Monitoring Dendritic Cell Migration using 19F / 1H Magnetic Resonance Imaging

Summary

Tracking of cells using MRI has gained remarkable attention in the past years. This protocol describes the labeling of dendritic cells with fluorine (¹⁹F)-rich particles, the in vivo application of these cells, and monitoring the extent of their migration to the draining lymph node with ¹⁹F/¹H MRI and ¹⁹F MRS.

Abstract

Continuous advancements in noninvasive imaging modalities such as magnetic resonance imaging (MRI) have greatly improved our ability to study physiological or pathological processes in living organisms. MRI is also proving to be a valuable tool for capturing transplanted cells in vivo. Initial cell labeling strategies for MRI made use of contrast agents that influence the MR relaxation times (T1, T2, T2*) and lead to an enhancement (T1) or depletion (T2*) of signal where labeled cells are present. T2* enhancement agents such as ultrasmall iron oxide agents (USPIO) have been employed to study cell migration and some have also been approved by the FDA for clinical application. A drawback of T2* agents is the difficulty to distinguish the signal extinction created by the labeled cells from other artifacts such as blood clots, micro bleeds or air bubbles. In this article, we describe an emerging technique for tracking cells in vivo that is based on labeling the cells with fluorine (¹⁹F)-rich particles. These particles are prepared by emulsifying perfluorocarbon (PFC) compounds and then used to label cells, which subsequently can be imaged by ¹⁹F MRI. Important advantages of PFCs for cell tracking in vivo include (i) the absence of carbon-bound ¹⁹F in vivo, which then yields background-free images and complete cell selectivityand(ii) the possibility to quantify the cell signal by ¹⁹F MR spectroscopy.

Introduction

The tracking of cells in vivo is a crucial aspect in several fields of biomedicine. For this, noninvasive imaging techniques that can selectively localize cells over a period of time are extremely valuable. Prior to the development of three-dimensional magnetic resonance imaging (MRI), the tracking of immune cell migration was limited to microscopic analyses or tissue biopsies. Cell tracking with the help of MRI has gained immense attention in the past few years, not only for immunologists studying immune cell behavior in vivo, but also for clinical and stem cell researchers. During the mid-90s, the first studies on iron oxide nanoparticles ¹ initiated a cascade of developments for tracking cells with MRI. Iron oxide particles shorten the MR relaxation time (T2*) of the labeled cells and thus cause signal depletion in MR images. Iron oxide particles have been employed to label macrophages ², oligodendrocyte progenitors ³ and many other cell types. Some of these particles have also been clinically approved by the FDA for labeling cellular vaccines in melanoma patients ⁴. Since in vivo or ex vivo labeling of cells with iron oxide particles relies on a shortening of the T2* signal and the latter could be also brought about by in vivo susceptibility-related T2* effects such as micro bleeds, iron deposits or air bubbles, it might be difficult to identify labeled cells in vivo from other background T2* signal extinctions ⁵.

In this article, we describe a technique for tracking dendritic cells (DC) in vivo by employing ¹⁹F/¹H magnetic resonance imaging (MRI). This cell tracking technology was only introduced in 2005 ⁶, several years after the first recognized applications for ¹⁹F in MRI had been reported ⁷. One important advantage of ¹⁹F over iron oxide particle cell labeling is the low biological occurrence of ¹⁹F in tissue; this makes it possible to track cells very selectively with basically background-free images. Furthermore, it is possible to overlay the ¹⁹F MR signal from the transplanted labeled cells with anatomical images obtained from conventional 1H MRI. ¹⁹F/¹H MRI is therefore considerably relevant for studies investigating cell migration in vivo. Cells studied with this method are labeled with ¹⁹F-rich particles. Synthetically-derived perfluorocarbons (PFCs) consisting primarily of carbon and fluorine atoms are commonly used to prepare the particles. These compounds are insoluble in water and need to be emulsified prior to application in vitro or in vivo. The usual size of the PFC particles that have been employed by other groups for in vivo ¹⁹F-MRI tracking experiments ranges between 100 nm and 245 nm ^6,8-10. We have however shown that the efficiency in labeling dendritic cells with perfluoro-15-crown-5-ether (PFCE) particles increases with increasing particle size (>560 nm).¹¹

Protocol

All animal procedures must be approved by the local institutional animal welfare committee prior to execution. During the MR measurements an adequate level of anesthesia and physiological monitoring (body temperature, respiratory rate) are indispensable requirements.

