Fluorescent Protein-Assisted Purification for Gene Expression Profiling Protocol

Overview

Introduction of Fluorescent Protein-Assisted Purification

In vivo real-time tracking of cell movement and measurement of cellular growth in animal models by targeted expression of bioluminescence markers, combined with the ability to purify specific cell types of interest from a complex tissue at a particular point in time via expression of fluorescent proteins are invaluable tools available to biomedical scientists. These objectives are achieved by targeting "molecular beacons" to specific cell types, either endogenously by expressing beacon-encoding genes under the control of cell type-specific gene promoters in transgenic mice, or exogenously by expressing these genes via vectors in ex vivo or cultured cells followed by transplantation in recipient animals. These techniques provide an excellent opportunity to study specific cell types not only at the cellular level, but also at the molecular level. The two most commonly used molecular beacons are firefly luciferase, encoded by the luc gene, and the green fluorescent protein (GFP) from jellyfish. GFP, and other fluorescent proteins, provide a unique advantage in that they allow the purification of live-targeted cell types residing in complex heterogeneous tissues, which can then be subjected to molecular analyses, including gene expression profiling. For the purpose of fluorescence-activated cell sorter (FACS) purification, targeted expression of the fluorescent proteins in the cells of interest is superior to labeling cells with fluorescent antibodies because: (1) The cell type of interest may not have a unique cell surface antigen that could be used for antibody labeling, and the use of antibodies against antigens that may be present on other cell types in the tissue risks cross-contamination; (2) Even if a unique cell type-specific cell surface antigen is known, a good antibody against that antigen may not be available; (3) Intracellular antigens cannot be used because those would require fixation and permeabilization of cells, and thus preclude the purification of live cells, which is required for obtaining high-quality RNA for gene expression profiling. Here, we describe two examples of purification of specific cell types from mouse tissues. The following protocols for the preparation of single-cell suspensions of mouse tissues for FACS-purification of fluorescently labeled cells can easily be applied to a wide variety of tissue types. The protocol for the isolation of total RNA from a relatively small number of FACS-purified mammalian cells can be applied universally.

Fluorescence Labeling of Rare Cells in Transgenic Mouse Skin

One of the most exciting applications of flow sorting is the ability to purify rare cell types within complex animal tissues. Tumbar et al. have used a tetracycline inducible system to target GFP to the epidermis and hair follicles in the skin, and identified rare skin stem cells. They used two transgenic mice: (1) K5-tetVP16 transgenic mice expressing the tetracycline-regulated transactivator under the control of the epidermal keratinocyte-specific keratin-5 promoter, and (2) Mice harboring a transgene-containing histone H2B-GFP fusion transcript under the control of tetracycline response element (TRE) promoter. These two transgenic mice were crossed to obtain double transgenic mice that expressed GFP in K5-expressing cells in the skin. This tet responsive system (Tet-off) allowed GFP expression to be shut off in the presence of tetracycline (or its analog doxycycline). Feeding double transgenic mice doxycycline-fortified chow shuts off any new expression of GFP. While the rapidly cycling skin cells completely lose GFP signal within 4–8 weeks after the start of doxycycline diet, the slow-cycling stem cells retain GFP signal even 4 months later. This small population of "label retaining cells" can be selectively FACS-purified from the skin of these mice following enzymatic digestion and mechanical dissociation of the skin tissue into a single-cell suspension.

Bifunctional Reporter-Aided Monitoring and Purification of Cancer Cells

Tumors are complex heterogeneous mixtures of cancer cells and a variety of noncancerous cell types. For gene expression profiling studies of cancer cells with respect to their growth properties and/or metastatic potential, it is crucial to purify the tumor cells away from the noncancerous tumor-associated cells (stromal cells). Coexpression of luciferase and GFP can be used to noninvasively image and track the growth, trafficking, and metastatic spread as well as FACS-purification of transplanted cancer cells from primary tumors as well as metastatic sites. We have developed lentiviral (LV) reporter that express both luciferase and GFP from a fused transcript under the control of the Pol2 and FerH promoters. We linked these two reporter genes (Luc/ GFP) to ensure colinear signals from both reporters. This fusion reporter strategy can also overcome the failure of LV infection to reach 100% efficiency because the Luc/GFP-labeled cells can be FACS-purified. We have determined that the Pol2 promoter provides sustainable long-term expression of the Luc/GFP reporter even in immunocompetent mice. Luciferase can be used to monitor the primary tumor growth, as well as metastatic spread to distant sites. Metastatic cells from distant metastatic tumors, and even dormant micrometastases, can then be FACS-purified using the GFP label. Here, we describe a method for tumor cell infection with LV-Luc/GFP, transplantation in syngeneic mice, and FACS purification of lung metastases.

