Xenobiotic efflux pumps are highly active in hematopoietic stem and progenitor cells (HSPCs) and cause extrusion of TMRM, a mitochondrial membrane potential fluorescent dye. Here, we present a protocol to accurately measure mitochondrial membrane potential in HSPCs by TMRM in the presence of Verapamil, an efflux pump inhibitor.
As cellular metabolism is a key regulator of hematopoietic stem cell (HSC) self-renewal, the various roles played by the mitochondria in hematopoietic homeostasis have been extensively studied by HSC researchers. Mitochondrial activity levels are reflected in their membrane potentials (ΔΨm), which can be measured by cell-permeant cationic dyes such as TMRM (tetramethylrhodamine, methyl ester). The ability of efflux pumps to extrude these dyes from cells can limit their usefulness, however. The resulting measurement bias is particularly critical when assessing HSCs, as xenobiotic transporters exhibit higher levels of expression and activity in HSCs than in differentiated cells. Here, we describe a protocol utilizing Verapamil, an efflux pump inhibitor, to accurately measure ΔΨm across multiple bone marrow populations. The resulting inhibition of pump activity is shown to increase TMRM intensity in hematopoietic stem and progenitor cells (HSPCs), while leaving it relatively unchanged in mature fractions. This highlights the close attention to dye-efflux activity that is required when ΔΨm-dependent dyes are used, and as written and visualized, this protocol can be used to accurately compare either different populations within the bone marrow, or the same population across different experimental models.
Hematopoietic stem cells (HSCs) are self-renewing, multi-potent, and capable of giving rise to all the cells of the blood1,2. Cellular metabolism is a key regulator of HSC maintenance, along with transcriptional factors, intrinsic signals and the microenvironment3,4,5. The proper control of mitochondrial function and quality is therefore critical to HSC maintenance6,7.
Mitochondrial membrane potential (ΔΨm) is a key parameter in the assessment of mitochondria as it directly reflects their functionality, which derives from the equilibrium of proton pumping activity in the electron transport chain and the proton flow through F1/FO ATP synthase. These are both required (depending on gene expression and substrate availability) for the oxygen-dependent phosphorylation of ADP to ATP8,9. Taking advantage of the electronegativity of the mitochondrial compartment, various potentiometric dyes have been developed to measure ΔΨm. One of them is tetramethylrhodamine methyl ester perchlorate (TMRM), which has been extensively used to measure ΔΨm by flow cytometry in a variety of cells10, including hematopoietic stem and progenitor cells11.
Mitochondrial dyes must be used with some caution in HSCs, however, because the high activity of the xenobiotic efflux pumps of these cells can result in dye extrusion12. Indeed, the extrusion of mitochondrial dyes such as Rhodamine 123 has allowed researchers to isolate HSCs13 or identify HSC “side populations” by exploiting the differential extrusion of the dyes Hoechst Blue and Hoechst Red14,15. It has also been shown that Fumitremorgin C, a specific blocker of the ATP-binding cassette sub-family G member 2 (ABCG2) transporter, does not affect the staining pattern of MitoTracker in HSPCs16. After the publication of these results, multiple studies were performed using mitochondrial dyes in the absence of xenobiotic efflux pump inhibitors, leading to the widespread impression that HSCs have only a small number of mitochondria with low ΔΨm16,17,18.
Recently, it was demonstrated, however, that Verapamil, a wide spectrum inhibitor of efflux pumps, significantly modifies the staining pattern of the mitochondrial dye MitoTracker Green19. This discrepancy is likely due to the fact that Fumitremorgin C is highly selective for Abcg2, while HSCs also express other transporters such as Abcb1a (which is only weakly sensitive to Fumitremorgin C)19. We have also reported that other mitochondrial dyes, such as TMRM, Nonyl acridine orange, and Mitotracker Orange (MTO) exhibit the same patterns as Mitotracker Green. More importantly, we have observed that the flow cytometric patterns of HSPCs reflect their ΔΨm in addition to mitochondrial mass11.
The intake of TMRM dye strictly depends on the negative charge of mitochondria, but the resulting accumulation of dye is in constant balance between its intake and clearance by efflux pumps20. The difference in xenobiotic efflux pump expression between HSCs and mature cell populations affects this balance and can lead to biased results. The use of dedicated inhibitors such as Verapamil should be considered in the analysis of ΔΨm by potentiometric dyes. Here we describe a modified protocol for accurate ΔΨm measurement by TMRM-based flow cytometry which corrects for xenobiotic transporter activity through the use of dedicated inhibitors.
All methods described here have been approved by the Institutional Animal Care and Use Committee (IACUC) of the Albert Einstein College of Medicine.
1. Preparation of Solutions
2. Bone Marrow Isolation
3. Immunostaining for Detection of HSC
4. TMRM Staining
5. Acquisition by Flow Cytometer
The protocol described above enables the easy isolation of BM-MNCs from a mouse model. Figure 1 summarizes the main steps of the protocol: bone isolation, flushing out of the bone marrow, red blood cell lysis, and antibody staining followed by TMRM staining to measure mitochondrial membrane potential in a specific hematopoietic population.
