This paper explains the application of fluorescent imaging using an activatable optical imaging probe to visualize the in vivo activity of key matrix metalloproteinases in two different experimental models of inflammation.
This paper describes a non-invasive method for imaging matrix metalloproteinases (MMP)-activity by an activatable fluorescent probe, via in vivo fluorescence optical imaging (OI), in two different mouse models of inflammation: a rheumatoid arthritis (RA) and a contact hypersensitivity reaction (CHR) model. Light with a wavelength in the near infrared (NIR) window (650 – 950 nm) allows a deeper tissue penetration and minimal signal absorption compared to wavelengths below 650 nm. The major advantages using fluorescence OI is that it is cheap, fast and easy to implement in different animal models.
Activatable fluorescent probes are optically silent in their inactivated states, but become highly fluorescent when activated by a protease. Activated MMPs lead to tissue destruction and play an important role for disease progression in delayed-type hypersensitivity reactions (DTHRs) such as RA and CHR. Furthermore, MMPs are the key proteases for cartilage and bone degradation and are induced by macrophages, fibroblasts and chondrocytes in response to pro-inflammatory cytokines. Here we use a probe that is activated by the key MMPs like MMP-2, -3, -9 and -13 and describe an imaging protocol for near infrared fluorescence OI of MMP activity in RA and control mice 6 days after disease induction as well as in mice with acute (1x challenge) and chronic (5x challenge) CHR on the right ear compared to healthy ears.
Autoimmune diseases such as rheumatoid arthritis (RA) or psoriasis vulgaris are graded as delayed-type hypersensitivity reactions (DTHRs).1 RA is a common autoimmune disease characterized by erosive synovitis and joint destruction.2 Inflamed arthritic joints demonstrate infiltration and proliferation of inflammatory cells, an increased expression of pro-inflammatory cells leading to pannus formation, cartilage and bone destructions.3,4 The cleavage of extracellular matrix molecules, such as collagen by matrix metalloproteinases (MMPs), is essential for tissue conversion and angiogenesis and causes tissue destructions.5,6 Contact hypersensitivity reactions (CHR) are characterized by aggregation of neutrophils leading to an oxidative burst.7 Similar to RA, MMPs in CHR are involved in tissue conversion, cell migration and angiogenesis in order to establish chronic inflammation.
To investigate RA, the glucose-6-phosphate isomerase (GPI)-serum injection mouse model was used.8 Serum from transgenic K/BxN mice containing antibodies against GPI, was injected into naïve BALB/c mice after which rheumatic inflammation began to develop within 24 h with a maximum of ankle swelling on day 6 after GPI-serum injection (see 1.1). To analyze chronic CHR, C57BL/6 mice were sensitized with trinitrochlorobenzene (TNCB) on the abdomen. The right ear was challenged up to 5 times starting 1 week after sensitization (see also 1.1 and 1.2).
Noninvasive small animal OI is a technique based on the in vivo investigation of fluorescent-, chemiluminescent- and bioluminescent-signals, which are mainly used in preclinical research. The acquired semi-quantitative data gives insights into the molecular mechanisms in the organs and tissues of healthy as well as diseased experimental animal models, and enables longitudinal follow up measurements (e.g. to assess therapeutic response profiles in vivo). A big advantage of longitudinal studies is the reduction of animal numbers, as the same animals can be measured in follow up studies at several time points instead of using different mice per time point. The resolution of OI allows detailed functional imaging of organs and even smaller tissue structures in experimental animals.
The use of specific excitation and emission filters with a narrow transmission spectrum, a protection against scattered light by a lightproof "dark box" and a sensitive charged-coupled device (CCD) camera, which is cooled in many devices down to -70 °C, allows highly specific and sensitive measurements of fluorescence signals.
By using fluorescent agents with excitation- and emission-spectra in the near-infrared fluorescence window (650 – 950 nm), signal-to-noise ratios can be improved significantly. The near-infrared fluorescence window is characterized by a relatively low absorption of the signal by hemoglobin and water as well as a low background auto-fluorescence.9 This allows a penetration depth of up to 2 cm in the tissue of small animals. OI-probes can address a target directly (e.g. by a fluorescence labeled antibody) or can be activated in the target tissue (e.g. by proteases). Activatable OI probes are optically silent in their inactivated form due to the Förster resonance energy transfer (FRET) to a quenching moiety, which transfers the excitation energy within the molecule to another domain. If the dye is cleaved (by a protease for example) the energy is no longer transferred within the molecule and a fluorescent signal can be detected by OI. This allows the design of OI probes with high specificity for distinct biological processes and excellent signal-to-noise-ratios.
