SEARCH

SEARCH BY CITATION

Keywords:

  • ATP-binding cassette transporter;
  • Adult bone marrow stem cells;
  • Progenitor cells;
  • Fluoresence-activated cell sorting analysis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The ability of cells to export Hoechst 33342 can be used to identify a subpopulation of cells (side population [SP]) with characteristics of stem cells in many tissues. The ATP-binding cassette transporters Bcrp1 (Abcg2) and Mdr1a/1b (Abcb1a/1b) have been implicated as being responsible for this phenotype. To further explore the involvement of these transporters in the SP phenotype, we have generated Bcrp1/Mdr1a/1b triple knockout mice and studied the effect of their absence on the SP in bone marrow and mammary gland. Whereas in bone marrow Bcrp1 was almost exclusively responsible for the SP, both transporters contributed to the SP phenotype in the mammary gland, where their combined absence resulted in a nearly complete loss of SP. Interestingly, bone marrow of Mdr1a/1b−/− mice frequently displayed an elevated SP, which was reversible by the Bcrp1 inhibitor Ko143, suggesting that Bcrp1 can compensate for the loss of Mdr1a/1b in bone marrow.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

A subpopulation of cells with characteristics of stem cells can be identified by the side population (SP) phenotype, which is based on the efflux of the fluorescent dye Hoechst 33342. This was first shown by Goodell and colleagues, who speculated that the ATP-binding cassette (ABC) transporter MDR1 (P-glycoprotein/ABCB1) was responsible for this phenotype [1]. Recently, however, it has become clear that another ABC transporter, Breast Cancer Resistance Protein (BCRP/ABCG2), is mainly responsible for the SP phenotype, at least in bone marrow [26]. Based on the initial findings in the hematopoietic compartment, SP cells have now been identified in many other tissues including the mammary gland [79], skeletal muscle, pancreas, lung, retina, liver, testis, heart, and epidermis [1017]. In addition to normal tissues, it has further been demonstrated that cancer cell lines and primary tumor cells contain an SP [1820]. Those findings led to the proposition that tumors might also contain a minor subpopulation of drug-resistant “cancer” stem cells which might be crucial for their malignancy.

Although the exact physiologic roles of BCRP and MDR1 in stem cells are not yet known, it seems likely that an important function of these transporters is to protect them against the cytotoxic actions of xenotoxins or endogenous compounds. In line with this, it has been suggested that BCRP might protect the stem cell compartment against hypoxic stress by reducing heme or porphyrin accumulation [21].

Previously, we have identified an SP in human and mouse mammary glands and shown that they constitute an undifferentiated subpopulation that can differentiate into ductal and lobular structures and into myoepithelial and luminal epithelial cell types [7]. To further characterize the respective roles of Bcrp1 and Mdr1a/1b in the SP phenotype, we have generated Bcrp1/Mdr1a/1b triple knockout mice and determined the relative transporter contributions to the SP phenotype in bone marrow and mammary gland.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Animals

Mice were housed and handled according to institutional guidelines complying with Dutch legislation. The animals that were used were Bcrp1−/− [22], Mdr1a/1b−/− [23], Bcrp1−/−/Mdr1a/1b−/−, and wild-type mice of a >99% FVB genetic background, between 9–14 weeks of age. Animals were kept in a temperature-controlled environment with a 12-hour light/12-hour dark cycle. They received a standard diet (AM-II; Hope Farms, Woerden, The Netherlands, http://www.hopefarms.nl) and acidified water ad libitum.

Clinical Chemical Analysis of Plasma

Standard clinical chemical analyses on plasma were performed on a Hitachi 911 analyzer (Hitachi, Ltd., Tokyo, http://www.hitachi.com) to determine levels of bilirubin, alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, creatinine, urea, Na+, K+, Ca2+, Cl, phosphate, total protein, and albumin.

Isolation of Mouse Mammary Epithelial Cells

Primary mouse mammary epithelial cells were harvested essentially as described by Smalley et al. [24], with minor modifications [25]. In brief, the fourth mammary fat pads (inguinal) were removed from 8- to 10-week-old virgin females and subjected to mechanical and enzymatic digestion to obtain epithelial “organoids.” Contaminating fibroblasts were removed by differential plating, and the organoids were enzymatically digested to single cells. Primary cells were then directly processed for Hoechst staining without intervening culture.