1. Generation of Mouse Bone Marrow-derived Dendritic Cells

1. Extract bone marrow cells from C57BL/6 mice as previously described ¹². This protocol dates back to 1992 ¹³ and was originally described by the group of Ralph M. Steinman (1943-2011), discoverer of the dendritic cell ¹⁴.
2. Briefly, extract tibia and fibula soon after sacrificing mice. Working under the laminar flow hood (to prevent microbial contamination), flush the bone marrow into a small Petri dish with wash medium (5% FBS, 1% penicillin/streptomycin and 1% HEPES in RPMI) ¹².
3. Transfer the cell suspension through a cell strainer (100 μm mesh size, nylon) into a sterile 50-ml centrifuge tube in order to remove bone and leftover tissue. After several wash steps and a lysis step ¹², count the viable cells using trypan blue and a hemocytometer.
4. Resuspend bone marrow cells at a concentration of 400,000 cells ml^-1 in culture medium (RPMI supplemented with 10% FBS, 1% penicillin/streptomycin, 1% HEPES and 1% L-glutamine) containing 75 ng ml^-1 of the mouse granulocyte macrophage-colony stimulating factor (mGM-CSF) obtained from supernatants of 293FT HEK cells transfected with a mouse GMCSF plasmid. Transfer 10 ml of cell suspension (4 x 10⁶ cells) in Petri dishes (100 mm x 20 mm) and incubate in a CO₂ incubator (37 °C, 5% CO₂). On day 3, replenish old medium with 10 ml fresh medium (keeping end concentration of mGM-CSF at 75 ng ml^-1). On day 6, remove 10 ml of old medium and replenish with 10 ml fresh medium (mGM-CSF end concentration = 75 ng ml^-1). On day 8, remove 10 ml of old medium and replenish with 10 ml fresh medium (mGM-CSF end concentration = 150 ng ml^-1).
5. On day 10, mature the differentiated DC with 1 μg ml^-1 lipopolysaccharide (LPS) for 24 hr.

2. Labeling of Dendritic Cells with ¹⁹F-rich Particles

1. During the 24 hr maturation step (1.5), label the DC with 1 mM fluorine-rich perfluoro-15-crown-5-ether (PFCE) particles (500-560 nm) also for 24 hr. Prepare large fluorine-labeled particles by emulsifying PFCE in Pluronic F-68 (total volume 2.5 ml) using a cell-disrupting titanium sonotrode and employing a continuous pulse program (amplitude 100%, power = 250 W) for 60 sec.
2. Following the above incubation steps, harvest the mature and fluorine-labeled DC using a tissue culture cell scraper, transfer into a 50-ml centrifuge tube and centrifuge at 500 × g for 5 min at room temperature.
3. To wash off any leftover particles and dead cells, leave harvested cells to adhere on poly-L-lysine dishes. To prepare for this step, coat Petri dishes with poly-L-lysine (1 mg ml^-1) at 37 °C for 1 hr, thereafter wash off the unbound poly-L-lysine with PBS.
4. Incubate above harvested cells on the poly-L-lysine-coated plates in serum-free medium and allow cells to adhere (1 hr at 37 °C, 5% CO₂).
5. Wash off the medium to discard non-adherent cells, remaining particles and dead cells and wash three times with pre-warmed (37 °C) PBS.
6. Harvest adherent cells by trypsin-EDTA treatment (0.25% trypsin-EDTA, 3 min incubation at 37 °C).
7. Following another wash centrifugation step, resuspend the required amount of mature DC in serum-free sterile PBS.
8. To determine whether the cells have been successfully labeled, it is recommended to check the physicochemical characteristics of the cells using flow cytometry (FACS). The sideward scatter (SSC) should reveal an increased granularity in these cells, as previously described ¹¹. It should be also kept in mind that fibroblasts might contaminate the DC cultures and could also take up the ¹⁹F particles in parallel. Therefore the DC purity should be assessed (by measuring the percentage of CD11c⁺CD11b⁺ cells using FACS) prior to in vivo application.