Materials of Fluorescent Protein-Assisted Purification

All experiments with mice must be performed under protocols approved by the Institutional Animal Care and Use Committees and the strict NIH guidelines for humane use of research animals.

Reagents and Cells

Liberase TL. Reconstitute at 38.5 mg/mL (200 U/mL) in dH₂O. Keep in small aliquots at −20°C for 6 months. Avoid repeated freeze/ thaw cycles.
DNAse I. Reconstitute 1 g in 20 mL dH2 O for a final stock concentration of 50 mg/ mL. Invert gently to mix. Do not vortex as DNAse I is sensitive to physical force. Store at −20°C in 1 mL aliquots for 1 year.
Digestion medium: 5 mL DMEM, 50 mL Liberase TL stock solution (0.385 mg/mL final concentration).
DFD solution: 24 mL DMEM, 6 mL fetal bovine serum (FBS), 300 mL DNAse I stock solution. Mix by inverting and chill on ice. Do not vortex, as DNAse I is sensitive to physical force. Make fresh as needed.
PBSA solution: 50 mL 1× phosphate-buffered saline (PBS), 0.5 g bovine serum albumin (BSA powder, store at 4°C). Shake well to dissolve. To dissolve completely, incubate at 37°C for 30 min. Shake well and then filter through a 0.45- mm syringe filter to eliminate any undissolved BSA particles. Chill on ice.
Cell culture medium: DMEM supplemented with 10% FBS.
Transfection medium: Opti-MEM I.
Inoculation medium: RPMI 1640.
ACK lysis buffer.
HEK-293 T cells.
ViraPower LV Expression System.
Lipofectamine 2000.
Trizol reagent.
70% ethanol: dilute absolute ethanol with DEPC-treated water.
80% ethanol: dilute absolute ethanol with DEPC-treated water.
RNeasy RNA isolation kit, Mini or Micro.

Equipment and Supplies

Dissection instruments; fine-tooth shaver or hair removal cream.
Medimachine system, 50-mm Medicons.
Filters: Filcon 70-mm purple filters, syringe-type; Filcon 30-mm green filters, cup-type; 70-mm white cell strainers.
20-mL Luer Lock syringes, with 18-gauge 1–0.5 in. needles.
Fine balance.
Refrigerated benchtop centrifuge with bucket rotors for 15-mL conical Falcon tubes.
Xenogen bioluminescence imaging system.
FACS.

Methods of Flow Cytometric FRET Analysis

Isolation of GFP+ Mouse Skin Cells by FACS

Obtaining Mouse Skin Tissue

Euthanasia: mice older than 1-week should be euthanized by CO2 asphyxiation. For certain experimental protocols, with IACUC approval, cervical dislocation can be used for euthanizing mice (see Note 1).
Shave the back skin as closely as possible. Alternatively, hair removal cream can be used under some circumstances (see Note 2).
Take the entire back skin, from shoulders to just above the tail, and put in a dish with 1× PBS on ice.
Immerse the skin in a second dish with cold 1× PBS and using two curved fine forceps remove all the fat and blood vessels from the underside of the skin. Scrape off fat layer until translucent skin is exposed from underneath. Put the cleaned skin in another dish in 1× PBS on ice (see Note 3).
Cut the skin tissue in small pieces, ~2–4 mm long, and weigh on a fine balance.