Isolation of Mouse Bone Marrow Cells

Femurs from mice were dissected out. The ends of each bone were snipped off, and a 5-ml syringe containing phosphate-buffered saline (PBS) and equipped with a 25-gauge needle (Terumo, Leuven, Belgium, http://www.terumo.com) was used to flush out the bone marrow from both ends of the bone. The resulting diluted bone marrow was washed with L15/10% fetal calf serum (FCS) and pelleted in 1 ml red blood cell lysing buffer (Sigma, St. Louis, http://www.sigmaaldrich.com), pipetted up and down briefly, and incubated for 5 minutes at 37°C. If red cells were still present after washing (red color in the pellet), the lysis procedure was repeated once more.

Hoechst 33342 Staining of Cells

Mouse bone marrow and epithelial cells were resuspended at 106 per ml in L15/10% FCS prewarmed to 37°C, and Hoechst 33342 (Molecular Probes, Leiden, The Netherlands, http://probes.invitrogen.com) was added at a final concentration of 5 μg/ml. This concentration was selected in initial optimization experiments, yielding the most discretely defined SPs for both mammary gland and bone marrow cells under the conditions applied in our laboratory. The curve of Hoechst concentration versus number of SP cells is steep [3], and optimal concentrations vary slightly between labs. Cells were incubated for 90 minutes at 37°C with occasional agitation. After incubation, cells were washed with cold medium and resuspended at 2–3 × 106 per ml, and propidium iodide (Molecular Probes) was added to a final concentration of 2 μg/ml.

Flow Cytometry

Cells were sorted on a FACSVantageSE cell sorter (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). The machine is equipped with two Coherent 90 C-4 argon ion lasers (Coherent, Santa Clara, CA, http://www.coherent.com), one with multiline UV optics (emission of 333.6–333.8 nm), the other with visible optics (tuned to 488 nm). Hoechst 33342 fluorescence was measured at both 424/44 nm and above 670 nm (split by a 610-nm short-pass dichroic mirror), both from UV excitation. Dead cells were excluded by propidium iodide fluorescence measured at 564–606 nm.

Histological Analysis and Immunohistochemistry

Tissues were fixed in 4% phosphate-buffered formalin, embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin and eosin according to standard procedures. For immunohistochemistry, tissues were deparaffinized in xylene and rehydrated. Endogenous peroxidase activity was blocked by using 3% (vol/vol) H2O2 in methanol for 10 minutes. Before staining, paraffin sections were pretreated by heat-induced epitope retrieval. Slides were incubated with 5% normal goat serum/PBS for 30 minutes, and subsequently sections were incubated overnight with a 1:400 dilution of BXP-53 at 4°C. mAb immunoreactivity was detected with the streptavidin-biotin immunoperoxidase (sABC) method by using biotinylated goat anti-rat IgG (1:100) (DakoCytomation, Carpinteria, CA, http://www.dakocytomation.us) as secondary antibody, and diaminobenzidine substrate (DAB) or amino-ethyl-carbazole (AEC) for visualization. After counterstaining with hematoxylin, slides were mounted. For negative control, the primary mAb was omitted.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Generation and Analysis of Bcrp1−/−/Mdr1a/1b−/− Mice

Previously, we have generated mice with a targeted disruption of the ABC transporters Bcrp1 (Abcg2) or of Mdr1a (Abcb1a) and Mdr1b (Abcb1b) [22, 23, 26]. In this study, we generated triple knockout mice lacking Bcrp1 and both Mdr1a and Mdr1b by crossing of the respective knockout lines. Bcrp1−/−/Mdr1a/1b−/− mice were fertile, had a normal life span, and were born at the expected mendelian ratio, indicating that there was no reduced embryonic viability. Standard plasma clinical chemical analysis and histological analysis revealed no abnormalities except that, similar to the Bcrp1−/− mice [22], levels of unconjugated bilirubin were slightly increased. Absence of Bcrp1 and Mdr1a/1b together, therefore, appears to be compatible with normal physiologic functioning of mice.

