Visualization and Quantification of Brown and Beige Adipose Tissues in Mice using [¹⁸F]FDG Micro-PET/MR Imaging

Summary

Functional imaging and quantitation of thermogenic adipose depots in mice using a micro-PET/MR imaging-based approach.

Abstract

Brown and beige adipocytes are now recognized as potential therapeutic targets for obesity and metabolic syndromes. Non-invasive molecular imaging methods are essential to provide critical insights into these thermogenic adipose depots. Here, the protocol presents a PET/MR imaging-based method to evaluate the activity of brown and beige adipocytes in mouse interscapular brown adipose tissue (iBAT) and inguinal subcutaneous white adipose tissue (iWAT). Visualization and quantification of the thermogenic adipose depots were achieved using [¹⁸F]FDG, the non-metabolizable glucose analog, as the radiotracer, when combined with the precise anatomical information provided by MR imaging. The PET/MR imaging was conducted 7 days after cold acclimation and quantitation of [¹⁸F]FDG signal in different adipose depots was conducted to assess the relative mobilization of thermogenic adipose tissues. Removal of iBAT substantially increased cold-evoked [¹⁸F]FDG uptake in iWAT of the mice.

Introduction

In response to changing nutritional needs, adipose tissue serves as an energy cache to adopt either lipid storage or mobilization mode to meet the needs of the body¹. Moreover, adipose tissue also plays a key function in thermoregulation, via a process called non-shivering thermogenesis, also called facultative thermogenesis. This is typically achieved by the brown adipose tissue (BAT), which expresses abundant level of mitochondria membrane protein uncoupling protein 1 (UCP1). As a proton carrier, UCP1 generates heat by uncoupling the proton transport and ATP production². Upon cold stimulation, thermogenesis in BAT is set in motion by activation of the sympathetic nervous system (SNS), followed by release of norepinephrine (NE). NE binds to the β3 adrenergic receptors and leads to elevation of intracellular cyclic AMP (cAMP). As a consequence, cAMP/PKA-dependent engagement of CREB (cAMP response element-binding protein) stimulates Ucp1 transcription via direct binding on CREB-response elements (CRE)². In addition to BAT, brown-like adipocytes are also found within white adipose tissue and are therefore named beige or brite (brown-in-white) cells^1,3. In response to specific stimuli (such as cold), these otherwise quiescent beige cells are remodeled to exhibit multiple brown-like features, including multilocular lipid droplets, densely-packed mitochondria, and augmented UCP1 expression^3,4,5.

Animal studies have demonstrated that brown and beige adipocytes possess multiple metabolic benefits beyond its fat-reducing effect, including insulin-sensitization, lipid-lowering, anti-inflammation, and anti-atherosclerosis^6,7. In humans, the amount of beige/brown fat is inversely correlated with age, insulin resistance index, and cardiometabolic disorders⁸. Moreover, activation of beige/brown adipocytes in humans by either cold acclimation or β3 adrenergic receptor agonist confers protection against a series of metabolic disorders^4,9,10. These pieces of evidence collectively indicate that induction of brown and beige adipose tissue is a potential therapeutic strategy for management of obesity and its related medical complications⁸.

Interestingly, although they share similar function, beige and classical brown adipocytes are derived from different precursors and activated by overlapping but distinct mechanisms¹. Therefore, in vivo imaging and quantification of brown and beige adipocytes are essential to achieve a better understanding of the molecular control of these adipose tissues. Currently ¹⁸F-fluorodeoxyglucose ([¹⁸F]FDG) positron emission tomography (PET) scan combined with computed tomography (CT) remains the gold standard for characterization of thermogenic brown and beige cells in clinical studies⁸. Magnetic resonance imaging (MRI) uses powerful magnetic fields and radio frequency pulses to produce detailed anatomical structures. Compared to CT scan, MRI generates images of organs and soft tissues with a higher resolution. Provided here is a protocol for visualization and quantification of functional brown and beige adiposes in mouse models after acclimation to cold exposure, a common and most reliable way to induce adipose browning. This method can be applied to characterize the thermogenic adipose depots in small animal models with high precision.