Preparation of Single-Cell Suspensions from Mouse Skin

Put 0.6–1 g of skin per 5 mL of digestion medium in a 15-mL Falcon tube. Invert the tube a few times to mix and incubate in 37°C water bath for 1 h. Mix once by inverting a few times at the mid-point of incubation (see Note 4).
Put the digested tissue in a dish on ice. Keep the cells cold as much as possible from here on.
Put a few skin pieces inside the Medicon chamber while rotating the blades, enough to cover the bottom of the chamber. Put the lid on, insert the Medicon into the Medimachine and run for 2–2.5 min (see Note 5).
Put two or three 15-mL conical Falcon tubes on ice and insert 70-mm purple filters in each. Using a 20-mL syringe, dispense 2 mL of DFD solution through the filters into each tube.
When the Medicon run is complete, take it out of the Medimachine and take the lid off. Using an 18-gauge needle on the 20-mL syringe, dispense DFD into the Medicon chamber to fill it. Then, insert the needle all the way into the little hole on the side and draw the DFD from the bottom of Medicon into the syringe. After removing the needle from the syringe, put this single-cell suspension in the Falcon tube through the purple filter (see Note 6).
Remove the remaining chunks of tissue from the Medicon chamber and repeat the cycle with new tissue pieces. When done with the Medicon, rinse it with one extra volume of DFD to completely draw any remaining cells.
When all the tissue has been processed into a single cell suspension, centrifuge the Falcon tubes at ~300×g for 5 min at 4°C.
Aspirate out as much of the supernatant as possible without disturbing the cell pellet. Resuspend the pelleted cells in one tube in 8 mL of PBSA. Transfer this suspension to other tubes sequentially to combine all of the cells.
Filter the pooled cell suspension through a 70-mm white strainer into a 50-mL Falcon tube. Rinse all the tubes with another 5 mL of PBSA and then rinse through the white strainer to combine.
Transfer the pooled cell suspension (13 mL) to a 15-mL Falcon tube and centrifuge at ~300×g for 5 min at 4°C. Aspirate the supernatant.
Resuspend the cell pellet in 8 mL of PBSA and transfer to a new 15-mL Falcon tube. Rinse the old tube with 5 mL of PBSA and combine into the new tube. Repeat centrifugation. Aspirate as much of the supernatant as possible without disturbing the pellet (see Note 7).
Suspend the final pellet in 2–4 mL of PBSA. This cell suspension can be directly used for the isolation of GFP+ cells on a FACS equipped with a 488-nm laser and a 530/30 bandpass filter. Just before FACS sorting, filter the cell suspension through a 30-mm green filter to remove clumped cells and debris (see Note 8).

In Vivo Tracking of Luc/GFP-Expressing Tumor Cells and Lung Metastases

The construction of pSico-Luc/GFP fusion reporter lentivirus vector has been described.

Production of Lentivirus

The day before transfection (day 1), plate 5×106 HEK-293 T cells in T-75 flask in 15 mL of DMEM culture medium containing 10% FBS. Do not include antibiotics in culture medium.
On the day of transfection (day 2), remove the culture medium from the HEK-293 T cells and replace with 7.5 mL of Opti-MEM I medium.
In a sterile 5-mL tube, mix 9 mg of the ViraPower™ Packaging Mix and 3 mg of LV expression plasmid DNA (12 mg total) in 1.5 mL of Opti-MEM I medium without serum. In a separate sterile 5-mL tube, dilute 36 mL of Lipofectamine™ 2000 in 1.5 mL of Opti-MEM^® I Medium without serum. Mix gently and incubate for 5 min at room temperature.
Combine the diluted DNA with the diluted Lipofectamine™ 2000 and mix gently (no pipetting). Incubate for 20 min at room temperature and then add the DNA-Lipofectamine™ 2000 complexes dropwise to each plate of cells. Mix gently by rocking the plate back and forth. Incubate the cells overnight. The next day (day 3), remove the medium and replace with 15 mL of complete culture medium without antibiotics.
24 h after changing medium (day 4), examine the cells with fluorescence microscope. HEK-293 T cells should express substantial level of GFP if transfection is successful.
Harvest virus-containing supernatants twice at 48 and 72 h posttransfection (days 4 and 5, respectively) by transferring medium into a 50-mL sterile, capped, conical tube, and then centrifuge supernatants at 2,000×g for 15 min at 4°C to pellet debris. Pipet viral supernatants into cryovials in 1 mL aliquots. Store viral stocks at –80°C, which can maintain the titer of the virus stably up to one year. Repeated freezing and thawing is not recommended.
If it is necessary to concentrate the virus supernatant, load the supernatant in the outer chamber of an Amicon centriprep-10 unit (Millipore Corp., Bedford, MA) and centrifuge at 3,000×g for 45–60 min at room temperature or 4°C. Discard the ultrafiltrate from the inner chamber and repeat the centrifugation step. It usually takes 2.5–3 h to concentrate supernatant 10- to 20-fold. Recovery of virus can be quantitated by titering material pre- and post-concentration. Losses are usually about 20%.
Titer the viral supernatant on a standard human adherent cell line, such as HOS cells (human osteosarcoma), HeLa, or HEK-293 cells. Plate the target cells 1 day prior to transduction in 24-well format, at 50,000 cells/well (day 1). In the next day, count the cell number from two wells (day 2). This would be the cell number used in the calculation of virus titer later. Incubate increasing amounts of virus supernatant (e.g., 2, 5, 10, 20, 50, 100, and 200 mL) with the target cells. Change the medium the next day. Two days later, observe the transduced cells with fluorescence microscope to confirm the expression of GFP. Bring the cells for flow cytometric analysis to measure GFP-positive percentage (% GFP+), and then identify the virus supernatant amount to reach 1%<%GFP+<10%. In this range, cells are assumed to be transduced with a single copy of virus. The titer is calculated as following: Infectious Units (IU)/mL=([cell number counted at day 2]×%GFP+)/(virus supernatant amount in mL×10⁻³).

