Isolating Brown Adipocytes from Murine Interscapular Brown Adipose Tissue for Gene and Protein Expression Analysis

Both mice and humans have two types of adipose tissues: white adipose tissue (WAT) and brown adipose tissue (BAT)¹. WAT stores energy in the form of triglycerides in white adipocytes, and the brown adipocytes (BAs) of BAT dissipate chemical energy as heat². Based on their developmental origin, BAs are further categorized into classical BAs (cBAs) that formed during embryo development and white adipocytes derived BAs (beige/brite cells, converted from white adipocytes under stress conditions)³. BAs are multilocular and express the thermogenic protein uncoupling protein 1 (UCP1)⁴. Interscapular BAT (iBAT) depot is one of the primary cBAs depots in small mammals⁵, whereas beige cells are dispersed within WAT⁶.

Due to their nature of dissipating energy, BAs have received much attention as a therapeutic target for reducing obesity⁷. To exploit BAs for the purpose of treating obesity, it is essential to understand the molecular mechanisms that control BAs function, survival, and recruitment. Adipose tissues including BAT and WAT are heterogeneous. Except for adipocytes, adipose tissues contain many other cell types, such as endothelial cells, mesenchymal stem cells and macrophages⁸. Although genetic tools to specifically deplete candidate genes in mice BAs are available, such as UCP1::Cre line⁹, techniques for purifying BAs from BAT or WAT are limited, making it hard to study BAs at a single-cell type level. Additionally, without obtaining pure BAs, the relationship between BAs and non-BAs will not be clearly delineated. For instance, platelet-derived growth factor receptor alpha (PDGFRα) has been used as a marker for undifferentiated mesenchymal cells, and it is expressed in the endothelial and interstitial cells of BAT. In cold stressed BAT, PDGFRα positive progenitor cells give rise to new BAs¹⁰. PDGFRα is activated by its ligand PDGF, a growth factor that regulates the growth and proliferation of mesenchymal cells¹¹; however, it is unclear whether BAs influence the behavior of PDGFRα positive progenitor cells by secreting PDGF.

Recently, a BAs isolation protocol has been published, which is based on fluorescence-activated cell sorting (FACS)¹². In this protocol, 3% bovine serum albumin (BSA) solution was used to separate BAs from non-BAs, and the enriched BAs were further purified by FACS. The application of this protocol is limited by the requirement of FACS process, which relies on both equipment and FACS operation experiences. In this study, a new protocol for isolating BAs from BAT was developed. The BAs isolated by this protocol can be directly used for gene and protein expression studies. Furthermore, data from this study suggest that BAs are a major PDGF resource.

All mice were maintained in pathogen-free conditions, and all procedures were approved by Masonic Medical research Institutional Animal Care and Use Committee (IACUC). UCP1::Cre⁹ and Rosa 26^tdTomato mice lines¹³ were reported previously. All mice were kept at room temperature with a 12 h light/dark cycle.

1. Preparing the Solutions and brown adipose tissue (BAT)

1. Prepare digestion solution and separation solution in 15 mL centrifuge tubes.
2. 10 mL of BAT Digestion solution: To 10 mL of sterile phosphate-buffered saline (PBS), add 3.5 mg/mL Dispase II, 1 mg/mL Collagenase II and 10 mM CaCl₂ to make BAT digestion solution.
3. 10 mL of 12% iodixanol solution (separation solution): Mix 1 mL of 10x PBS, 2 mL of 60% iodixanol, 0.01 mL of 1 M MgCl₂, 0.025 mL of 1 M KCl, 0.1 mL of 0.2 M ethylenediaminetetraacetic acid (EDTA) and 6.865 mL of ddH₂O to obtain 12% iodixanol.
4. Euthanize one adult mouse with CO₂ overdose. Briefly, fill the mouse cage with 100% CO₂ at a displacement rate of 10-30% cage volume per min. 5 min later, confirm death by checking for the absence of visibly breathing.
5. Dissect the animal and collect interscapular BAT. Remove WAT and muscle layers under a stereo microscope.
6. To have sufficient digestion, cut each BAT lobe into ~ 3 mm³ parts and place them into a clean 50 mL flask with a metal stir bar and 5 mL digestion solution. Before starting digestion, let the flask containing BAT and digestion buffer sit on ice for 1 h.
   ​NOTE: In the following steps, around 80 mg BAT was used for brown adipocytes isolation.