The following protocol explains in detail the preparation of the animals, the OI measurements using an Activatable OI probe to image MMP-2, -3, -9 and -13 activity in vivo and two experimental models of inflammation (RA, CHR).
All procedures described in this paper, followed the guidelines and international standards of the care and use of laboratory animals and were approved by the local Animal Welfare and Ethics Committee of the Country Commission Tuebingen, Germany. 8 – 12 weeks old BALB/c and C57BL/6 mice were kept on a 12 h:12 h light:dark cycle and were housed in IVCs and standardized environmental conditions at 22 ± 1 °C in groups of 2 – 5 with water and food access ad libitum.
1. Material Preparation
2. Induction of Rheumatoid Arthritis and Chronic Contact Hypersensitivity Reaction
3. Animal Preparation for Optical Imaging
4. Optical Imaging
5. Data Analysis
NOTE: Perform data analysis using the image software following manufacturer's protocol.
To induce rheumatoid arthritis (RA) in naïve BALB/c mice, animals were injected i.p. with auto-antibodies (1:1 dilution with 1x PBS) against GPI on day 0. The maximum inflammation (ankle swelling) in this GPI-serum induced RA model is on day 6 post injection11. Therefore, 2 nmol of the activatable OI dye was prepared and injected i.v. in the tail vein of arthritic mice and healthy control animals on day 5. 24 h after injection (day 6), mice were anesthetized and imaged as described in in section 3 (Figure 1A).
To induce CHR, C57BL/6 mice were sensitized with 5% TNCB on day 0 at the shaved abdomen. On day 7 post sensitization the 1st challenge on the right ear using 1% TNCB started. Up to five challenges (day 15) were performed to induce chronic inflammation.7 12 h after the last challenge, 20 nmol of the OI dye was injected i.v. On day 15, 24 h post injection, OI was performed as described above (Figure 1B).
Figure 2 displays a clearly enhanced fluorescence signal in the hind paws, ankles and front paws of the RA mouse on day 6 after disease induction, compared to control ankles (Figure 2A). Biodistribution analysis of the MMP activity in organs of RA mice and healthy mice exhibited a strongly enhanced signal in the hind paws, ankles and fore paws exclusively in the joints of RA mice. Interestingly, we determined an enhanced MMP signal in both kidneys of all measured RA mice, when compared to the kidneys of healthy control mice (Figure 2A; right). Moreover, we measured an enhanced MMP signal in the liver and intestine regardless of whether mice suffer from RA. No differences between RA and healthy mice were found in the pancreas, lung, spleen and heart.
During chronic CHR (five TNCB challenges at the right ear), we measured highly increased MMP activity on the right inflamed ear compared to the healthy left control ear (Figure 2B).
Thus, our data is indicating a major role of MMP activity during disease progression in both inflammatory diseases models.
Toolsets for Bioluminescence and Fluorescence allow the quantification of, for example pmol of fluorochrome or the number of cells which express fluorescence. Data can be displayed in 1. Counts: displays an uncalibrated measurement of the photon incident on the camera, 2. Radiance (photons): indicates a calibrated measurement of the photon emission in "photons/second/cm2/steradian", which is recommended for luminescence imaging or 3. Radiant Efficiency (fluorescence): shows fluorescence emission radiance per incident excitation power, recommended for fluorescence measurements.
A big advantage of working with radiance units while analyzing OI images is, that camera settings can be changed during one experiment, and this has no impact on the images itself or the measured ROI data, which further allows comparing images from different IVIS systems.
The region of interest (ROI) indicates the specific area of interest from the user in an optical image. The toolbar enables to draw 3 different ROIs: measurement (measures the signal intensity in the ROI of the image), average background (measures the average signal intensity in the user defined area for the background), or subject ROI (identifies a subject animal in an image).
If the subject displays a high unspecific background signal itself, obtain a background-corrected ROI measurement by subtracting the background signal (ROI) from the target ROI. Set the image minimum close to zero to determine the auto fluorescence or naturally occurring background luminescence in your image. Link the image to a self-defined (user specific) background image (ROI) which were measured primarily to the measured specific image.
Semi-quantitative analyses of the images were performed by drawing standardized regions of interest (ROIs) in both arthritic and control ankles, as well as in inflamed and healthy ears using Living Image software. Only descriptive statistical analyses were performed on in vivo data in this manuscript, due to insufficient number of animals. Analysis of the signal intensity of arthritic and healthy ankles or paws showed a 7-fold enhanced MMP signal intensity in arthritic, compared to control ankles. A 3-fold higher MMP signal intensity was detected in the front paws of arthritic mice compared to that of the healthy animals (Figure 3A).