Labeling of Tumor Tissue and In Vivo Monitoring of Growth and Metastasis

Put 1 g of fragmented B16BL6 melanoma tumor tissue freshly harvested from subcutaneous sites of mice per 5 mL of digestion medium in a 15-mL Falcon tube. Invert the tube a few times to mix and incubate in 37°C water bath for 30 min. Mix once by inverting a few times at the mid-point of incubation (see Note 4). 2. To make single-cell suspension.
Aspirate out as much of the supernatant as possible without disturbing the cell pellet. Resuspend the pelleted cells in one tube in 8 mL of PBSA. Count cell number and transfer 10⁶ cells into 15-mL Falcon tubes. Centrifuge at ~300×g for 5 min at 4°C.
Aspirate the supernatant and resuspend the cell pellet with 500 mL of virus supernatant with IU higher than 10⁷. Adjust the final volume to 1–1.5 mL with RPMI 1640 serum-free medium.
Centrifuge at 800–1,200×g for 0.5–1 h (depending on the fragility of tumor cells). Aspirate the supernatant. Resuspend the cells with 100 mL of serum-free RPMI 1640 medium and inoculate subcutaneously into mice immediately. Caution: use rotors with biosafety cover for centrifugation.
Monitor the tumor growth by measuring the physical size and bioluminescence imaging on the Xenogen bioluminescence imaging system. When the tumors reach 1 cm³ in mice, harvest tumor and prepare cell suspension following steps 1–3.
Resuspend the pelleted cells in 5 mL of ACK lysis buffer and incubate at room temperature for 10 min. Centrifuge at 300×g, 4°C, for 10 min. Resuspend the cell pellet with 8 mL of PBSA. Transfer this suspension to other tubes sequentially to combine all of the cells.
Filter the pooled cell suspension through a 70-mm white cell strainer into a 50-mL Falcon tube. Rinse all the tubes with another 5 mL of PBSA and then rinse through the white strainer to combine. Transfer the pooled cell suspension (13 mL) to a 15-mL Falcon tube and centrifuge at ~300×g for 5 min at 4°C. Aspirate the supernatant.
Resuspend the final pellet in 2–4 mL of PBSA. This cell suspension can be directly used for FACS.
Following sorting, pool the sorted GFP+ cells, centrifuge at 300×g, 4°C, for 10 min. Aspirate the supernatant and resuspend the cells with serum-free RPMI 1640 medium to reach 106 cells/100 mL.
Inoculate 106 tumor cells subcutaneously into each mouse. The tumors are uniformly labeled with luciferase-GFP fusion gene, allowing in vivo bioluminescence monitoring and isolation of GFP+ cells from the harvested tumors.

Preparation of Total RNA from Sorted Cells for Gene Expression Profiling

Centrifuge sorted GFP+ cells for 10 min at ~300×g at 4°C.
Remove as much of the supernatant as possible without disturbing the pellet (see Note 9).

Figure 1. Monitoring of B16BL6 melanoma tumors labeled with Pol2-Luc/GFP lentiviral vector by bioluminescence (BL) imaging.

3. Lyse the cells in 1 mL of Trizol reagent by vortexing vigorously. Transfer to an Eppendorf tube. This lysate can be stored indefinitely at −80°C.

4. Add 200 mL of chloroform and shake vigorously (or vortex at low setting) for 15 s. Let stand at room temperature for 5 min.

5. Centrifuge at 20,000×g for 15 min at 4°C.

6. Carefully take the aqueous (top) phase and transfer to a new Eppendorf tube. Measure the volume transferred.

7. Add an equal volume of 70% ethanol and vortex to mix.

8. Transfer the mixture to RNeasy MinElute spin column (see Note 10). Centrifuge at 10,000×g for 15 s at room temperature. If the volume is more than 700 mL, reload the rest to the column until all the volume has been run through the column. Discard all flow through.

