to DICOM format using Truequant. Both micro-CT and FMT datasets were imported into Amira software (Visage Imaging, Andover, MA) which was used to coregister the data. A pseudo color scale based on fluorophore concentration was applied to the FMT data for better visualization.
2.5. Histopathology and Immunohistochemistry
After the final imaging session, tissues were fixed in 10% neutral buffered formalin for 24 hr, then processed, and embedded in paraffin by routine methods for histomorphologic evaluation and immunohistochemistry analysis. Replicate four-micron thick sections were cut and mounted onto positively charged slides. One slide set was stained with hematoxylin and eosin (H&E) according to routine methods for histopathologic examination. A second set was further processed for IHC staining. Briefly, heat-induced epitope retrieval (HIER) was performed with a commercial buffer (DIVA Decloaker; Biocare Medical, Concord, CA) using a pressure cooker. After cooling the slides were washed in dH2O, rinsed in phosphate buffered saline containing 0.1% Tween 20 (PBST, pH 7.2), and loaded onto an automated IHC stainer (intelliPATH, Biocare, Concord, CA). Endogenous peroxidase was blocked with 3% H2O2 for 20 min, slides were washed in PBST (3 × 3 min), and then incubated for 30 min in serum free blocking buffer (Background Sniper; Biocare). Slides were incubated in 1.0 ug/mL rat-anti mouse Ly-6B.2 alloantigen to detect neutrophils (MCA771GA; AbD Serotec, Raleigh, NC) for 1 hr and washed in PBST as before. The bound antibody was detected using a rat-on-mouse HRP-polymer kit (RT517; Biocare) according to the manufacturer’s instructions. After additional PBST washes the slides were incubated with diaminobenzidine substrate (DAB) for 5 min followed by counterstaining with hematoxylin.
Histomorphologic evaluation was conducted by a trained pathologist and results were peer-reviewed. Histomorphologic diagnoses were made separately for ovary and surrounding tissues, oviduct, uterus, cervix, and vagina and graded semiquantitatively on a scale of + to +++++ according to routine procedures (+ = very slight, ++ = slight, +++ = moderate, ++++ = marked, and +++++ = severe).
3. Results
3.1. Elastase Signal Localization and Time Course Following Chlamydia Genital Tract Infection
Mice were either challenged intravaginally with C. muridarum (naïve/challenged) as described or received a mock challenge with PBS. The term naïve is used here to be consistent with terminology used later in the paper to distinguish between vaccinated and nonvaccinated (naïve) animals. Seven, 14, 21, and 28 days following challenge, separate groups of mice were injected with Elastase680 and then euthanized after the washout period. The genital tracts were harvested and imaged ex vivo. One week following infection the Elastase probe signal was detected primarily in the uterine horns in Chlamydia-infected mice and not in PBS mock-infected mice. At 14 days the probe signal was detected primarily in the oviduct area, and by 21 days after challenge probe signal detection was reduced in all genital tract regions (Figure 1(a)). Region of interest (ROI) analysis was performed over the entire genital tract and used to quantitate the neutrophil response over the region. Seven and 14 days after challenge the quantified signal was significantly higher in Chlamydia-infected mice than in mock-infected mice (Figure 1(b)).
(a)
(b)
3.2. In Vivo Live Animal Imaging with Elastase680 Is a Biomarker at the Oviduct Level
Challenged and mock-challenged mice were also scanned using noninvasive imaging (live in vivo imaging). Figure 2(a) shows an example of an animal scan coregistered with a microCT scan of the same animal; the image shows the raw signal without any processing. Fiducial markers in the animal holder allowed for coregistration of the two modalities. Signal from the uterine horns during in vivo imaging was masked by excreted probe in the bladder and therefore analysis could only occur at the ovary/oviduct level. Signal intensity in this area peaked at approximately 14 days after challenge and was significantly higher than signal in a similar area in mock-infected animals (Figure 2(b)). ROI analysis on ex vivo images of the ovary/oviduct also showed signal intensity peaking 14 days after challenge (Figure 2(c)). There was a good correlation between ex vivo imaging and in vivo imaging (Figure 2(d)).