Analysis of ears with acute and chronic CHR compared to the left unchallenged control ear showed a significantly increased MMP signal even after the 1st challenge, which further increased after the 3rd and remained at the same level until the 5th challenge (Figure 3B). As a consequence of contamination with the contact allergen on the left ear due to scratching of the mice, we determined a slightly increased MMP signal in the left (not TNCB challenged) control ear. The results indicate an important role of MMPs, evinced by a highly increased fluorescence signal during GPI-arthritis and chronic inflammation, as measured by OI.
Figure 1: RA and CHR Model, Time Course of Disease Induction and Time Points for OI. (A) Time course of RA induction in BALB/c mice (B) and chronic CHR induction in C57BL/6 mice. Overview of the respective in vivo imaging time point and associated injection time points of the OI dye for measuring MMP activity in both models. Please click here to view a larger version of this figure.
Figure 2: MMP OI in RA, CHR and Control Animals. (A) Representative results of the measured MMP activity in an arthritic mouse 6 days after GPI-arthritis induction and a healthy control mouse (left side). Significantly higher signal (Radiance (p/s/cm2/sr)) intensity was observed in the arthritic compared to the healthy mouse. The image on the right displays the MMP activity in the hind paws and includes the ankles, fore paws, pancreas, lung, spleen, heart, kidneys, liver and intestine of a sacrificed GPI-arthritis mouse after OI dye injection. OI uncovered a high signal in the fore paws, ankles and hind paws of RA mice and the liver and intestine regardless of whether the mice were injected with GPI serum (RA mice) or control serum (healthy control mice). (B) MMP fluorescence signal in ears with chronic CHR (left) and control animals (right). The inflamed right ear (challenged 5 times) displays a highly enhanced signal compared to control ears. Please click here to view a larger version of this figure.
Figure 3: Semi Quantitative Analysis of Inflamed and Healthy Ankles and Ear Tissue Using OI. (A) Semi-quantitative analysis of the fluorescence signal in RA and healthy animals in the front paws and ankle joints. Arthritic front paws and ankle joints display a significantly (**p < 0.05) (n = 3) higher MMP signal compared to healthy ankle joints and front paws. The data is presented as the mean ± SD. (B) Semi-quantitative analysis of the left and right (challenged) ears after the 1st, 3rd and 5th challenge. MMP signal is increased significantly from the first to the 5th challenge and already showed a maximum signal intensity in the right challenged ear after the 3rd challenge (**p < 0.05) (n = 3) (Control ears are partly contaminated with TNCB due to the mice scratching. This produced an enhanced signal which was not significant). The data is presented as the mean ± SEM. The 2-tailed Student's t-test was used to analyze differences between inflammation (GPI-arthritic ankles, right treated ears) and control (control ankles, left untreated ears) for both models. Please click here to view a larger version of this figure.<!–
OI is a very useful, fast and inexpensive tool for non-invasive in vivo molecular imaging in preclinical research. A particular strength of OI is the capability to monitor highly dynamic processes like inflammatory responses. Moreover, OI allows one to follow the course of a disease for an extended period of time, ranging from days to weeks.
OI has several advantages over other in vivo imaging modalities such as positron-emission tomography (PET) or magnetic resonance imaging (MRI), as it is very time-, and cost-efficient. High-throughput analysis with up to five animals per acquisition is possible. Furthermore, a large number of probes, targeting many different biological processes, are available.12 To provide further anatomical information, OI can be combined with other imaging techniques such as MRI or x-ray. Another limitation is its low tissue penetration depth compared e.g. to functional PET imaging. Furthermore a large amount of probes aiming at many different biological targets are available.12
Effects of anesthesia should not be underestimated in in vivo imaging. While specific effects of anesthesia on OI results are not well characterized, our group has shown in several studies the influence of anesthesia on PET imaging results.13,14,15 For example, Fuchs et al. analyzed blood parameters, such as pCO2, pH and lactate before and after PET scans using different breathing and anesthesia protocols. It was shown that, significant changes in pCO2 and lactate levels in anesthetized compared to conscious oxygen or air breathing mice led to significant changes in PET results.14 They conclude that this effect is mainly caused by oxygen breathing and subsequent to respiratory acidosis in animals, leading to the different results in the PET imaging.
Standardization of anesthesia parameters, like O2-concentrations and doses of anesthetic drugs, as well as prevention of hypothermia help to avoid systematic failures. Of course, OI is limited by physical properties e.g. light scattering, tissue auto-fluorescence and light attenuation9. Several groups are addressing these problems by combining the macroscopic results of OI with other techniques like in vivo fluorescence microscopy or flow cytometry.