2. BAs isolation procedure

1. Place the flask on a magnetic stirrer that is enclosed in an incubator. Set the stirring speed at 60 rpm and the temperature of the incubator at 35 °C, respectively. The digestion will last for around 30 min. If the BAT slices form clumps around the stirrer bar during digestion, use a 1 mL pipette tip to disrupt the aggregated tissues.
2. Place a 70 µm strainer filter on top of a clean 50 mL centrifuge tube. Pipette around 4 mL cell suspension through the strainer. Wash the strainer with 4 mL of 12% iodixanol solution. Pipette up and down to mix the cells and the iodixanol solution. Transfer the cell mixture into two clear 5 mL polystyrene test tubes.
   NOTE: At the end of digestion, the digestion solution should be cloudy, indicating sufficient digestion. Use fresh digestion solution every time. Once prepared, the digestion solution should be used within 2 h.
3. Repeat step 2.2 and 2.3 once more if the digestion is not sufficient.
4. Leave the clear polystyrene tubes containing the separation solution and BAs on ice for 1 h. The BAs will form a layer on the top.
5. Take out 20 µL of the isolated BAs for microscope examination.

3. RNA and protein isolation from BAs

1. Pipette the BAs layer into two 1.7 mL microcentrifuge tubes for RNA and protein isolation. Carefully remove the excessive iodixanol solution without disrupting the BAs layer.
2. Add 1 mL Trizol to simultaneously isolate RNA, DNA, and protein into the cell solution. Mix sufficiently to lyse the cells.
3. Add 200 µL of Chloroform to separate the phases.
4. Centrifuge the tube for 9,981 x g for 10 min at 4 °C.
5. After centrifugation, use the aqueous phase for RNA isolation. In this study, total RNA was used for reverse transcription, and quantitative RT-PCR was carried out with Real-Time PCR. Primer sequences are listed in Table of Materials.
6. Transfer 300 µL of the organic phase into a 2 mL microcentrifuge tube.
7. To the organic phase, add 2.5 volume 100% ethanol and vortex for 10 s.
8. Add 200 µL of 1-Bromo-3-chloropropane and vortex for 10 s.
9. Add 600 µL of double distilled water and vortex for 10 s.
10. Let the mixed solution stand for 10 min at room temperature.
11. Centrifuge at 9,981 x g for 10 min at 4 °C. At this step, the phases will be separated. The protein phase is localized in the middle layer.
12. Remove the top aqueous solution. Add 1 mL 100% ethanol into the remaining solution.
13. Centrifuge for 10 min, 9,981 x g, 4 °C. After centrifugation, the protein pellet will form. Discard the supernatant.
14. Wash the pellet with 1 mL 100% ethanol.
15. Centrifuge for 10 min, 9981 x g, 4 °C. Save the pellet and discard the supernatant.
16. Air-dry the pellet for 10 min at room temperature.
17. Measure the weight of the wet pellet. Add in 1% SDS solution at a ratio of 20 µL/mg pellet.
18. Dissolve the pellet by putting the tube in a heated shaker. Set the temperature at 55 °C, and the speed at 11 x g. It usually takes 5-10 min to completely dissolve the protein pellet.
    NOTE: The concentration of the dissolved protein can be measured by BCA assay¹⁴.

Preparation of interacapular BAT for brown adipocytes isolation
The brown adipocytes (BAs) isolation process is depicted in Figure 1A. The whole process, from preparing BAT and digestion/separation solutions to obtaining isolated BAs will take around 4 h.

In adult mice, abundant BAT exists in the interscapular region. This interscapular BAT (iBAT) is covered by muscle layers and WAT (Figure 1B). Before starting the digestion procedure, the muscle layers and WAT need to be removed to yield out clean iBAT (Figure 1C). In a published BAs isolation protocol, minced BAT was used for BAs isolation¹². In this study, digestion of 3 mm³ size BAT (Figure 1D) yielded out more BAs than minced BAT.

Separation of BAs from non-BAs with 3% BSA solution
After iBAT digestion, dissociated BAs were mixed with non-BAs in the digestion product. Because BAs contain lipid droplets, their density is lower than non-BAs; however, the BAs density is not low enough to let them efficiently float to the top of a regular PBS solution. PBS containing 3% bovine serum albumin (BSA) has been used to separate BAs from non-BAs¹², which was successfully repeated in this study (Figure 2A).

Rosa 26^tdTomato is a reporter mouse line, which expresses strong tdTomato (tdTom) fluorescence protein following Cre-mediated recombination¹³. Ucp1::Cre transgenic mice express Cre recombinase in the BAs⁹. Ucp1::Cre mouse line was crossed with Rosa 26^tdTomato mice to genetically label BAs with tdTom (Figure 2B). For validating the BAs isolation procedure, iBAT from Ucp1::Cre;tdTom/+ mice was dissociated. Most of the cells enriched in the top layer of BSA solution were raspberry shape and contained multilocular lipid droplets. Furthermore, most of these raspberry shape cells were tdTom positive (Figure 2C), confirming that they were brown adipocytes.

