Proc. Natl. Acad. Sci. USA Vol. 92, pp. 8901-8905, September 1995 Medical Sciences

Modification of staining allows identification of hematopoietic stem cells with preferential short-term or long-term -repopulating ability J. MARK J. M. ZIJLMANS*t, JAN W. M. VISSER*t, KARIN KLEIVERDA§, PHILIP M. KLUIN§, ROEL WILLEMZE*, AND WILLEM E. FIBBE* *Laboratory of Experimental Hematology, Department of Hematology, and §Department of Pathology, University Medical Center, 2300 RC Leiden, The Netherlands; and tLindsley F. Kimball Research Institute, Laboratory of Stem Cell Biology, New York Center, New York, NY 10021 Communicated by J. J. van Rood, University Hospital, Leiden, The Netherlands, June 12, 1995

ABSTRACT We have developed a modified rhodamine In addition to methods exploiting surface marker expres- (]Rho) staining procedure to study uptake and efflux in sion, other techniques are based on functional differences. murine hematopoietic stem cells. Distinct populations of One of these involves the use of rhodamine 123 (Rho), a Rho++ (bright), Rho+ (dull), and Rho- (negative) cells could supravital fluorescent that accumulates intracellularly and be discriminated. Sorted Rho- cells were subjected to a that may bind to mitochondria (5, 6). Cells with smaller or less second Rho staining procedure with the P-glycoprotein block- active mitochondria exhibit reduced uptake of Rho (6). Rho ing agent verapamil (VP). Most cells became Rho positive efflux from the cell is mediated by the transmembrane carrier [Rho-/Rho(VP)+ cells] and some remained Rho negative protein P glycoprotein, encoded by the multidrug-resistance [Rho-/Rho(VP)- cells]. These cell fractions were character- gene (7-9). Hydrophobic drugs used in anticancer therapy such ized by their marrow-repopulating ability in a syngeneic, as vinca alkaloids, anthracyclines, and taxanes are extruded sex-mismatch transplantation model. Short-term repopulat- from the cell by P glycoprotein. P glycoprotein-mediated ing ability was determined by recipient survival for at least 6 transport can be blocked by verapamil (VP). Rho uptake weeks after lethal irradiation and transplantation-i.e., ra- studies have been used to purify murine HSCs (10-13). The dioprotection. Long-term repopulating ability at 6 months Rho' (dull) cells representing primitive stem cells exhibited after transplantation was measured by in situ MRA, while Rho++ (bright) cells representing more mature hybridization with a Y-chromosome-specific probe, by graft progenitor cells had the ability to form colonies or in function and recipient survival. Marrow-repopulating cells vitro colonies. were mainly present in the small Rho- cell fraction. Trans- Recent developments in HSC characterization have dem- plantation of 30 Rho- cells resulted in 50%o radioprotection onstrated heterogeneity within the cell population having and >80% donor repopulation in marrow, spleen, and thymus MRA. By using counterflow centrifugation, cells with short- 6 months after transplantation. Cotransplantation of cells term repopulating ability (STRA) can be distinguished from from both fractions in individual mice directly showed that cells with long-term repopulating ability (LTRA) (14). Recip- within this Rho- cell fraction, the Rho-/Rho(VP)+ cells ients transplanted with cells mediating LTRA only did not exhibited mainly short-term and the Rho-/Rho(VP)- cells survive lethal irradiation. These cells, however, could sustain exhibited mainly long-term repopulating ability. Our results long-term hematopoiesis ifcotransplanted with cells exhibiting indicate that hematopoietic stem cells have relatively high STRA. In the present study, we have developed a modified P-glycoprotein expression and that the cells responsible for Rho staining method selecting for Rho efflux capacity that long-term repopulating ability can be separated from cells allows identification of distinct populations of Rho++, Rho+, exhibiting short-term repopulating ability, probably by a and Rho- cells. The Rho- cells were