In vivo imaging of MMP activity by MMP-specific activatable fluorescent probes has been used in several experimental models of inflammation like CHS,7 but also in diseases like cerebral ischemia,16 aortic aneurysms,17 myocardial infarction18 and atherosclerotic plaques19 as well as various tumor models.20
For example, Cortez-Retamozo et al. investigated the in vivo activity of MMPs in a model of allergic airway inflammation by injecting an MMP activatable probe, 24 h after intranasal application of ovalbumin in previously sensitized mice.21 To identify the location of MMP activity further this group combined tomographic OI, NIR fiber optic bronchoscopy and intravital microscopy in order to monitor the course of disease and treatment response in greater detail.21
Other approaches to measure in vivo MMP activity mostly use inhibitors of MMP that bind to the active site of MMPs. McIntyre et al. described an in vivo tumor associated-MMP-7 detection using a "polymer-based fluorogenic substrate", which acts as a "proteolytic beacon" for MMPs in a mouse xenograft model.22 They show that the activity of MMP-7 could be detected selectively by optical imaging.23 Further, Olson et al. investigated "activatable cell penetrating peptides" (ACPPs) which are able to target many xenograft tumor models, but cannot adsorb into the cell until the linker is proteolyzed. They investigated ACPPs which are selective for MMP-2 and MMP-9.24,25
Labeling of MMP inhibitors with radioactive isotopes for single-photon emission computed tomography (SPECT) or PET imaging was used to study MMP expression in atherosclerosis26 and cancer27. In a recent study a fluorescence-labeled MMP inhibitor was compared to an MMP-activatable probe.28 The fluorescently-labeled MMP inhibitor showed a lower signal to noise ratio compared to an activatable probe but required a shorter uptake time.
New biomarkers to assess key features of pathophysiology can be utilized and validated by OI. For example, the response to danger associated molecular patterns (DAMPs) is an early event during an inflammatory reaction. Vogl et al. used a Cy5.5 labeled antibody against alarmin S100A9 in models of acute ear skin dermatitis, collagen-induced arthritis and infection with Leishmania major to develop sensitive biomarker for early signs of inflammation.29
In conclusion, OI represents a fast and efficient method for preclinical research, which is capable of uncovering new mechanisms of diseases in vivo, characterizing novel molecular targets, as well as monitoring therapeutic effects. Nevertheless, a subsequent validation process with more quantitative techniques like PET might be required.
The authors have nothing to disclose.
We thank Daniel Bukala, Natalie Altmeyer and Funda Cay for excellent technical support. We thank Jonathan Cotton, Greg Bowden and Paul Soubiran for editing the manuscript. This work was supported by the Werner Siemens-Foundation and the Medical Faculty of the Eberhard Karls University Tübingen (‘‘Promotionskolleg’’) and by the DFG through the CRC 156 (project C3).
Cornergel | Gerhard Mann GmbH | 1224635 | ophthalmic ointment |
Forene | Abbott GmbH | 4831850 | isoflurane |
U40 insulin syringe | Becton Dickinson and Company | 324876 | |
Heparin | Sintetica | 6093089 | |
High-Med-PE 0.28×0.61mm | Reichelt Chemietechnik GmbH+Co | 28460 | polyethylene tubing, inner diameter 0.28 mm, outer diameter 0.61 mm |
BD Regular Bevel Needles, 30 G | Becton Dickinson & Co. Ltd. | 305106 | 30 G injection cannula |
RTA-0011 isoflurane vaporizer | Vetland Medical Sales and Services LLC | – | |
Artagain drawing paper | Strathmore Artist Paper | 446-8 | coal black |
IVIS Spectrum | Perkin Elmer | 124262 | Optical imaging system |
BD Regular Bevel Needles, 25 G | Becton Dickinson and Company | 305122 | |
2-Chloro-1,3,5-trinitrobenzene | Sigma Aldrich GmbH | 7987456F | TNCB |
MMPSense 680 | Perkin Elmer | NEV10126 | fluorescent imaging dye |
Oditest | Koreplin GmbH | C1X018 | mechanical measurment |
Miglyol 812 | SASOL | – | Oil |
BALB/C, C57BL/6 | Charles River Laboratories | – | Mice used for experiements |
PBS | Sigma Aldrich GmbH | For dilution of the RA serum | |
Pipette (100µl) | Eppendorf | Used for TNCB application | |
shaver | Wahl | 9962 | Animal hair trimmer |
Living Image | Perkin Elmer | Imaging software to measure OI |