BM-MNCs contain several cell populations, including HSCs. The antibody cocktails used in this protocol are well-established in the purification of HSCs (CD34− and CD34+), multipotent progenitor cells (MPPs), Lin− as well as Lin+ cells, respectively 21. The gating strategy for isolating these fractions is shown in Figure 2.
After the identification of populations of interest, TMRM intensity, which should appear as a bright signal, was assessed. TMRM staining in serum-free expansion medium (SFEM) is highly recommended, as TMRM profiles in HSPCs can undergo alteration when staining is performed in PBS +2% FBS (Figure 3A).
Figure 3C shows the average intensity of each population, which is normalized by the intensity of BM-MNCs. HSCs express high activity levels of xenobiotic efflux pumps capable of extruding TMRM dye20, and indeed, we found TMRM profiles in HSPCs were changed in the presence of Verapamil (Figure 3B,C). Similar results were obtained by other inhibitors such as Cyclosporin H (Figure 3C). Thus, the accurate amount of TMRM loaded in the mitochondria by ΔΨm can be measured after inhibition of the efflux pumps by Verapamil or Cyclosporin H (Figure 3C).
Finally, FCCP can be used to verify the accuracy of TMRM staining. FCCP depolarizes mitochondria, resulting in a reduction in TMRM intensity (Figure 3D). This approach can also be used to determine the background intensity of the staining and/or as a negative control.
Figure 1: Protocol flowchart. Graphical summary of the procedure to isolate and stain BM-MNCs to determine ΔΨm. Critical steps are highlighted by picture inserts (A-D). Femurs and tibias from adult C57BL/6 mice were isolated and their ends are removed (A). Long bones as in A after flush out (B). Isolated BM-MNCs before (C) and after (D) ACK lysis. Please click here to view a larger version of this figure.
Figure 2: Gating setup. Schematic representation of gating strategy to identify the different hematopoietic populations, including CD34−-HSC and CD34+-HSC, MPP, Lin− and Lin+ cells. The panels were modified from Bonora, M. et al.11. Please click here to view a larger version of this figure.
Figure 3: Flow cytometry analysis of mitochondrial membrane potential. (A) Representative distribution of ΔΨm in HSCs stained with TMRM in PBS+2%FBS (green) or in serum-free expansion medium (SFEM) (red). (B, C) Representative distribution of ΔΨm in CD34−-HSC and Lin− cells (B) and quantification of ΔΨm in CD34−-HSC, CD34+-HSC, MPP, Lin− and Lin+ cells (C) stained with TMRM in presence or absence of efflux pump inhibitors. TMRM intensity of each population was normalized by the TMRM intensity of own BM-MNCs (modified from Bonora, M. et al.11). (D) Representative histogram of TMRM intensity distribution in HSCs before (pink) and after (light blue) FCCP addition. Please click here to view a larger version of this figure.
Mitochondrial membrane potential measurement is a cornerstone of the analysis and assessment of mitochondria, which are critical to the metabolic state of the cell. Here, we describe a protocol for the analysis of ΔΨm by TMRM staining. TMRM is a cell-permeant fluorescent dye which accumulates in active mitochondria due to ΔΨm, and its respective levels remain in equilibrium between the extracellular, cytoplasmic and mitochondrial compartments10. This protocol can be adapted for various dyes, including tetramethylrhodamine, ethyl ester (TMRE) and JC-1. Appropriate staining conditions are critical to achieving accurate results. These include: protection from light exposure, proper incubation time, maintenance of constant temperature (37 °C) and extracellular concentration. When TMRM dye levels have not yet reached perfect equilibrium, unexpected changes in staining intensity can be detected during data acquisition in the flow cytometry process. Therefore the recommended staining period for TMRM is 1 h rather than 30 min (as mentioned in one TMRM manufacturer’s protocol). It is also suggested that the sample should be acquired at least two times within 5-min in order to compare the stability of the dye.
Another critical step of the protocol is the setting up of an efficient color match among antibodies and dye. Hematopoietic populations can be detected by flow cytometry using well-established surface markers21.The gating strategy described here (Figure 2) is compatible with TMRM staining, and allows for the measurement of its intensity at different differentiation stages simultaneously. TMRM matches the PE channel, but its intensity is usually much weaker than that of antibodies conjugated with PE (data not shown). As this can affect color compensation, causing unexpected shifts in the spectra of some populations, the sample with TMRM staining should be used as a compensation control for the PE channel.