Protocol

The protocol described below follows the animal care guidelines of The University of Hong Kong. The animals used in the study were 8-week-old C57BL/6J mice.

1. Animal surgical procedures and cold challenge

1. Perform interscapular BAT (iBAT) dissection.
   1. Anesthetize the mice by intraperitoneal injection of ketamine/xylazine (100 mg/kg bodyweight ketamine and 10 mg/kg bodyweight xylazine). After anesthesia, shave the hair of the mouse from the neck to just below the scapulae.
   2. Place the mice on the heating pad after disinfection and make a 2 cm incision along the dorsal midline of the mice.
   3. Remove the iBAT pads (bilateral). In the sham-operated group, make the same incision but leave the iBAT pads intact.
   4. Close the incision using 7 mm stainless steel wound clips after the bleeding stops.
   5. After surgery, give meloxicam (5 mg/kg in drinking water) to the mice for 6 days and house them in an intensive care unit (ICU) for 14 days. Remove the clips as soon as the wound is healed (7-10 days).
2. Cold challenge of the mice: House the mice at thermoneutrality (30 °C) for 14 days. On day 13, pre-chill the animal cages in cold (6 °C) overnight. On day 14, put the mice at 6 °C in the environmental chamber for 7 days. Place two mice in each cage.

2. Micro-PET/MR calibrations and workflow setup

NOTE: Micro-PET/MR imaging is performed using a sequential PET/MR system (see Table of Materials). Each mouse is placed on the imaging bed; first scan with the MR for an anatomical reference (scout view) before advancing to the center of the PET field-of-view (FOV) for a static [¹⁸F]FDG PET acquisition, followed by MR imaging for anatomic reference. An imaging workflow is created in the scanner-operating software (see Table of Materials) to enable automated, sequential PET/MR scans prior to the imaging session.

1. Create an imaging workflow in the operating software that includes static PET acquisition, MRI acquisitions for attenuation correction, and anatomical reference using T1-weigthed 3D imaging and T2-weighted 2D imaging, respectively.
2. To acquire PET, set 400-600 keV level discrimination, F-18 study isotope, 1-5 coincidence mode and 20 min scans.
3. To acquire T1-weighted MR (for attenuation correction), set Gradient Echo-3D (TE = 4.3 ms, TR = 16 ms, FOV = 90 x 60 mm, number of excitations (NEX) = 3, 28 slices with 0.9 mm thickness, voxel size = 0.375 x 0.375 x 0.9 mm).
4. To acquire T2-weighted MR (anatomical reference), set Fast-spin Echo 2D (TE = 71.8 ms, TR = 3000 ms, FOV = 90 x 60 mm, NEX = 5, 32 slices with 0.9 mm thickness, voxel size = 0.265 x 0.268 x 0.9 mm³).
5. To reconstruct PET, use Tera-Tomo 3D (TT3D) algorithm (8 iterations, 6 subsets) with 1-3 coincidence mode, and with decay, dead-time, random, attenuation, and scatter corrections to create images with an overall of 0.3 mm³ voxel size.
6. Perform a PET Activity Test of the micro-PET/MR scanner one day before the start of imaging study to check the accuracy of PET quantitation.
   1. Prepare a 5 mL syringe filled with [¹⁸F]FDG as recommended by the manufacturer guidelines (140-220 μCi/5-8 MBq in water or saline).
   2. Record the activity of the syringe using a dose calibrator (see Table of Materials) and note the time of measurement.
   3. Select Interpolated Ellipse ROI to draw a volume-of-interest (VOI) on the reconstructed image to compare the recovered activity to the value measured as described above. The recovered activity for a well-calibrated scanner is accurate within ±5%.