further subdivided by a reduced mitochondrial Rho-binding capacity. second Rho staining with the use of VP in Rho-/Rho(VP)+ and Rho-/Rho(VP)- cells. Functional characterization of The lifelong production of mature lymphoid and myeloid these stem cell populations in a syngeneic mouse transplan- blood cells is derived from a small population of primitive tation model demonstrates highly purified cell fractions with hematopoietic stem cells (HSCs). The frequency of these cells preferential STRA or LTRA. in murine bone marrow is estimated to be 1/104-105 (1). HSCs are functionally characterized by their ability to repopulate the MATERIALS AND METHODS bone marrow including the HSC compartment of irradiated recipients (marrow-repopulating ability; MRA). For many Mice. BALB/c mice ranging from 8 to 12 weeks of age were years, purification of HSCs has been pursued, not only for purchased from Broekman (Someren, The Netherlands). Mice biological studies- e.g., susceptibility to hematopoietic growth were housed in laminar flow cabins for the duration of the factors-but also for clinical purposes-i.e., gene therapy, experiment and were fed commercially available rodent chow. transplantation without contaminating tumor cells, or trans- From 1 week before until 6 weeks after total body irradiation, plantation with in vitro amplified stem cells. Methods to purify the drinking fluid was supplemented with ciprofloxacin (85 HSCs rely on positive selection-i.e., binding of the lectin mg/liter) and polymyxin B (70 mg/liter). The experimental wheat germ agglutinin (WGA) (2,3) or expression of the Sca-1 protocol was approved by the institutional ethical committee antigen (4)-combined with depletion techniques based on on animal experiments. the absence of myeloid or lymphoid lineage markers. Using these techniques a 500- to 1000-fold enrichment of HSCs can Abbreviations: HSC, hematopoietic stem cell; Rho, rhodamine 123; be obtained (3, 4). VP, verapamil; MRA, marrow-repopulating ability; STRA, short-term repopulating ability; LTRA, long-term repopulating ability; WGA, wheat germ agglutinin; FISH, fluorescence in situ hybridization. The publication costs of this article were defrayed in part by page charge tTo whom reprint requests should be addressed at: Department of payment. This article must therefore be hereby marked "advertisement" in Hematology, University Medical Center Leiden, Building-1 C2R, accordance with 18 U.S.C. §1734 solely to indicate this fact. P.O. Box 9600, 2300 RC Leiden, The Netherlands. 8901 Downloaded by guest on September 27, 2021 8902 Medical Sciences: Zijlmans et al. Proc. Nati. Acad. Sci. USA 92 (1995) Stem Cell Purification. Mice were sacrificed by CO2 as- presence of donor cells in bone marrow (myeloid cells), thymus phyxia. Both femora were flushed using a syringe with a fine (T lymphocytes), and spleen (B lymphocytes). needle. Flushing, incubation, and washing procedures were In some experiments, a competitive repopulation assay was done in RPMI 1640 medium supplemented with 2% fetal done by transplanting female recipients with a combination of bovine serum, penicillin (500 ,ug/ml), and streptomycin (250 male and female cells from different stem cell fractions. In jig/ml) (GIBCO). Bone marrow cells were separated by density- these experiments, peripheral blood samples obtained by tail gradient centrifugation (Ficoll/Isopaque; 1.077 g/cm3). Low bleeding at intervals ranging from 3 to 16 weeks after trans- density cells were labeled with the monoclonal antibody 15-1.1 plantation were assayed for the presence of male donor cells. (rat IgG2b) binding to myelomonocytic cells (15). Cells were washed once and labeled with isothiocyanate- RESULTS conjugated WGA (0.2 ,ug/ml; Vector Laboratories) and phy- coerythrin-conjugated goat anti-rat IgG (Caltag, South San Stem Cell Purification. Low density bone marrow cells Francisco, CA). WGA+/15-1.1- cells (±5%) were sorted (8-15%) were labeled with fluorescein isothiocyanate- using a FACStar flow cytometer (Becton Dickinson) equipped conjugated WGA, 15-1.1, and phycoerythrin-conjugated goat with a 5-W argon tuned at 488 nm (0.2 W). Sorting gates anti-rat IgG. The WGA+/15-1.1- cell population represented were set such that the WGAdim cells were also included. 