3. In Vivo Application of ¹⁹F-labeled Dendritic Cells

1. Administer DC (10⁶ - 10⁷) in a volume of 50 - 100 μl intracutaneously into the hind limbs of a C57BL/6 mouse by positioning the mouse in a holder and carefully inserting a 26 ½ G needle into the upper layers of the skin prior to application (use 0.5 ml tuberculin syringes with permanently-attached needles to minimize dead volume). To ensure an appropriate intracutaneous application, it is important that the inserted part of the needle be partially visible beneath the skin. If the needle is not visible any more, the needle has been inserted too deep into the skin layers. For the untrained individual, the intracutaneous application could be performed under anesthesia.
2. Image the limbs by in vivo fluorine (¹⁹F) / proton (¹H) MRI (see next section) 21 hr following injection.

4. In Vivo ¹⁹F/¹H Magnetic Resonance Imaging

The detailed instructions for setting up the MR scans refer to a Bruker MR scanner using the control software Paravison (version 5.1). Names referring to vendor-specific functions and items have been highlighted in italics. For other MR scanners these steps may have to be adjusted according to the manufacturers' guidelines.

1. Arrange access to a dedicated small animal MR scanner. For adequate image quality, a system with a magnetic field strength of 9.4 Tesla or more and tailored radio frequency coils (e.g. ¹H/¹⁹F dual-tunable volume RF coil, 35 mm inner diameter, 50 mm length; Rapid Biomed, Würzburg, Germany) are recommended.
2. Prior to MRI, prepare the setup of the MRI scanner for anesthesia and physiological monitoring. Ensure that the breathing mask of the animal bed/holder is connected to the isoflurane system and that the temperature probe and breathing pad are connected to a remote animal monitoring system (e.g. Model 1025, SA Instruments Inc., New York, USA). The latter serves to monitor the animal's vital parameters such as body temperature and respiration.
3. Anesthetize mice by inhalation narcosis using a mouse chamber connected to an isoflurane system. Adjust the flow rate for air and O₂ at 0.2 L min^-1 and 0.1 L min^-1, respectively and 3% isoflurane (adjusted from a vaporizer) for about 2 min until the required level of anesthesia is reached (no response following toe pinch). Optimal flow rates may vary depending on the setup being used. Be aware that high flow rates might lead to dehydration. Careful monitoring of the animals' conditions is required at all times as mentioned above.
4. Position the mouse prone on the mouse bed/holder of the small animal MR scanner (Figure 1).
5. While keeping the flow rate for air and O₂ constant, adjust the isoflurane vaporizer to 0.8 - 1.5% until an optimal breathing pattern is reached. A respiratory rate of 50 - 70 breaths per min is recommended.
6. Keep the body temperature at 36-37 °C during the experiment by employing a warm water (or alternatively warm air) circulation system.
7. Keep the eyes of the mouse moist with sterile eye lubricating ointment during the measurement.
8. After securing the animal on the imaging holder and the connections to the monitoring system, move the holder to the center of the ¹H/¹⁹F RF coil (that is positioned at the isocenter of the animal scanner magnet).
9. Set the position of the mouse such that the knee lies within the isocenter of the magnet, either with an automated method (e.g. using a positioning laser combined with a motor-driven animal holder) or by manually calculating the distance the holder needs to be moved into the magnet to reach the isocenter.
10. For confirming the correct positioning of the animal in the scanner and later planning of the image slice geometry (i.e. size, position and rotation in space), acquire scout images in three standard orientations (axial, coronal, sagittal) using fast and low resolution image acquisition methods (e.g. TriPilot protocol, which uses a FLASH pulse sequence).
11. Tune the RF coil to the ¹H resonance frequency (e.g. 400.1 MHz for 9.4 T) and to the ¹⁹F resonance frequency (e.g. 376.3 MHz for 9.4 T) and match the characteristic impedance of the coil to 50 Ohm using the tuning monitor of the animal MR scanner. Tuning and matching is necessary to achieve the optimum conditions for transmission and signal reception.