(a)
(b)
(c)
(d)
3.3. Elastase680 Is Able to Distinguish between Nonvaccinated and Vaccinated Animals
Next we performed longitudinal imaging on nonvaccinated and vaccinated mice to determine if noninvasive imaging can differentiate between these groups. Mice were either mock challenged with PBS (mock), challenged (naïve) with C. muridarum, or vaccinated with EBs (EB vaccinated) and then challenged with C. muridarum intravaginally as described above. Five mice from each group were euthanized at two weeks after challenge for ex vivo analysis while the remainder () were allowed to progress to genital tract gross pathology development and observational scoring (11 weeks postinfection). Day 14 after challenge was chosen for ex vivo evaluation because previous studies showed this to be the peak time point for the Elastase680 signal. At 14 days after challenge there was a statistical difference in quantified signal between vaccinated and naïve mice using ex vivo analysis on the ovary/oviduct region of the reproductive tracts with the Elastase680 probe (Figure 3(a)). However, this differentiation did not occur until day 14. In the longitudinal group, at earlier time points there was an observable trend but there was no statistical difference in the Elastase680 signal between naïve and vaccinated mice (Figure 3(b)). Both sets of mice had a higher Elastase680 signal than mock-challenged mice although vaccinated mice had a lower signal intensity than naïve mice at both day 7 and day 10 after challenge. Eighty days after challenge 80% of the naïve mice developed gross pathology while only 25% of the vaccinated mice developed gross pathology (Table 1). As there is a close association between neutrophil infiltration and the level of late stage pathology [17, 28] the findings at day 80 could have been predicted at day 14 when the differentiation between the treatment groups using Elastase680 signal was observed.
(a)
(b)
3.4. Immunohistochemistry and Histopathology Confirm Elastase680 Signal
To confirm that neutrophils were present in each section of the genital tract identified by imaging, we performed histopathology to estimate semiquantitatively the levels of acute inflammation on excised genital tracts. Table 2 summarizes the notable histopathological findings related to C. muridarum challenge in mice. A detailed table of pathological findings can be found in the supplemental sections (Table S1 in Supplementary Material available online at http://dx.doi.org/10.1155/2015/264897). Prior to 14 days, the majority of infected mice had histomorphologic changes that were acutely inflammatory in nature and localized primarily in the oviducts and/or the connective tissues around the oviducts and ovaries. Acute inflammation in the uterus was very slight to moderate and present mostly in the endometrium of the uterine horns. The acute inflammatory infiltrate was comprised primarily of neutrophils and lower levels of lymphocytes and macrophages in all areas. Approximately 14 days after challenge, similar acute inflammatory changes were present with a high incidence in the oviducts and/or periovarian areas in infected mice. At 21 days after challenge, inflammation in the periovarian tissues, oviducts, and uterus was more chronic in nature, with mostly lymphocytes and substantially fewer neutrophils infiltrating the tissues. Mock-infected mice had no notable histomorphologic changes in the reproductive tract as expected.
To further characterize the neutrophilic component of the inflammatory response observed by routine histomorphologic evaluation, we performed immunohistochemistry using an antibody (Ly6B.2) that specifically labels murine neutrophils (Figure 4, Table 3). Approximately seven days after challenge aggregates of Ly 6B.2 immunoreactive neutrophils were most predominant in the uterine horns and oviducts and in some of the surrounding stroma in the infected mice. As the infection progressed through 14 days the most abundant neutrophil aggregates were found to be associated with the oviducts and adjacent tissues as indicated by the severity score in Table 3. At day 21 after challenge, the number of Ly 6B.2 positive neutrophils was greatly reduced compared to earlier time points, consistent with the histopathology evaluation of routinely prepared sections. Both the percentage and intensity (percent area of positive staining) of Ly6B.2 were reduced in vaccinated mice verses naïve mice at all time points. Consistent with the in vivo imaging findings, the vaccinated mice did have neutrophil infiltration in the oviduct area at day 7 and day 10 after challenge and the level of neutrophil infiltration dropped off significantly at day 14. Overall immunohistochemistry and histopathology results correlated well with Elastase680 imaging results. Thus, noninvasive imaging using the neutrophil biomarker Elastase680 accurately reflected neutrophil infiltration in the ovary/oviduct area.
4. Discussion
Previous studies have shown gross pathology or histological evaluation following Chlamydia reproductive tract challenge and the effectiveness of live Chlamydia EB vaccination in preventing the damage associated with this infection [4–8, 11, 13, 19]. However, evaluation of Chlamydia-induced gross pathology and the effect of interventional strategies in the mouse model have been difficult to quantify, making effective comparison and development of treatments slow and challenging. Past approaches have included applying gross pathological observational scoring of hydrosalpinx and/or uterine dilatation at time points distal from challenge (70–80 days) or microscopic histopathology evaluation at earlier time points. Both of these approaches are terminal procedures capturing a point in time measurement and are subjective assessments. Development of a less subjective and quantitative method for tracking responses to C. muridarum early in the course of infection without the need to euthanize animals would allow faster and more thorough evaluation of C. muridarum pathogenesis and response to interventional therapies than is currently available. Furthermore, the longitudinal nature of these studies would allow for better interpretation of data, as each animal can act as its own control. In this study we have shown that Neutrophil Elastase 680 (Elastase680) based fluorescence molecular tomography (FMT) can be used to track inflammatory responses to C. muridarum in the mouse model early in the course of infection. We further demonstrated utility in application to Chlamydia vaccine evaluation.