4-6% of the cells. Rho staining (100 ng/ml, 20 min, 37°C, WGA+/15-1.1- sorted cells were stained with Rho (Molecular followed by washing twice at room temperature) of WGA+/ Probes) at a concentration of 0.1 ,gg/ml for 20 min at 37°C. 15-1.1- sorted cells allowed identification of Rho++ and Rho+ Cells were washed twice at room temperature, resuspended, cells (Fig. 1A). After incubation in Rho-free medium (37°C), and incubated in medium without Rho for 20 min at 37°C. To an additional Rho- population could be recognized. This improve the distinction between Rho- and Rho' cells, we Rho- cell population comprised 5-8% after 20 min (Fig. 1B) modified the procedure in that an excitation and emission and 30% after 60 min of incubation in Rho-free medium. For wavelength of 514 and 580 nm was used rather than the transplantation experiments, the Rho++, Rho+, and Rho- standard excitation and emission wavelengths of 488 and 530 cells were sorted after 20 min of incubation in Rho-free nm. An unstained control sample was used to set the sorting medium. In some experiments, the Rho- cell population was gate for Rho- cells (99% of unstained cells were within this sorted and restained with Rho. Now, the incubation in Rho- gate). Distinct populations of Rho-, Rho+, and Rho++ cells free medium (20 min, 37°C) was done in the presence of VP, were sorted. In some experiments, sorted Rho- cells (5-8%) with most cells (85-92%) becoming Rho(VP)+ and a minority were stained with Rho again (0.1 ,ug/ml; 20 min; 37°C), washed (8-15%) remaining Rho(VP)-. The Rho-/Rho(VP)+ cells twice at room temperature, and incubated in medium without represented 2-4/104 and the Rho-/Rho(VP)- cells repre- Rho, now in the presence of VP (10 ,uM) for 20 min at 37°C. sented 2-6/105 bone marrow cells. The distinct identification Distinct populations of Rho(VP)+ and Rho(VP)- cells were of the Rho- and Rho-/Rho(VP)- cells was possible only sorted. Cells were used for transplantation within 18 hr after when excitation and emission wavelengths of 514 and 580 nm sacrificing the donor mice. were used (Fig. 2 A and B) instead of the standard excitation Irradiation. Recipient mice were given total body irradia- and emission wavelengths of 488 and 530 nm (Fig. 2 C and D). tion of 8.5 Gy divided in two equal parts in posterior-anterior Determination of Total Body Irradiation Dose. To deter- and anterior-posterior positions at a dose rate of 4 Gy/min mine the optimal dose of total body irradiation, mice were with a Philips SL 75-5/6-mV linear accelerator (Philips Med- irradiated at doses of 8.1, 8.5, and 8.9 Gy and the survival time ical Systems, Best, The Netherlands). was determined. After a dose of 8.1 Gy, all mice died after an Peripheral Counts. Blood was obtained by tail interval varying from 13 to 29 days. At a dose of8.9 Gy, all mice bleeding or cardiac puncture. White blood cells, red blood died between 6 and 10 days after irradiation, probably because cells, and were counted on a Sysmex F800 counter of irradiation toxicity. At 8.5 Gy, all mice died but no early (TOA Medical Electronics, Kobe, Japan). deaths and no substantial survival beyond day 16 were ob- Fluorescence in Situ Hybridization (FISH). The presence served: 24 of 26 mice (92%) died between days 11 and 16. of male donor cells was demonstrated in female Based on these results, a dose of 8.5 Gy was used in subsequent recipient experiments to allow reliable measurement of early bone mice by FISH using the murine Y-chromosome-specific marrow repopulating ability before day 16 after transplanta- probe M34 (16). Cells were harvested from bone marrow, tion. Also, because the irradiation dose of 8.5 Gy is supralethal, thymus, spleen, or peripheral blood. Cell preparations were autologous recovery of bone marrow cells