Bcrp1 and Mdr1a/1b Both Contribute to the Mammary SP Phenotype

The relative contributions of Bcrp1 and Mdr1a/1b to the SP phenotype in the mammary gland were determined by comparing mammary preparations from wild-type, Bcrp1−/− and Bcrp1−/−/Mdr1a/1b−/− mice. For determination of SP, we used pools of mammary glands (fourth inguinal pair) from 4–13 mice per analysis. Wild-type mammary glands had levels of epithelial SP cells that were consistent with our previous observations [7]. Glands from Bcrp1−/− mice still had a clear but significantly reduced SP as compared with wild-type mice (0.04% vs. 0.22%, respectively; p = .024; Figs. 1A, 1B, 1D), suggesting that also in the mammary gland Bcrp1 makes a significant contribution to the SP phenotype. Mammary glands from Bcrp1−/−/Mdr1a/1b−/− animals had no detectable SP cells, indicating that together these transporters are fully responsible for the SP phenotype in the mammary gland (Figs. 1C, 1D).

thumbnail image

Figure Figure 1.. Bcrp1 and Mdr1a/1b are required for the mammary side population (SP) phenotype. (A, B, C): Flow-cytometric SP analysis of mammary epithelial samples from (A) wild-type, (B)Bcrp1−/−, and (C)Bcrp1/Mdr1a/1b−/− mice. Percentages of SP cells within the gated regions of the fluorescence-activated cell sorting (FACS) traces are shown in each panel. (D): Distribution of percentages of SP cells from wild-type, Bcrp1−/−, Mdr1a/1b−/−, and Bcrp1−/−/Mdr1a/1b−/− mice. Numbers of mice used to generate the data were 28 (wild-type), 40 (Bcrp1−/−), 12 (Mdr1a/1b−/−), and 17 (Bcrp1/Mdr1a/1b−/−). Each separate analysis is represented by a single point. The bar indicates the mean.

Download figure to PowerPoint

Loss of Mdr1a/1b Frequently Results in an Increased SP Phenotype in Bone Marrow, Which Is Reversible by Ko143

We next determined the relative contributions of Bcrp1 and Mdr1a/1b to the SP phenotype in bone marrow by comparing wild-type, Bcrp1−/−, Mdr1a/1b−/− and Bcrp1−/−/Mdr1a/1b−/− mice (Fig. 2). A clear but significantly reduced SP was detected in bone marrow of Bcrp1−/− mice as compared with wild-type mice (0.05% ± 0.07% versus 0.18% ± 0.20%, respectively; p < .01). In contrast to the mammary gland, however, bone marrow of Bcrp1−/−/Mdr1a/1b−/− mice still had a detectable SP (0.05% ± 0.08%; p < .01 as compared with wild-type), showing that the SP cells remaining in Bcrp1−/− bone marrow may not exclusively depend on Mdr1a/1b for efflux of Hoechst 33342 (Fig. 2B). Our data showing that Bcrp1 is responsible for the majority of the SP phenotype in hematopoietic cells are consistent with a previous suggestion that Bcrp1 is exclusively responsible for the SP phenotype in hematopoietic cells [3]. Surprisingly, the percentage of SP cells detected in the bone marrow of Mdr1a/b−/− mice was frequently higher than wild-type controls, but this did not reach statistical significance due to high interindividual variation (mean 0.35% ± 0.36% vs. 0.18% ± 0.20%, respectively; p = .17). To further characterize this unexpected increase in SP in Mdr1a/1b−/− mice, we measured SPs in individual mice rather than in pooled samples. It should be noted that in other studies, hematopoietic SP was determined in pools of mice, masking possible interindividual differences [24]. To compensate for the inherent variability in the measurement of percentage of SP, a larger number of samples was analyzed from each genotype (Fig. 2B). In total, between 28 and 46 individual marrows (using either one, or pools of two or four marrows per experiment) were analyzed for the genotypes shown. We found that the SP in the Mdr1a/1b−/− mice was highly variable and that some mice had exceptionally high SPs. The low variability of the SP in the Bcrp1−/−/Mdr1a/1b−/− mice shows that the variation in the Mdr1a/1b−/− mice is not due to the technical procedure. It should be noted that some variation was also observed in the wild-type and Bcrp1−/− mice which might be due to variable expression of the respective transporters. Treatment of the cells with the specific Bcrp1-inhibitor Ko143 [27] strongly reduced the SP phenotype in both wild-type and Mdr1a/1b−/− animals (0.02% ± 0.02%; p < .01 and 0.05% ± 0.06%; p < .01 as compared with untreated wild-type, respectively), suggesting that Bcrp1 is primarily responsible for the SP phenotype in Mdr1a/1b−/− mice. In cases in which a significant SP fraction was detected in the Ko143-treated Mdr1a/1b−/− samples, the corresponding untreated half of the sample was found to have higher-than-normal SP levels, suggesting the involvement of a third transporter compensating for the loss of Mdr1a/1b.