Elastase680 is a noninvasive imaging neutrophil biomarker and has been previously shown to track the presence of infiltrating neutrophils associated with acute inflammation [20, 21]. Other studies have labeled Chlamydial strains with either a fluorescent dye [29] or bioluminescent protein [30] to visualize Chlamydial infection in mice. These studies provided the ability to track the bacteria in live animals but do not predict downstream pathology. This is an important parameter for vaccine discovery especially since different mouse strains have different responses to C. muridarum infection [31]. Studies have shown a close association of downstream pathology and the level of neutrophil recruitment in mouse models of C. muridarum infection [17, 28]. In vivo imaging using Elastase680 allowed for noninvasive monitoring of neutrophil infiltration associated with C. muridarum challenge. Peak signal in the upper genital tract of naïve/challenged animals occurred two weeks after challenge before declining to mock challenge levels by four weeks after challenge. This acute inflammatory time course correlated with neutrophil infiltration as determined by histopathology and immunohistochemistry and aligned with previous reports showing that peak inflammatory response occurs about two weeks following C. muridarum challenge [16, 17, 19].
One of the limitations of using Elastase680 in vivo in this model is its clearance through the bladder where the probe accumulates before being excreted from the body. This excess signal obscures any detectable signal in the uterine horns. However, signal located in the uterine horns was easily detectable upon removal of the genital tract and analysis ex vivo. At one week after challenge Elastase680 signal is first detectable in the uterine horns of naïve/challenged animals and neutrophil presence was confirmed with histopathology and immunohistochemistry and supported by previous studies [16]. However, in vivo evaluation of the oviduct region was ultimately sufficient for the purposes of tracking C. muridarum infection and predicting outcome in the mouse model.
Elastase680 was also able to distinguish between EB vaccinated and unvaccinated mice. At two weeks after challenge in EB vaccinated mice, little or no Elastase680 signal was present in the oviduct region. Again, this was confirmed with histopathology and immunohistochemistry which demonstrated few or no neutrophils present. However, at earlier time points EB vaccinated animals had showed a transient increase in Elastase680 signal, indicating that there was some transient neutrophil infiltration in EB vaccinated animals. In animals that were longitudinally followed with imaging and allowed to progress to pathology at day 80 after challenge there was no pathology observed in the EB vaccinated animals. Studies have linked the duration of time neutrophils are present with subsequent pathology in mouse models [17, 28]. Noninvasive imaging may enhance the study of this link by monitoring the same animal over time before pathological evaluation. In addition, mice can be kept alive for fertility studies and Elastase680 could serve as an earlier biomarker for vaccine efficacy. An additional benefit of the use of longitudinal, noninvasive imaging is a reduction in the number of mice required for studies, since mice do not need to be euthanized at time points during the course of the study.
Optical imaging using Elastase680 to monitor acute inflammation caused by C. muridarum infection can be applied in a consistent manner to reduce study time and animal numbers or to evaluate individual mice over time. Recently it was reported that plasmid free C. muridarum can also cause hydrosalpinx in some strains of mice [32]. Since hydrosalpinx is a chronic pathology that results from acute inflammation it is possible that Elastase680 could be used to aide in discovery of the levels of hydrosalpinx caused by different strains of C. muridarum. More importantly imaging is quantitative and not subjective. Not only can imaging be applied to investigate reduction in pathology with vaccination, but also it may have utility in monitoring the possibility of enhanced pathology with ineffective vaccination as was observed with some early studies in both nonhuman primates and human volunteers with chlamydia EB vaccines [33–35]. Imaging with Elastase680 and other developing immunological probes provides new biomarker tools to support the effective evaluation of pathogenesis following chlamydial infection and the effect of vaccination.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Supplementary Materials
A detailed histopathologic summary of findings in reproductive tracts of female mice infected with C. muridarum. Mice were sacrificed at the indicated time point and the reproductive tracts were prepared for histopathology assessment as described in the methods. The data presented here is a compilation of several studies. The numbers indicate the percentage of mice that had findings. Severity was graded semi-quantitatively on a scale of 1 to 5 according to routine procedures (1=very slight, 2=slight, 3=moderate, 4=marked, and 5=severe). Acute inflammation was mostly neutrophilic while chronic inflammation was lymphocytic in nature.