was expected to be made and hybridization was performed as described (17, 18). minimal. Blinded samples were scored by two independent observers Short-Term Survival Is Mediated Mainly by Rho- and with a Leitz Diaplan microscope, each examining at least 200 Rho-/Rho(VP)+ Cells. To test the presence of HSCs in the nuclei. Results were expressed as mean percentage of hy- purified fractions, decreasing cell numbers were transplanted bridization signal positive nuclei in each sample. In nine into lethally irradiated recipient mice and the percentage of control samples from female and male mice 0.5% ± 0.3% and 98.0% ± 1.5% positive nuclei, respectively, were found. A lB Experimental Protocol. HSCs were characterized in a bone a,0 D marrow repopulating assay after transplantation into lethally E irradiated (8.5 Gy) recipients. Female recipients were trans- z .,ill planted with male donor-derived cells from each of eight ..,''41I'l 111. fractions at decreasing cell numbers. In each experiment, 0 L groups of 8-10 mice for each cell dose and cell fraction were 1.Lx transplanted (40-120 mice per experiment, nine experiments). Bone marrow-repopulating ability was distinguished in Rhodamine Fluorescence Intensity (Log) STRA and LTRA. STRA was measured either by survival of at 6 FIG. 1. Rho fluorescence histograms showing Rho efflux capacity mice for least weeks after lethal irradiation (radioprotec- of sorted WGA+/15-1.1- cells. Gray line represents an unstained tion) or by the presence of donor cells in the recipients 3-6 control sample. (A) Immediately after Rho incubation (0.1 ,ug/ml, 20 weeks after transplantation. LTRA was measured 6 months min, 370C) and washing, Rho++ and Rho+ cells can be discriminated. after transplantation by survival; by enumerating peripheral (B) After an additional incubation in Rho-free medium (20 min, 37°C), blood erythrocytes, leukocytes, and platelets; and by the a separate Rho- cell population is present (8%). Downloaded by guest on September 27, 2021 Medical Sciences: Zijlmans et al. Proc. Natl. Acad. Sci. USA 92 (1995) 8903 A B the percentage of Y-chromosome-positive cells in bone mar- row, spleen, and thymus of mice 6 months after transplantation . ,117! was determined by FISH with a Y-chromosome-specific probe. 514 nm More than 80% donor cells were found in bone marrow, spleen, and thymus in 69 of 75 (92%) mice transplanted with IA't 30, 100, or 300 cells from the Rho-, Rho-/Rho(VP)+, and Rho-/Rho(VP)- cell fractions (Table 1). Immunophenotyp- ing with T-lymphocyte- and B-lymphocyte-specific monoclo- c D nal antibodies in three mice 6 months after lethal irradiation and transplantation showed 95-99% T cells in the thymus and li. .j 40-52% B cells in the spleen cell preparations. Although the

488 nm .li "i i:) spleen is not a purified B-lymphocyte population, the >80% I if donor repopulation implicates the presence of donor-derived ,i B cells. We did not observe lineage-restricted repopulation '1.''9 patterns (myeloid vs. lymphoid). It can thus be concluded that Rhodamine Fluorescence Intensity (Log) all three stem cell populations have multilineage LTRA. Increased Late Mortality After Transplantation of FIG. 2. Rho fluorescence histograms representing two sequential Rho-/Rho(VP)+ Cells. In some groups of recipient animals, Rho staining procedures, the second with the use of VP. Gray line a substantial mortality was observed between 6 weeks and 6 represents an unstained control sample. Excitation wavelengths of 514 months after lethal irradiation and transplantation. This late nm (A and B) and 488 nm (C and D) were used. (A and C) Rho mortality was determined by the number of cells in the graft fluorescence histogram of sorted WGA+/15-1.1- cells stained with and by the cell fraction used for transplantation. The percent- Rho followed by incubation in Rho-free medium (20 min, 37°C). A age of mice between 6 weeks and 6 months after distinct Rho- cell population is observed only at 514 nm excitation. (B dying and D) Rho fluorescence histogram