Separation of BAs from non-BAs with 6% iodixanol solution
3% BSA separation solution enriched BAs. It was unclear whether these isolated BAs could be used for gene and protein expression analysis. RNA and protein were then extracted from the enriched BAs. RNA extraction was successfully performed according to the standard RNA isolation procedure. However, the standard Trizol protein isolation protocol, also known as Guanidinium thiocyanate-phenol-chloroform method (GTPC method), did not work well, which was tedious and had very low protein yield. Therefore, an improved protein isolation method was adopted to extract protein from Trizol-lysed BAs.

In this improved GTPC protocol, ethanol, bromo-chloropropane, and water was used for extracting protein from the organic phase¹⁵. After adding ethanol, bromo-chloropropane and water into the organic phase, and after centrifugation, protein pellet formed between the aqueous phase and the organic phase (Figure 3A). Protein pellet was then washed with 100% ethanol and dissolved in 1% SDS. This improved GTPC method was used to extract protein from iBAT and BSA solution-enriched BAs. Although the BAT protein pellet was easily dissolved in 1% SDS, major part of the isolated-BAs protein pellet was not soluble. Then the dissolved protein was examined with SDS-PAGE gel. As shown in a Coomassie blue stained SDS-PAGE gel (Figure 3B), a massive protein band around 60 kDa was present in the isolated BAs but not in the BAT samples. Because the molecular weight of BSA is 66 kDa, and abundant BSA exists in the BAs separation solution, this dominant protein band should be BSA. These data suggest that the BSA from the BAs separation solution interferes with protein extraction.

Iodixanol is a nonionic and iso-osmotic gradient medium¹⁶ that has been widely used for cell¹⁷ and adeno-associated virus (AAV) purification¹⁸. To avoid BSA interference of protein expression studies, iodixanol was used to replace BSA in a new BAs separation solution. 3% BSA solution has a density of 1.03, which is similar with 6% iodixanol. In 6% iodixanol solution, BAs floated to the top in 30-60 min (Figure 3C). The BAs isolated with this solution showed out typical raspberry shape and contained multilocular lipid droplets (Figure 3D). Proteins extracted from these isolated BAs were nicely separated in SDS-PAGE gel (Figure 3E).

To verify whether the 6% iodixanol solution efficiently separated BAs from non-BAs, we genetically labeled the BAs with tdTom and examined the cells residing in the clear 6% iodixanol solution. After separating the BAs from non-BAs (step 2.4), the 6% iodixanol solution below the BAs layer was diluted 6 times with PBS and was then centrifuged at 600 x g for 5 min. After centrifugation, a small red cell pellet was formed on the bottom, which might be stromal vascular fraction cells. As shown in Figure 3, cells from the BAs layer were tdTom positive cells (Figure 3F); however, cells recovered from the pellet were tdTom negative (Figure 3G). Additionally, no obvious lipid droplets were visible in the tdTom negative cells. These data suggest that our new protocol can efficiently separate the BAs from the non-fat cells.

Together, these data demonstrate that isolating BAs with 3% BSA solution interferes with following-up biochemistry studies and suggest that 6% iodixanol solution is better than 3% BSA solution for isolating BAs.

Gene and protein expression analysis with isolated BAs.
For validating this new BAs isolation procedure on molecular level, the expression of three genes was compared between BAT and isolated BAs: Ucp1, Pdgfa and Pdgfra. In BAT, Pdgfra is expressed in endothelial cells and interstitial cells, and PDGFRα positive cells are putative progenitor cells¹⁰. The mRNA levels of Ucp1 and Pdgfa were both significantly higher in the isolated BAs than in the BAT (Figure 4A,B). On the contrary, the mRNA of Pdgfra was only detected in BAT (Figure 4B).

PPARγ is a transcriptional factor controlling adipose tissue development, UCP1 is a mitochondria protein, and PDGFRα is a membrane receptor protein. These three proteins represent proteins distributed in different cellular compartments. Western blots were performed to test whether protein extracted from Trizol-lysed BAs and BAT was suitable for protein expression analysis. UCP1 and PPARγ were detected in both BAs and BAT (Figure 4C,D), confirming that the total protein isolated from the Trizol-lysed BAs or BAT is suitable for western blot. Furthermore, consistent with the qRT-PCR results, UCP1 protein was enriched in BAs (Figure 4C); whereas PDGFRα was only detected in BAT but not in pure BAs (Figure 4D). In summary, these data demonstrate that our new BAs isolation method is efficient and suggest that BAs enriched by this method can be directly used for gene and protein expression studies.

Figure 1: Preparation of iBAT for brown adipocytes isolation. (A) Workflow of the brown adipocytes isolation procedure. (B) Ventral view of the interscapular tissue containing BAT, WAT and muscle layers. (C) Interscapular BAT (iBAT). Muscle layers and WAT adjacent to iBAT were removed. (D) A representative image of iBAT pieces used for brown adipocytes isolation. B-D, Scale bar= 5 mm. Please click here to view a larger version of this figure.