12. Perform the necessary automatic system settings including shimming to fine-tune the homogeneity of the magnetic field (e.g. ADJ_SHIM), system frequency adjustments to tune the RF to the Lamor frequency of the mouse (e.g. ADJ_SF) and reference gain to adjust the RF amplitude (e.g. ADJ_REFG). All these settings are important to make the static and variable magnetics fields (B₀, B₁) in the region of interest as homogeneous as possible.
13. Acquire a second set of scout images (as above) along the three orthogonal orientations, after having adjusted the field of view (FOV) such that both lower limbs are visible from pelvis to foot (LxWxH = 50x25x25 mm). These images serve to plan the orientations of the final 3D ¹H/¹⁹F scans.
14. Set up a TurboRARE 3D protocol for ¹H-scan: TR/TE = 1,500/53 msec, FOV = 50x25x25 mm, Matrix = 400x200x200, RARE Factor = 16 (scan time approx. 60 min). Adjust FOV with the help of the scout images, start the scan (traffic-light).
15. Load a single pulse FID-sequence with a TR of at least 1,000 msec. Open Edit Scan and set nucleus to ¹⁹F. Use manual gain settings by deselecting the automatic reference gain in the Edit Method of the Toolbox.
16. Start measurement without recording data to display an MR signal in real time by using GSP (go setup). If the ¹⁹F spectral signal within the acquisition window is too low, add more averages in Edit Method and/or modulate the pulse attenuator (TX0 or SP0 slider) until a signal is clearly visible. Adjust the basic frequency in order to center the ¹⁹F spectral peak at 0 Hz in the acquisition window. Apply basic frequency and press Stop. Note that the isoflurane used as anesthesia may generate a background signal. At 9.4 Tesla MRI the chemical shift between the ¹⁹F-containing particles and the isoflurane is about 1,800 Hz. Ensure that the excitation bandwidth is smaller than 3,600 Hz to avoid exciting the isoflurane together with the ¹⁹F-labeled cells or particles. The basic frequency should be set correctly on the signal generated by the ¹⁹F-labeled cells. Alternatively, another method of anesthesia could be used as the experimenter sees fit.
17. Clone (duplicate) the TurboRARE 3D protocol from the previous ¹H scan. Go to Edit Method and deselect the automatic reference gain (as above). Set ¹⁹F nucleus from Edit Scan. Change the matrix to 128x64x64 and the RARE-Factor to at least 40 (up to 64). For a TR/TE 1,000/6 msec and 64 averages, the scan time is approximately 70 min. Set the receiver gain to 101 (toolbox) and start the scan without any further adjustments using GOP (go pipeline).
18. When the scans are finished retract the mouse holder from the MR scanner. Disconnect the mouse carefully from the holder. If the mouse is not sacrificed for ex vivo analysis (e.g. histology or spectroscopy, see below) directly after the MR measurements, closely monitor until it has completely recovered from anesthesia. Body temperature regulation is affected by the anesthesia, so during the recovery process, put the mouse in a separate cage that is placed on a warm temperature-regulated pad. Once the mouse has completely recovered from anesthesia, return it to its holding cage and to the animal room.
19. To overlay the ¹⁹F images to ¹H images, use the View Menu option on the same software (Paravison). The Underlay function can be found under Correlate. Save the underlay, load the new image and highlight the ¹⁹F layer by defining a new color from the Look up table. To discriminate between the ¹⁹F and ¹H scans, alter the threshold for the chosen color (cut Look up table) such that the ¹⁹F signal will have the chosen color and the background ¹H image will remain in grayscale. Other post-processing programs (e.g. ImageJ) are available to create the ¹H/¹⁹F image overlays.
20. Should an in vivo quantification of the ¹⁹F signal be necessary, follow a simple quantitative protocol as previously described ^15,16. In brief, quantification can be done by placing a reference tube with a known concentration of PFCE-Emulsion within the FOV next to the mouse. After acquiring the MR images, quantify the intensities of the ¹⁹F signals using region-of-interest (ROI) analysis and compare to the in vitro calibration curve as shown in Figure 2.