of sorted WGA+/15-1.1-/Rho- transplantation of 300, 100, 30, and 10 Rho- cells was 16% cells stained with Rho again followed by incubation in Rho-free (6/37), 25% (9/35), 35% (9/26), and 60% (3/5), respectively. medium in the presence of VP (10 ,uM, 20 min, 37°C). Distinct Survival curves for mice transplanted with a fixed cell number Rho(VP)+ and Rho(VP)- cell populations are observed only at 514 (300 cells) from the Rho++, Rho+, Rho-, Rho-/Rho(VP)+, nm excitation. and Rho-/Rho(VP)- fractions are shown in Fig. 4. Trans- plantation of Rho++ cells did not result in survival beyond day mice surviving at least 6 weeks (radioprotection) was deter- 30. About 1/3 of the mice transplanted with Rho+ cells lived mined (Fig. 3). The radioprotective capacity was highly en- beyond day 30, but most of them died before day 100. Some riched in the Rho- fraction of WGA+/15-1.1- cells and in the late mortality was observed after transplantation of Rho- cells Rho-/Rho(VP)+ subfraction. The cell dose resulting in 50% (survival, 88% and 74% on days 100 and 180, respectively) but survival was calculated to be 104 for unmodified bone marrow it was more pronounced after transplantation of Rho-/ cells, 3 x 103 for light density cells, 200 for WGA+/15-1.1- cells, Rho(VP)+ cells (survival, 89% and 56% on days 100 and 180, 3 x 104 for Rho++ cells, 800 for Rho+ cells, 30 for Rho- cells, respectively). Although short-term survival was only 65% after 25 for Rho-/Rho(VP)+ cells, and 200 for Rho-/Rho(VP)- transplantation of Rho-/Rho(VP)- cells (see also Fig. 3), we cells. Short-term survival was obtained with cells from all observed no substantial late mortality. fractions, thus obviating the need to perform transplantations Improved Long-Term Bone Marrow Function After Trans- with helper cells to provide short-term survival. plantation of Rho-/Rho(VP)- Cells. To test the hypothesis Percentage Donor Repopulation 6 Months After Transplan- that the late mortality of mice transplanted with Rho-/ tation. To assess the LTRA of transplanted male donor cells, Rho(VP)+ cells was caused by late graft failure, we studied bone marrow function 6 months after lethal irradiation and

Survival (%) transplantation. Peripheral blood counts of red blood cells,

100- -T white blood cells, and platelets were determined in recipient mice 6 months after transplantation of 30, 100, and 300 cells from the Rho-/Rho(VP)+ and Rho-/Rho(VP)- cell frac- tions (Table 2). counts were almost normal (9.9 x 1012 cells per liter) after transplantation of only 30 Rho-/ Rho(VP)- cells, while they were decreased after transplanta- tion of Rho-/Rho(VP)+ cells (6.7 x 1012 cells per liter after transplantation of 30 cells and 8.6 x 1012 cells per liter after Table 1. Percentage male donor cells in bone marrow, spleen, and thymus cell preparations in female recipient mice 6 months 1000 10000 100000 after transplantation Number of Cells Transplanted Transplantation

bone marrow Iight density WGA+ /15-1 .1- Fraction No. of cells Marrow Spleen Thymus

O Rho + + -7-Rho + -*Rho- Rho- 30 72 ± 33 76 ± 25 64 ± 28 100 91± 6 88± 4 96± 2 .&- Rho-/Rho(VP) + Rho-/Rho(VP)- 300 92± 5 Rho-/Rho(VP)+ 30 90 ± 3 87 ± 2 93 ± 6 FIG. 3. Short-term survival of lethally irradiated (8.5 Gy) mice 100 79 28 81 ± 14 81 ±32 after transplantation of bone marrow-derived cell fractions. Each data 300 89± 7 point represents survival percentage of 10-40 mice. Survival data of Rho-/Rho(VP)- 30 92 ± 2 88 ± 4 96 ± 3 mice transplanted with Rho-/Rho(VP)+ or Rho-/Rho(VP)- cells 100 93± 3 86± 3 89± 9 are expressed as means ± SD of four different experiments with 8-10 300 95 ± 2 mice transplanted per cell fraction and per cell dose in each experi- ment. Results are expressed as means + SD of 7-11 mice per group. Downloaded by guest on September 27, 2021 8904 Medical Sciences: Zijlmans et al. Proc. Natl. Acad. Sci. USA 92 (1995)

Survival (%) Percentage Male Peripheral Blood Leukocytes IUU x 80

60 --

40:

40 _ ...... 20 -