3. Injection of [¹⁸F]FDG

1. Order a clinical dose of [¹⁸F]FDG (10 mCi/370 MBq) from the supplier for its arrival to the imaging lab approximately 30 min before the first scheduled injection. Make sure to wear appropriate personal protective equipment (PPE), such as a lab coat, gloves, personal radiation dosimeter e.g. Fingers, whole body when receiving the package containing radioactive materials. Change gloves regularly to prevent cross contamination of the radioactivity and increase distance from the radioactive source as much as possible.
2. Use the forceps to carefully transfer the [¹⁸F]FDG stock vial behind an L-block table top shield.
3. Dispense an aliquot of [¹⁸F]FDG and dilute with sterilized saline to give a total activity concentration at 200-250 μCi/7-9 MBq) in 100-150 μL.
4. Draw the [¹⁸F]FDG solution into a 1 mL syringe with needle (see Table of Materials), measure the radioactivity using a dose calibrator set to F-18, and record the time of measurement.
5. Record the weight of the mouse prior to injection. Inject the prepared [¹⁸F]FDG solution via tail vein. Take note of the injection time and residue of the radioactivity of the syringe to enable decay correction.
6. Put the mouse back in the cage and allow [¹⁸F]FDG uptake for 60 min before PET scans.
7. Calculate the injected [¹⁸F]FDG activity using the following formula¹¹:
   Injected activity (μCi/MBq)
   = Activity in the syringe before injection
   - activity in the syringe after injection

4. Micro-PET/MR acquisition

1. Turn on the air heater to the mouse bed to allow warm air to pass through it.
2. Anesthetize the mouse using 5% isoflurane (1 L/min medical O₂). Once induced, transfer the mouse to the warm mouse bed and maintain anesthesia at 2%-3 % isoflurane via a nose mask cone. Position the mouse head-prone onto the bite bar and make sure the mouse does not protrude outside of the diameter of the bed. Apply eye lubricant to avoid drying and formation of corneal ulcers.
3. Monitor the body temperature and the respiratory rate by a thermal probe and a respiratory pad, respectively. Maintain the body temperature at 36-37 °C, and the respiratory rate at 70-80 breaths per minute (bpm) by adjusting the isoflurane level.
4. Perform a scout view to determine the mouse position. Adjust the mouse bed position to include the whole mouse body, and to ensure the center FOV of MR is in the center of mouse body.
5. Under the PET Acquisition in the study list window, select Scan Range on Previous Acquisition to use the scout view position as described above. Click on Prepare to move the animal bed from MR to PET. Select OK to initiate the PET scan. Record the injection dose and time measured before and after [¹⁸F]FDG administration in the Radiopharmaceutical Editor. Enter the weight of the mouse under the Subject Information menu.
6. Once the PET scan is completed, select Prepare to move to MR and complete all the MR acquisitions in the study list window. Select OK to start the MR scans.
7. After the whole workflow is completed, briefly evaluate the quality of the acquired MR images using the post-processing software (see Table of Materials). Click on the Home button to move the mouse bed from MR to the original position.
8. Carefully remove the mouse from the scanner and return it to a clean housing cage with a heated pad underneath to allow recovery in warm environment. Supply the mouse with food and water. The system is now ready for the next mouse in queue.
9. To reconstruct data, select PET Acquisition under the Raw Scan menu to load the completed PET scan. Select T1-weighted MR Acquisition for material map creation. Reconstruct the data as described above (see step 2.5).
10. Follow the local and institute regulations regarding the care and handling of post-PET imaging mouse. Consider all used syringes/needles, gloves, bedding, and fecal matter as radioactive waste that require special handling/disposal in accordance with the local regulations.