The major advantage of the current protocol is that it enables the removal of the effect of xenobiotic efflux pumps, whose efficient extrusion of TMRM dye modifies both its equilibrium across the plasma membrane and its mitochondrial accumulation. This is a critical step in the accurate assessment of HSC mitochondrial function by dye absorption. As HSCs express more active efflux pumps than committed cells19, Verapamil or similar drugs, such as Cyclosporine H, a Ca2+-independent multidrug resistance inhibitor22 (Figure 3B–C), should be used in the staining procedure for HSCs to more accurately assess levels of mitochondrial activity. Because this critical issue was pointed out only recently11,19 and is still under discussion23, the main aim of this report is to provide an easy and detailed protocol which will help to standardize the procedure for obtaining reproducible data and reduce the apparent discrepancies between the results of different research groups. The protocol here described shows in detail each step, including doses, timing and specific media. For instance, TMRM profiles in HSPCs can differ depending on their medium (Figure 3A). A detailed analysis of the mechanisms involved will require further exploration, but since serum-free expansion medium (SFEM) is a well-established method for HSC culture, SFEM rather than PBS is strongly recommended when performing TMRM staining for HSCs. Further, the accurate assessment of mitochondrial membrane potential demonstrated here through the inhibition of efflux pumps highlights possible future applications of Verapamil, as well as other dyes (e.g. MitoTracker, MitoSOX), in the investigation of HSCs.
Importantly, TMRM intensity as shown by flow cytometry may not precisely reflect mitochondrial volume. Flow cytometry measures total fluorescence intensity per cell, but the distribution of mitochondrial dyes to mitochondria depends on ΔΨm. It is therefore difficult to discern whether the critical contribution to a TMRM reading derives from ΔΨm or mitochondrial volume11. We have confirmed that MTO, one of the dyes most frequently used to measure mitochondrial mass, is affected by both ΔΨm and efflux pump activity. We have also observed that Verapamil changes the intensity profile of MTO in HSPC populations, but have found no correlation between MTO intensity and mitochondrial volume11. Combined measurement of mitochondrial volume and ΔΨm should be used to normalize data and eliminate the effects of membrane potential, and further exploration of mitochondrial content should be carried out using ΔΨm-independent techniques.
Such techniques would include 3D volume analysis based on fluorescent markers (e.g., immunostaining of mitochondrial proteins), electron microscopy11, and quantitation of mitochondrial/nuclear DNA ratios. The use of mitochondrial dyes (i.e., TMRM) in combination with efflux system inhibitors will doubtless prove of great benefit in the elucidation of mitochondrial biology through flow cytometry.
The authors have nothing to disclose.
The authors thank all members of the Ito laboratory, especially K Ito and H Sato, and the Einstein Stem Cell Institute for comments and the Einstein Flow Cytometry and Analytical Imaging core facilities (funded by National Cancer Institute grant P30 CA013330) for help carrying out the experiments. K.I. is supported by grants from the National Institutes of Health (R01DK98263, R01DK115577, and R01DK100689) and the New York State Department of Health as Core Director of Einstein Single-Cell Genomics/Epigenomics (C029154). K.I. Ito is a Research Scholar of the Leukemia and Lymphoma Society.
ACK lysing buffer | Life Technologies | A1049201 | |
B220-biotin | BD Bioscience | 553086 | |
CD3e-biotin | Life Technologies | 13-0031-85 | |
CD4-biotin | Fischer Scientific | BDB553782 | |
CD8-biotin | Life Technologies | 13-0081-85 | |
CD11b-biotin | BD Bioscience | 553309 | |
CD19-biotin | BD Bioscience | 553784 | |
CD34-FITC | eBioscience | 11-0341-85 | |
CD48-APC | eBioscience | 17-0481-82 | |
CD135-biotin | eBioscience | 13-1351-82 | |
CD150-PerCP/Cy5.5 | Biolegend | 115922 | |
c-kit-APC/Cy7 | Biolegend | 105826 | |
Cyclosporin H | Millipore Sigma | SML1575-1MG | |
DAPI solution (1mg/mL) | Life Technologies | 62248 | |
Fetal Bovine Serum (FBS) | Denville | FB5001-H | |
FCCP | Millipore Sigma | C2920-10MG | |
Gr1-biotin | Biolegend | 108404 | |
IgM-biotin | Life Technologies | 13-5790-85 | |
Il7Rα-biotin | eBioscience | 13-1271-85 | |
Nk1.1-biotin | Fischer Scientific | BDB553163 | |
Phosphate buffered saline (PBS) | Life Technologies | 10010023 | |
Sca-1-PE/Cy7 | eBioscience | 25-5981-81 | |
SCF murine | PEPROTECH | 250-03-10UG | |
StemSpan SFEM medium | STEMCELL technologies | 9605 | |
Streptavidin-Pacific Blue | eBioscience | 48-4317-82 | |
Ter119-biotin | Fischer Scientific | BDB553672 | |
TMRM | Millipore Sigma | T5428-25MG | |
TPO | PEPROTECH | 315-14-10UG | |
Verapamil hydrochloride | Millipore Sigma | V4629-1G |