v V C . 20 . .T 0 20 40 60 80 100 120 140 160 180 Days after Transplantation I~~~~~ 0 20 40 60 80 100 120 Control I Rho+ + 7 Rho± Days after Transplantation * Rho- A Rho-/Rho(VPI + X Rho-/Rho(VP)- FIG. 5. Peripheral blood chimerism of mice at various time in- FIG. 4. Cumulative survival of lethally irradiated (8.5 Gy) mice tervals after cotransplantation of Rho-/Rho(VP)+ and Rho-/ Rho(VP)- cells. Female recipient mice were transplanted with 200 after transplantation of 300 cells from the Rho++ (n = 16), Rho+ (n Rho-/Rho(VP)+ cells derived from a female donor and 200 Rho-/ = 22), Rho- (n = 42), Rho-/Rho(VP)+ (n = 28), or Rho-/ Rho(VP)- cells from a male donor (group A, 10 mice; +) or with 200 Rho(VP)- (n = 20) fraction. Late mortality was observed mainly in mice transplanted with Rho-/Rho(VP)+ cells. Rho-/Rho(VP)+ cells derived from a male donor and 200 Rho-/ Rho(VP)- cells from a female donor (group B, 10 mice; X). Data are expressed as percentage male cells in individual mice. Lines connect transplantation of 100 cells). counts were similarly the mean of 10 mice per group at 19, 40, and 110 days after decreased in mice transplanted with Rho-/Rho(VP)+ cells. transplantation. Differences in counts were less pronounced. STRA Is Derived from Rho-/Rho(VP)+ and LTRA from Transplantation of only 30 Rho- cells in lethally irradiated Rho-/Rho(VP)- Cells. To directly compare the relative recipients resulted in 50% survival at 6 weeks and >80% donor STRA and LTRA of Rho-/Rho(VP)- cells and Rho-/ repopulation in bone marrow, spleen, and thymus at 6 months Rho(VP)+ cells, cotransplantation experiments were per- after transplantation. In this highly purified cell fraction, formed with a mixture of both cell populations. For these STRA and LTRA could be separated by modification of Rho experiments, equal numbers of Rho-/Rho(VP)+ cells from a staining with the use of VP. The resulting Rho-/Rho(VP)- female donor and Rho-/Rho(VP)- cells from a male donor cells are relatively depleted of STRA but enriched for LTRA, (group A) or Rho-/Rho(VP)+ cells from a male donor and while the Rho-/Rho(VP)+ cells are relatively enriched for Rho-/Rho(VP)- cells from a female donor (group B) were STRA and depleted for LTRA. cotransplanted in lethally irradiated female recipients. After Previously, HSCs were purified with the use of WGA, 15-1.1, transplantation, male peripheral blood cells in the group A and Rho (15). MRA was found in the Rho' cell population, mice would be derived from the Rho-/Rho(VP)- stem cells representing ±50% of all WGA+/15-1.1- cells (6). We found and in the group B mice from the Rho-/Rho(VP)+ stem cells. that MRA is mainly present in the small (5-8%) cell popula- At intervals ranging from 3 to 16 weeks, peripheral blood tion becoming Rho- upon incubation in Rho-free medium. leukocytes were examined by FISH for the presence of male This implicates a 6-10 times further purification of marrow- cells (Fig. 5). A decrease in the percentage of male cells in repopulating cells. Rho efflux capacity has been reported to group B mice (from 75% to 18%) and an increase in group correlate with expression of multidrug-resistance mRNA or its A mice (from 28% to 49%) were observed, directly showing product the P-glycoprotein transmembrane carrier protein in that Rho-/Rho(VP)+ cells had better STRA and Rho-/ tumor cell lines or subtypes of peripheral blood leukocytes (19, Rho(VP)- cells had better LTRA. 20). P-glycoprotein expression has also been found in human hematopoietic progenitor cells (9). Therefore, P-glycoprotein DISCUSSION expression is probably higher in cells becoming Rho- upon incubation in Rho-free medium. After a second exposure of We demonstrate that MRA is mainly present in the small the WGA+/15-1.1-/Rho- cells to Rho with the use of VP, Rho- population of WGA+/15-1.1- bone marrow cells, func- most cells became Rho(VP)+ due to the blocking of P- tionally characterized by the strongest Rho efflux capacity. glycoprotein-mediated efflux, but some cells remained Table 2. Peripheral blood cell counts in mice 6 months after transplantation