thumbnail image

Figure Figure 2.. Bcrp1, but not Mdr1a/1b, is required for the bone marrow side population (SP) phenotype. (A): Flow-cytometric SP analysis of bone marrow samples from wild-type, Bcrp1−/−, Mdr1a/1b−/−, and Bcrp1/Mdr1a/1b−/− mice. Percentages of SP cells within the gated regions of the fluorescence-activated cell sorting (FACS) traces are shown in each panel. (B): Distribution of percentages of SP cells from wild-type, Bcrp1−/−, Mdr1a/1b−/−, and Bcrp1/Mdr1a/1b−/− mice. Numbers of individual marrows used to generate the data were 46 (wild-type), 34 (Bcrp1−/−), 28 (Mdr1a/1b−/−), 28 (Bcrp1/Mdr1a/1b−/−), six (wild-type/Ko143 treated), and six (Mdr1a/1b−/−/Ko143 treated). Each separate analysis is represented by a single point. The bar indicates the mean.

Download figure to PowerPoint

Immunohistochemical Localization of Bcrp1 in Mammary Gland

Finally, we were interested whether Bcrp1 could be used as a marker to immunohistochemically detect cells in the virgin mammary gland representing the SP. The low background of staining demonstrated using Bcrp1 null tissues, allowed the identification of a small number of positive cells in wild-type virging lands (Figs. 3A–3D, 3F). As might be expected from the low proportion of SP cells in the virgin gland (normally less than 1%), very few positive cells were detected (Figs. 3A–3C). The few cells detected were closely associated with mammary ducts, were often single, and were almost always basal to smooth muscle actin–positive cells (Fig. 3F) as assessed by immunohistochemical staining. No positive cells were ever detected in the luminal cells of mid- and large-sized ducts. As there are major alterations in cellular proliferation and apoptosis throughout the estric cycle, staged glands were stained for Bcrp1, for a marker of proliferation (proliferating cell nuclear antigen [PCNA]), and for apoptosis (terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling [TUNEL] staining). No major changes in the pattern of Bcrp1 expression were noted during the cycle (data not shown). In addition, no obvious relationship was detected between Bcrp1-expressing cells and those staining for PCNA or TUNEL (data not shown).

thumbnail image

Figure Figure 3.. Localization of Bcrp1 in the virgin mouse mammary gland. Immunohistochemical detection of Bcrp1 in virgin mouse mammary glands: (A–C) wild-type, (D, E)Bcrp1−/−. (F): Smooth muscle actin staining in virgin wild-type mammary gland. Arrows indicate positively staining cells. Bars = 100 μM.

Download figure to PowerPoint

Although Bcrp1-expressing cells in the virgin gland are clearly basal to myoepithelial cells, they are intimately associated with mammary ducts and could thus be represented within the primary SP. However, we cannot formally exclude the possibility that cells intimately associated with the epithelium express levels of Bcrp1 below the histological detection limit and are capable of exporting sufficient Hoechst 33342 dye to be represented within the SP cell population. The niche that contains mammary epithelial stem cells remains to be identified, and the Bcrp1-expressing cells (as determined histologically and phenotypically) must thus be considered as candidate stem cells.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Interest in the ABC transporters Bcrp1 and Mdr1a/1b has been high after their implication in the regulation of stem cells [1, 2]. In this report, we observed that mice lacking both transporters (Bcrp1−/−/Mdr1a/1b−/−) develop normally, excluding the possibility that the survival of individual mutant animals was due to functional complementation between Mdr1a/1b and Bcrp1. It remains possible, however, that additional genes functionally compensate for the loss of Bcrp1 and Mdr1 during development.