5. Post-imaging analysis

1. Open the Image Analysis software (see Table of Materials) and click on Load DICOM Data to retrieve the corresponding MR and PET images.
2. Perform co-registration of MR and PET image by dragging these images to the display window. Click on the Automatic Registration function.
   1. Select Rigid transformation under the Registration Setup drop-down menu. Check Shift and Rotation under the Rigid/Affine menu.
   2. Select T1-weighted MR acquisition as the Reference and PET acquisition as the Reslice under the Global Role Selection menu.
   3. Inspect the registration in all three dimensions to make sure a perfect alignment between MR and PET images. To adjust it manually, click on Manual Registration.
3. Use Interpolated Ellipse ROI to draw VOI on the tissue of interest, i.e., iBAT and inguinal white adipose tissue (iWAT) using MR image for reference. Use the Brush Tool and Eraser Tool to define the VOI border; hence, the anatomy of tissues. Make sure there is no overlap uptake by using PET image to avoid spillover from the neighboring organs. Repeat the process slide-by-slide until the whole VOI is delineated. If necessary, edit the VOIs to maintain consistent VOI volumes between each mouse.
4. Use Ellipsoid VOI to draw a 3 mm³ VOI on the lung as a reference organ. Avoid any spillover from the neighboring heart and muscle.
5. Upon completion, click on Show ROI Table to rename each VOI. Record the mean radioactivity with the VOI and tissue volume into a spreadsheet. Archive the VOI drawings and the imaging data to a data storage device.
6. Calculate the standardized uptake value (SUV) for all VOIs using the following equation¹¹:
   SUV_mean = VOI radioactivity in kBq / (Decay - corrected injected dose in kBq / mouse body weight in kg), assuming a tissue density of 1 g/mL. 

Representative Results

Three groups of mice (n = 3 per group) underwent micro-PET/MR imaging in this study, where they were housed at either thermoneutrality (30 °C) or cold (6 °C) for 7 days. One group of mice (n = 3) had their iBAT removed (iBATx) prior to cold treatment (Figure 1A). This method led to an alteration to the white adipose tissue activity in all three mice. In particular, a remarkable increase in [¹⁸F]FDG uptake was observed in iWAT using micro-PET/MR imaging (Figure 1B-C). This co-registered imaging data is demonstrated as maximal intensity projection (MIP), where iWAT was clearly delineated to allow quantification of the [¹⁸F]FDG uptake. Consistently, the multilocular adipocytes, which are characteristic morphology for beige adipocytes, were more pronounced in iWAT from iBATx mice, compared to the sham operated group (Figure 1D).

To verify whether changes on iBAT and iWAT activities upon this prolonged cold induction can be monitored by micro-PET/MR imaging, imaging studies were performed on the mice exposed to 30 °C and 6 °C and the results between groups were compared. PET/MR imaging also demonstrated that mice subjected to 6 °C have markedly elevated [¹⁸F]FDG uptake on iBAT in sham operated mice (Figure 2A), which is consistent with the previous reported literature¹¹. Mice with their iBAT removed (iBATx) prior to cold treatment showed the highest [¹⁸F]FDG uptake in the iWAT among the 30 °C and 6 °C group (Figure 2B). PET images were further quantified using an SUV-based approach. In iBAT, cold exposure caused a 7-fold increase in [¹⁸F]FDG uptake when compared to the 30 °C group. In iWAT, [¹⁸F]FDG uptake was higher in cold-acclimated iBATx mice than the remaining groups (Figure 2C). Removal of iBAT in the cold-induced mice resulted in an 8-fold increase in the uptake of iWAT compared to the thermoneutrality mice, whereas only a modest increase (2-fold) was observed when iBAT was present in mice.

Figure 1: Micro-PET/MR Imaging of inguinal white adipose tissue (iWAT) in mice. Interscapular brown adipose tissue was surgically removed (iBATx). After recovery, the mice were housed at 6 °C for 7 days before analysis. (A) Flow chart for the surgical and the subsequent procedures. (B) Illustration of the position of the mouse and the PET/MR scanner. (C) Maximal intensity projection (MIP) of co-registered PET/MR images. White arrows: Location of iWAT. A: Anterior L: Left. (D) Hematoxylin and Eosin (HE) staining of iWAT in sham and iBATx mice after cold exposure. Scale bar = 100 µm. Please click here to view a larger version of this figure.