of cells from Rho- subfractions Transplantation Fraction RBC per liter WBC per liter Platelets per of Rho- No. of cells X 10-12 x 10-9 liter x 10-9 Rho(VP)+ 30 6.7 ± 1.6 8.6 ± 0.7 405 ± 302 100 8.6 ± 0.4 11.0 ±4.3 642 ± 197 300 9.2 ± 1.3 10.9 ± 4.6 551 ± 254 Rho(VP)- 30 9.9 ± 0.4 8.1 ± 0.8 660 ± 237 100 9.5 ± 0.3 15.4 ± 2.6 905 ± 123 300 11.3 ± 1.7 13.4 ± 3.9 1050 ± 323 RBC, red blood cells; WBC, white blood cells. Results are expressed as means ± SD of 4-8 mice per group. Downloaded by guest on September 27, 2021 Medical Sciences: Zijlmans et al. Proc. Natl. Acad. Sci. USA 92 (1995) 8905 Rho(VP)-. It should be noted that efflux is still possible during LTRA and Rho-/Rho(VP)+ cells with predominant STRA. the Rho incubation, since VP was added only after Rho Both subfractions, however, can be considered as HSCs since incubation. Virtually all cells (93-97%) became Rho(VP)+ if transplantation of either fraction alone resulted in long-term VP was also present during Rho incubation (data not shown), chimerism in myeloid cells, T lymphocytes, and B lymphocytes. indicating that probably all cells have the ability to uptake Rho. It is unclear whether these purified stem cell fractions repre- LTRA was found mainly in the Rho-/Rho(VP)- cell popu- sent homogeneous populations of cells with both LTRA and lation, indicating that the most primitive HSCs have either a STRA or mixed populations of cells with only LTRA or reduced Rho-uptake capacity or the highest P-glycoprotein STRA. Transplantation of single cells will be required to activity. resolve this issue. Fluorescence intensity differences between Rho-, Rho', In conclusion, HSCs with both STRA and LTRA can be and Rho++ cells and between Rho-/Rho(VP)- and Rho-/ purified by selection for maximal Rho efflux capacity, likely Rho(VP)+ cells were more pronounced with 514-nm excita- reflecting a higher level of P-glycoprotein expression. Within tion than with 488-nm light. This may be explained by a this cell fraction, using Rho and a P-glycoprotein blocking difference in the excitation spectra of Rho in cells that bind the agent, cells with mainly STRA or LTRA could be discrimi- dye intracellularly to mitochondria as opposed to cells retain- nated. The Rho-/Rho(VP)- cells exhibiting preferential ing it in unbound form in the cytoplasm. The spectrum of the LTRA may be the most suitable candidate stem cells for bound dye is shifted to the red by >10 nm (5, 21). Excitation somatic gene transfer. by 514-nm light is more efficient for the bound form than for the free form, whereas this difference is less significant or even We thank Maarten van de Keur from the Department of Hema- reversed at 488 nm. Rho- cells and Rho-/Rho(VP)- cells tology for assistance in fluorescence-activated cell sorting procedures have a fluorescence intensity comparable to an unstained and Arie Boon and the Department of Radiotherapy of the University control sample when excited at 514 nm but a positive fluores- Medical Center Leiden for assistance in irradiating the animals. cence intensity when excited at 488 nm, apparently because they retain some free dye. We observed earlier that the 1. Visser, J. W. M. & De Vries, P. (1988) Blood Cells 14, 369-384. mitochondria in the more primitive Rho+ cells are smaller than 2. Visser, J. W. M., Bauman, J. G. J., Mulder, A. H., Eliason, J. F. those in the more mature Rho++ cells (6). Our present results & De Leeuw, A. M. (1984) J. Exp. Med. 59, 1576-1590. excitation also indicate that the Rho- stem cells 3. Ploemacher, R. E., Van Der Loo, J. C. M., Van Beurden, C. A. J. using 514-nm & Baert, M. R. M. (1993) Leukemia 7, 120-130. and especially the more primitive Rho-/Rho(VP)- stem cells 4. Spangrude, G. J., Heimfeld, S. & Weissman, I. L. (1988) Science show reduced intracellular, mitochondrial Rho binding. 