5. Lymph Node Magnetic Resonance Spectroscopy (MRS)

1. After sacrificing the mouse, remove the popliteal lymph node and place in a small 5 mm NMR tube and add 100 μl of 2% PFA.
2. Install a ¹⁹F spectroscopy coil (e.g. a toroidal coil) which has an opening of at least 5 mm for holding the NMR tube containing the lymph node.
3. Place the NMR tube inside the coil and move the coil into the isocenter of the magnet.
4. Tune the RF coil to the ¹⁹F resonance frequency (e.g. 376.3 MHz for 9.4 T) and match the characteristic impedance of the coil to 50 Ohm using the tuning monitor of the animal MR scanner. Tuning and matching is necessary to achieve the optimum conditions for transmission and signal reception.
5. Set all shim parameters within the Shimming Toolbox to 0 and press apply. Usually it is not necessary to apply shimming methods since the volume of the sample is relatively small. If a sufficient ¹⁹F signal is available, automated methods may be applied for shimming, system frequency and receiver gain as described above.
6. Load a single pulse FID sequence with a TR of at least 1,000 msec. Open Edit Scan and set the nucleus to ¹⁹F. Use manual gain settings by deselecting the automatic reference gain in the Edit Method of the Toolbox. Set the number of averages to 1.
7. Start the sequence using GSP. If the ¹⁹F spectral signal within the acquisition window is too low, add more averages in Edit Method and/or modulate the pulse attenuator (TX0 or SP0 slider) until a signal is clearly visible. Adjust the basic frequency so that the ¹⁹F spectral peak is at the center of the acquisition window at 0 Hz and apply basic frequency. Adjust the pulse attenuator again to maximize the signal, to reach 90 ° excitation. When signal is at maximum press Stop.
8. Set the averages in Edit Method between 64 and 256 (or more if necessary, depending on the signal) and start the acquisition using GOP. The detected signal correlates linearly with the number of averages. During quantification keep in mind that the detected signal needs to be normalized to one average. Furthermore, a calibration standard of a known concentration or a calibration curve is necessary to calculate the number of cells from the normalized ¹⁹F signal (Figure 2).
9. Once the scan is finished, remove the sample, place the next sample and proceed with step 5.3.

Results

Eighteen to twenty-one hours following intracutaneous application, ¹⁹F-labeled dendritic cells (DC) migrate into the draining popliteal lymph node. The movement of DC via the lymphatic ducts into the draining politeal lymph node can be appreciated by overlaying the ¹H anatomical images with the ¹⁹F DC images (Figure 2A). We have previously reported on the migration of these cells in vivo, as well as the impact of ¹⁹F-particle size on DC immunobiology, ...

Discussion

This method of employing ¹⁹F/¹H MRI to follow the movement of DC into the lymph node gives the opportunity to study the migration patterns of immune cells in vivo. Dendritic cells are excellent examples of rapidly migrating immune cells that are able to maneuver through three-dimensional structures without tightly adhering to specific substrates ¹⁷. Although the low spatial resolution (μm range) of the described technique is not comparable with the high resolution (nm ra...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
REAGENTS
C57BL/6 mice	Charles River, Berlin
RPMI	Gibco	21875-091
FBS Superior	Biochrom AG	S 0615
HEPES	Gibco-Invitrogen	15630-056
Penicillin-Streptomycin	Gibco	15140-122
L-glutamine	Gibco	25030-024
Dulbecco's PBS	Sigma Aldrich	D8662
PFA	Santa Cruz	sc-281692
Perfluoro-15-crown-5-ether	ChemPur	391-1996
Pluronic F-68	Sigma Aldrich	P5556
Petri dishes (35 x 10 mm)	VWR, Germany	391-1996
27 ½ G syringes	VWR, Germany	612-0151
Nylon cell strainers (100 μm mesh)	VWR, Germany	734-0004
NMR tubes	VWR, Germany	634-0461
EQUIPMENT
Dissection tools	FST
CO2 incubator	Binder
Small animal MR system	Bruker Biospin		9.4T BioSpec 94/20 USR, ParaVision Acquisition and Processing Software
1H/19F dual-tunable volume RF coil	Rapid Biomed, Würzburg, Germany		35 mm inner diameter, 50 mm length
19F spectroscopy coil	in-house		tune/match loop coil, 4 turns, inner diameter 5 mm, 10 mm long, two capacitors for tuning and matching
Isoflurane inhalation system	Föhr Medical Instruments GmbH
Animal monitoring system Model 1025	SA Instruments Inc., New York, USA