Figure 2: Representative in vivo [¹⁸F]FDG uptake in brown adipose tissue at the interscapular region (iBAT) and inguinal subcutaneous white adipose tissue (iWAT). Mice housed at thermoneutrality (30 °C), cold-acclimated (6 °C) and cold-acclimated + iBATx were subjected to [¹⁸F]FDG PET/MR imaging. (A) Sagittal section of PET/MR images showing iBAT in mice. (B) Axial section of PET/MR images showing bilateral iWAT. (C) Quantitative analysis of [¹⁸F]FDG uptake in iBAT (left) and iWAT (right). Yellow arrows: Location of iBAT. White arrows: Location of iWAT. n = 3 for each group. Values of SUV_ratio are presented as mean ±SD. Please click here to view a larger version of this figure.

Discussion

In this study, a PET/MR -based imaging and quantification of functional brown and beige adipose tissue in small animal was described. This method uses the non-metabolizable glucose analog [¹⁸F]FDG as an imaging biomarker so as to identify the adipose tissues with high glucose-demand in a non-invasive manner. MR offers good soft tissue contrast and can better differentiate adipose fat tissues from the neighboring soft tissues and muscle. When combined with PET, this enables imaging and quantifying of the activated adipocytes as a result of high glucose utilization in an accurate manner. The experimental conditions outlined here highlights the feasibility of using [¹⁸F]FDG PET to study up-regulation of iBAT and iWAT in vivo, and is potentially useful for evaluation of the thermogenic impact of new drug candidates. In addition, this protocol can be easily modified into high throughput format by simultaneously imaging multiple mice using a specially designed animal bed, thereby increasing the statistical power and confidence in the imaging data at a reduced cost and time^12,13.

Currently, [¹⁸F]FDG PET/CT remains the most common approach to visualize BAT in humans and rodents and standard protocols have been well established^8,11. In recent years, there are also several studies using [¹⁸F]FDG PET/MR imaging to assess BAT in humans^14,15,16. In contrast, no detailed description on [¹⁸F]FDG PET/MRI for small animals is available. Described here is a detailed protocol that relies on the use of a combined PET and MR imaging system in mice. This method takes advantage of the higher resolution of MRI, especially on the detection of fat tissues, making them easy to identify and segment compared to the commonly used CT method. Therefore, the current approach enables an improved accuracy of PET quantification compared to the PET/CT method, which is of great value for studies in small animals with more delicate adipose depots. When analyzing the results of tissues of interest at their baseline uptake, MRI becomes an essential tool to accurately draw the VOIs to ensure consistency of their volumes between mice and avoid inclusion of neighboring organs. In addition, accurate image processing such as image registration and VOI delineation are important to allow reliable quantification. The anatomical location of the glucose-responsive BAT is distinct between humans and mouse. While the functional BAT locates at the interscapular region, [¹⁸F]FDG PET/MR imaging-based analysis mainly identify functional BAT in the supraclavicular region in humans^14,15,16.

The fasting or fed status of the mice should also be taken into consideration when performing the [¹⁸F]FDG uptake experiment. In some studies, the mice are fasted for several hours or even overnight before the uptake experiment since it is supposed that endogenous glucose will compete with [¹⁸F]FDG. In the protocol, the [¹⁸F]FDG at fed status was measured and strong uptake signal was still observed in both iBAT and iWAT. This, thus, demonstrates that it is not necessary to put the mice at a fasted status for robust uptake signals, which is less physiologically relevant. Actually, caution should be taken when examining BAT and beige adipocytes in fasted animals since a previous finding has reported that the hypothalamic neuropeptide Y (NPY)-mediated hunger signal acts on the medullary motor systems to inhibit BAT thermogenesis by reducing the sympathetic innervation¹⁷. Consistently, in humans, it is suggested that upon high calorigenic diets, the thermogenic adipocytes burn out extra calories so as to maintain energy balance. In contrast, upon nutrient deprivation, counter-regulatory mechanisms are activated to suppress energy waste.