241, 58-62. In most studies LTRA is measured only by the percentage 5. Darzynkiewicz, Z., Traganos, F., Staiano-Coico, L., Kapuscinsky, donor cells in the recipient mice (3, 4, 9, 13). We found almost J. & Melamed, M. R. (1982) Cancer Res. 42, 799-806. equally high percentages of donor repopulation 6 months after 6. Visser, J. W. M., De Vries, P., Hogeweg-Platenburg, M. G. C., transplantation of Rho-/Rho(VP)+ cells and after transplan- Bayer, J., Schoeters, G., Van Den Heuvel, R. & Mulder, D. H. tation of Rho-/Rho(VP)- cells. Still, a reduced LTRA of (1991) Semin. Hematol. 28, 117-125. Rho-/Rho(VP)+ cells was suggested since mice at 6 months 7. Lampidis, T. J., Munck, J. N., Krishan, A. & Tapiero, H. (1985) after transplantation of Rho-/Rho(VP)+ cells had lower Cancer Res. 45, 2626-2631. peripheral blood cell counts and increased late mortality in 8. Kessel, D. (1989) Cancer Commun. 1, 145-149. with with 9. Chaudhary, P. M. & Roninson, I. B. (1991) Cell 66, 85-94. comparison mice transplanted Rho-/Rho(VP)- 10. Bertoncello, I., Hodgson, G. S., Bradley, T. R., Hunter, S. D. & cells. In previous studies, increased mortality between 4 and 12 Barber, L. (1985) Exp. Hematol. 13, 999-1006. months after transplantation was observed in mice trans- 11. Mulder, A. H. & Visser, J. W. M. (1987) Exp. Hematol. 15, planted with limiting cell numbers, suggesting that late mor- 99-104. tality was due to graft failure (22). Our data also implicate that 12. Ploemacher, R. E. & Brons, N. H. C. (1989) Exp. Hematol. 17, autologous stem cell recovery was minimal. Probably, this is 263-266. related to the relatively high dose of irradiation (8.5 Gy) that 13. Spangrude, G. R. & Johnson, G. R. (1990) Proc. Natl. Acad. Sci. is beyond the threshold dose resulting in 100% mortality (8.1 USA 87, 7433-7437. Gy). Our data indicate the importance of measuring LTRA 14. Jones, R. J., Wagner, J. E., Celano, P., Zicha, M. S. & Sharkis, not only by percentage donor repopulation but also by graft S. J. (1990) Nature (London) 347, 188-189. function and survival at 6 months after transplantation. Trans- 15. Visser, J. W. M. & De Vries, P. (1990) Methods Cell Biol. 33, involving a combination of Rho-/ 451-468. plantation experiments 16. Singh, L., Winking, H., Jones, K. W. & Gropp, A. (1988) Mol. Rho(VP)- and Rho-/Rho(VP)+ stem cells directly indicated Gen. Genet. 212, 440-449. a 3-fold better STRA of Rho-/Rho(VP)+ stem cells and a 17. Kibbelaar, R. E., Van Kamp, H., Dreef, E. J., Wessels, J. W., 3-fold better LTRA of Rho-/Rho(VP)- stem cells. Beverstock, G. C., Raap, A. K., Fibbe, W. E., Den Ottolander, Separation of STRA from LTRA by bone marrow cell G. J. & Kluin, P. M. (1991) Cytogenet. Cell Genet. 56, 132-136. fractionation with counterflow centrifugation has been de- 18. Fibbe, W. E., Hamilton, M. S., Laterveer, L. L., Kibbelaar, R. E., scribed (14). Clear distinction between STRA and LTRA in Falkenburg, J. H. F., Visser, J. W. M. & Willemze, R. (1992) J. highly purified stem cell populations has not yet been de- Immunol. 148, 417-421. scribed. Within the highly purified Sca+/Thyl.1low stem cell 19. Drach, D., Zhao, S., Drach, J., Mahadevia, R., Gattringer, C., population, cells with STRA only could be separated from cells Huber, H. & Andreef, M. (1992) Blood 80, 2729-2734. with both STRA and LTRA by coexpression of the Mac 20. Klimecki, W., Futscher, B. W., Grogan, T. M. & Dalton, W. S. 25). (1994) Blood 83, 2451-2458. antigen (23) or by a medium/high Rho retention (24, 21. Emaus, R. K., Grunwald, R. & Lemasters, J. L. (1986) Biochim. These results raised the issue on the existence of stem cells with Biophys. Acta 850, 436-448. LTRA only. In the present study, although with a different 22. Mauch, P. & Hellman, S. (1989) Blood 74, 872-875. preenrichment procedure (WGA+/15-1.1-), we also have 23. Morrison, S. J. & Weissman, I. L. (1994) Immunity 1, 661-673. purified a Rho- cell population having both STRA and 24. Li, C. L. & Johnson, G. R. (1992) J. Exp. Med. 175, 1443-1447. LTRA. After an additional purification procedure, we were 25. Spangrude, G. J., Brooks, D. M. & Tumas, D. B. (1995) Blood 85, able to discriminate Rho-/Rho(VP)- cells with predominant 1006-1016. Downloaded by guest on September 27, 2021