9. Apply 700 mL of RW1 buffer (included in RNeasy kit) and centrifuge at 10,000×g for 15 s. Discard flow through and collection tube.

10. Transfer spin column to a new collection tube (provided in the kit). Apply 500 mL of RPE buffer (ethanol added, see Note 11) to the column and centrifuge at 10,000×g for 15 s. Discard the flow through.

11. Add 500 mL of 80% ethanol to the spin column and centrifuge at 10,000×g for 15 s. Discard the flow through.

12. Repeat wash with 500 mL of 80% ethanol, and centrifuge for 2 min at 10,000×g (see Note 12).

13. Transfer column to a new collection tube. Open the cap of the spin column and centrifuge at maximum speed for 5 min to dry the column completely (see Note 13). Discard the flow through and collection tube.

14. Transfer column to an Eppendorf tube. 10-14 uL (for Micro kit) or 30-50 uL (for Mini kit) of RNAse-free water directly in the center of the column membrane and centrifuge for 1 min at maximum speed to elute RNA.

15. Store RNA at −20°C, or at −80°C for long-term storage.

Notes of Fluorescent Protein-Assisted Purification

CO₂ asphyxiation is the method of choice for euthanasia of mice. However, if exposure to CO₂ is suspected to adversely affect the gene expression profile of the cells of interest, then cervical dislocation can be used. Cervical dislocation must be performed with appropriate IACUC approval and by experienced personnel.
Hair removal cream can be used, but this can adversely affect the gene expression profiles in the epidermal cells.
Removing all the blood vessels and especially fat from the underside of the skin is extremely important because the presence of fat in the final cell suspension causes cells to clump and interfere in the FACS sorting. Do not use too much 1× PBS while scraping because it causes the skin to float up from the sides and makes it difficult to scrape. Scrape the underside of the skin until you expose the translucent skin. Remove as much of the fat as possible.
Liberase TL is a cocktail of collagenases and neutral proteases (NP). Keep in mind that given enough time, NP damage the surface proteins and increase cell death. Incubating for longer periods and/or using more Liberase are thus counter-productive. Liberase TL has an NP:Collagenase ratio that is much lower than the majority of other available Liberase formulations. This relatively lower NP level in the TL formulation is important to keep the cell viability relatively high. However, at least some level of cell death is expected with 1 h incubation. On the other hand, 1 h incubation is important for optimal dissociation of tougher (fibrous) tissues, e.g., skin, and thus higher yield in FACS. For less fibrous tissues, 30–45 min incubation is sufficient.
Do not overfill the Medicon chamber with tissue, as it might jam. The time of medimachine run can be extended up to 4 min for tougher and fibrous tissues to increase tissue dissociation. Do not exceed 4 min of running time, as longer runs only increase fine debris (e.g., connective tissue) that would clog the Medicon membrane. You can use two Medicons at a time so that you can process/refill one Medicon while the other is running. Change to a new pair midway. This is because the Medicon membrane clogs with debris and cells are trapped.
Change the 70-mm purple filters when you change the Medicons midway.
The cell pellet is mostly translucent. However, depending on how well the blood vessels and/or blood have been removed, there may be a reddish hue to the pellet. Small amounts of red blood cells would not affect the FACS yield.
Before filtering through the 30 mm filter, it is very important to dilute the final cell suspension appropriately so that it is not too cell dense. Too much cell density causes large clumps, which are filtered out and reduce yield. Large cell pellets should be diluted to 3.5–4 mL before filtration. Smaller pellets should be diluted to 2–3 mL. Volumes larger than 4 mL take too long to process through the FACS.
Depending on how many cells have been collected by sorting, a pellet may or may not be visible. Generally, cell pellets containing less than 2×10⁵ cells are not visible. In such cases, be careful while aspirating the supernatant. You can leave a little bit of the supernatant in the bottom, as long as it is less than 10% of the volume of Trizol used.
Depending on the number of cells being processed, either RNeasy Mini or Micro kit can be used. Use the Mini kit for larger number of cells (millions), and Micro for hundreds to less than a million cells.
Buffer RPE is provided in the kit as a concentrate and absolute ethanol must be added before use. See the bottle label for how much ethanol to add.
The second 80% wash ensures that no salt is carried over to the eluate. Even small amounts of carryover salts can inhibit subsequent labeling enzymatic reactions.
It is important to dry the column completely before RNA elution to make sure that no ethanol is carried over, which can inhibit subsequent enzymatic reactions.