We analyzed the relative contribution of the ABC transporters Bcrp1 and Mdr1a/1b to the SP phenotype in mammary gland and bone marrow of mice. It was found that both Bcrp1 and Mdr1a/1b contributed to the mammary epithelial SP phenotype. By contrast, a role for Bcrp1 but not Mdr1a/1b was found in the bone marrow. Interestingly, we also observed that in bone marrow, the percentage of SP cells was frequently increased in Mdr1a/1b knockout mice as compared with wild-type mice. A similar increase in SP in Mdr1a/1b bone marrow has also been independently observed by the group of Sorrentino [4]. In that study, however, the increase was tentatively ascribed to different genetic backgrounds of the mice. Because our mice are all of the same genetic background (>99% FVB), this can be excluded. Our results thus suggest that mice with a deficiency in Mdr1a/1b can compensate for this loss by upregulation of Bcrp1. Whether Mdr1a/1b can also be upregulated to compensate for the loss of Bcrp1 is not known. The observation that the SP is not completely ablated on treatment of Mdr1a/1b nulls with a specific and efficient BCRP/Bcrp1 inhibitor suggests that another Hoechst 33342 transporter may be upregulated on loss of Mdr1a/1b. In contrast, the consistent loss of SP in the mammary gland of triple null animals suggests that in this tissue Bcrp1 and Mdr1a/1b are the only transporters capable of effluxing Hoechst 33342.

The SP phenotype has been a useful marker for isolating hematopoietic stem cells and has been suggested to be a stem cell marker [3, 8, 28]. However, there is some controversy over how universal this association may be [29, 30]. Recently, we have demonstrated that Bcrp1 is highly expressed in the apical membrane of mammary alveolar epithelial cells during late pregnancy and lactation, whereas it is not detectable in mammary epithelial cells of virgins [31]. Here we show that in virgin mice, a small number of cells in mammary ducts were found to express Bcrp1. Isolated, positively staining cells were occasionally detected in the basal region of mammary ducts. In most cases, these cells were basal to the smooth muscle actin–positive myoepithelial cells, were not polarized, and had significant levels of cytosolic Bcrp1 staining (Figs. 3A–3C). Because these cells are closely associated with the epithelium, they are likely to be represented within the cell population that is subjected to Hoechst 33342 staining. If all the histochemically positive Bcrp1-expressing cells were identical to the phenotypically identified SP cells, this would imply that stromally located Bcrp1-positive cells were epithelial progenitors, because we previously showed that cultured SP cells differentiated into epithelial cell clones [7]. Previous studies have argued that mammary epithelial progenitors lie within the epithelium adjacent to the basement membrane [32, 33]. A recent report, however, has suggested that stromal adipocytes may transdifferentiate into secretory epithelial cells and vice versa [34]. The data presented here are consistent with the latter observation. Alternatively, significant numbers of cells that do not detectably stain for Bcrp1 might give rise to epithelial progenitors within the SP. This could result from Bcrp1 activity at levels below the immunohistochemical detection limit or from the activity of transporters such as Mdr1a/b that are required for the mammary SP phenotype. Unfortunately, high-quality antibodies are not available for study of the cellular localization of Mdr1a/b in the mammary gland. In previous work, we showed that mammary SP cells were undifferentiated by comparison with non-SP cells, although they differentiated after in vitro culture [7]. In addition, we showed that transplanted mammary SP cells from virgin animals primarily generated lobular epithelial outgrowths [7]. Taken together, these observations suggest that Bcrp1 expression in the virgin mammary gland may mark lobular epithelial progenitors and account for their isolation as SP cells.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Our data show that the relative contributions of the ABC transporters Bcrp1 and Mdr1a/1b to the SP phenotype in different tissues, in this case bone marrow and mammary gland, can vary. As a consequence, the sensitivity of tissues to cytotoxic drugs that are substrates of these ABC transporters will also vary, which in turn might affect the chemotherapeutical treatment of patients. Our data also show that loss of expression of Mdr1a/1b in bone marrow resulted in an increased SP and suggest that this is due to compensatory upregulation of Bcrp1 and possibly a third transporter.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We thank our colleagues for critical reading of the manuscript and Elly Mesman for excellent technical assistance. Matthew Smalley is kindly acknowledged for comments and discussions. This work was supported by the Dutch Cancer Society, Cancer Research UK, and the Breast Cancer Campaign.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References