Figure 2: Separation of brown adipocytes from digestion solution. (A) Images of dissociated brown adipocytes before and after separation. 3% BSA solution was used to separate brown adipocytes from non-brown adipocytes. Scale bar = 1 cm. (B) Schematic view of genetic labeling brown adipocytes with tdTomato fluorescence protein. (C) Images of isolated brown adipocytes. DIC, differential interference contrast. Scale bar = 50 µm. Please click here to view a larger version of this figure.

Figure 3: Extraction of total protein from lysed brown adipocytes. (A) Separation of protein phase from the organic phase. (B) Coomassie staining of SDS-PAGE gel. Total protein was extracted from BAT or 3% BSA solution purified brown adipocytes. BSA protein band was indicated by an arrow. (C) Brown adipocytes layer formed on top of 6% iodixanol solution. Scale bar = 1 cm. (D) Brown adipocytes isolated with 6% iodixanol solution. Brown adipocytes were indicated by yellow arrows. Scale bar = 50 µm. (E) Coomassie staining of SDS-PAGE gel. Total protein was extracted from BAT or 6% iodixanol solution enriched BAs. (F) Images of tdTom labeled brown adipocytes. (G) Images of cells recovered from the iodixanol solution below the BAs layer. F and G, scale bar = 50 µm. Please click here to view a larger version of this figure.

Figure 4: Gene and protein expression analysis of isolated brown adipocytes. Brown adipocytes isolated with the iodixanol method was used in these gene and protein expression studies. (A,B) qRT-PCR measurement of gene expression. mRNA levels were normalized to 36B4. N=3. Student t test, *, P<0.01; **, P<0.01. (C) Western blotting of PPARγ and UCP1. (D)Western blotting of PDGFRα. C and D, Ponceau S-stained membrane was used as loading control. Please click here to view a larger version of this figure.

In this study, a new method of isolating BAs for gene and protein expression analysis was developed.

In a published BAs isolation protocol, 3% BSA solution was used to enrich BAs¹². Nevertheless, the enriched BAs achieved by this published protocol could not be directly used for protein expression analysis. This is because the concentrated BSA existing in the BAs solution interferes with following-up protein extraction. When the BAs enriched in the 3% BSA solution were treated with Trizol reagent, sticky protein aggregates would form, the majority of which was not soluble in the GTPC protein extraction process. Additionally, in the GTPC protein extraction product, most of the protein was BSA (Figure 3B). In this new protocol, 3% BSA separation solution was replaced with 6% iodixanol for purifying BAs. 6% iodixanol solution efficiently separated the BAs from non-BAs, and the isolated BAs had preserved morphology (Figure 3D). Superior to 3% BSA solution, 6% iodixanol solution did not interfere with protein extraction, and the extracted protein was suitable for western blot analysis (Figure 4C,D).

Adipose tissue contains large amount of lipids, and lipids contamination in extracted protein samples obstructs protein concentration measurement. Recently, a protocol to remove lipids from protein extractions has been published. In this protocol, a series of low temperature centrifugations are required, which is tedious and needs a large amount of starting materials¹⁹. In the current study, an improved GTPC method¹⁵ was adopted to isolate protein from BAT. Different from the classical GTPC protocol, this improved GTPC protocol used ethanol, bromo-chloropropane and water to extract protein out of the organic phase. Our data showed that protein isolated with this improved GTPC method was compatible with bicinchoninic acid assay (BCA) based protein concentration measurement.

In the current BAs isolation protocol, before the start of the enzyme dissociation process, the BAT and digestion solution mixture were placed on ice for one hour. The purpose of this procedure is to reduce the cell metabolism and gene expression rate²⁰, as well as to let the digestion enzymes efficiently perfuse into the brown adipose tissue.

The current study aimed to develop a straitforward method to isolate brown adiopocytes from interscapular BAT for gene expression study; however, white adipocytes contamination may exist in the isolated BAs. This is because interscapular BAT is sometimes attached by white adipose tissue, which cannot be completely removed during the BAT preparation step. The current protocol cannot separate BAs from WAs based on cell density. Therefore, if very high purity is essential, the isolated BAs need to be sorted by FACS before being analyzed.

The interscapular BAT (iBAT) contains abundant classical BAs that derived from the Myf5 cell lineage during embryonic development²¹. Here iBAT was used as an example for isolating BAs. Similar with the published protocol¹², our method can also be used for isolating beige cells from WAT. Nevertheless, to acquire pure beige cells from WAT, the beige cells need to be labeled with fluorescence protein and be enriched by FACS. Following our current protocol, the FACS enriched beige cells can be used for both gene and protein expression analysis, which will greatly improve the efficiency of biological materials utilization.