Another consideration for [¹⁸F]FDG PET imaging involves the routes for radiotracer administration into mice. Intraperitoneal and intravenous techniques are two common ways to inject [¹⁸F]FDG into mice, and both methods result in a relatively similar biodistribution of [¹⁸F]FDG in mice 60 min post injection¹⁸. While the intraperitoneal method is relatively easy to perform and the injection can be done quickly to avoid unwanted stress imposed onto the mice, direct injection accidentally into the bowel is common and is not immediately identified, leading to unreliable PET results¹⁹. Intravenous method is the preferred method and employed in this study. Successful tail vein injection can be determined when a visible blood flashback is observed prior to infusion, indicating that the needle is properly positioned inside the vein for infusion. One limitation to this technique is the difficulty to notice a visible blood flashback, potentially due to low blood pressure and the presence of dark hair on the tails. This can be overcome by warming the tail with a warm washcloth to increase the blood flow, hence improving the visibility of the vein for needle insertion.

An accurate scanner and a relevant equipment are other important factors for reliable PET image quantification. It is imperative to perform routine quality control examinations on PET and MR components of the scanner. MR quality control involves the signal-to-noise ratio assessment on different T1- and T2-weighted sequences, which should be performed on a weekly basis as recommended by the scanner manufacturer. For PET, accuracy of activity must be determined using a syringe containing a known concentration of radioactivity on a weekly basis or before the start of an important study. This quality control test also allows determination of co-registration of PET and MR images. Calibration must be performed if the recovered activity falls outside the recommended range or mis-registration between PET and MR images is found. In addition, the dose calibrator should be regularly calibrated according to the manufacturer guidelines since this is an important tool for quality control of the scanner as well as radioactivity measurement for PET imaging.

This study shows that the activation of adipose depots in both iBAT and iWAT in mice can be visualized and quantified using [¹⁸F]FDG PET/MR imaging upon exposure to cold temperature. However, the current study is limited by the fact that [¹⁸F]FDG uptake in iWAT was relatively low unless in the absence of iBAT. This indicates that compared to the iBAT that is readily activated by cold stimulus, beige adipocytes are relatively reluctant to be mobilized and act more like a backup thermogenic depot of iBAT in mice. More efficient methods to induce the [¹⁸F]FDG signal in the iWAT and/or other adipose depots in normal mice, such as the use of beige-specific activators or stronger cold challenge condition, are to be identified, which is beyond the scope of the current study.

Materials

Name	Company	Catalog Number	Comments
0.9% sterile saline	BBraun		0.9% sodium chloride intravenous infusion, 500 mL
5 mL syringe	Terumo	SS05L	5 mL syringe Luer Lock
Dose Calibrator	Biodex		Atomlab 500
Eye lubricant	Alcon Duratears		Sterile ocular lubricant ointment, 3.5 g
Insulin syringe	Terumo	10ME2913	1 mL insulin syringe with needle
InterView Fusion software	Mediso	Version 3.03	Post-processing and image analysis software
Isoflurane	Chanelle Pharma		Iso-Vet, inhalation anesthetic, 250 mL
Ketamine	Alfasan International B.V.	HK-37715	Ketamine 10% injection solution, 10 mL
Medical oxygen	Linde HKO	101-HR	compressed gas, 99.5% purity
Metacam	Boehringer Ingelheim		5 mg/mL Meloxicam solution for injection for dogs and cats, 10 mL
nanoScan PET/MR Scanner	Mediso		3 Tesla MR
Nucline nanoScan software	Mediso	Version 3.0	Scanner operating software
Wound clips	Reflex 7	203-100	7mm Stainless steel wound clips, 20 clips
Xylazine	Alfasan International B.V.	HK-56179	Xylazine 2% injection solution, 30 mL