Lifelong hematopoiesis in both reconstituted and sublethally irradiated mice is provided by multiple sequentially recruited stem cells

Experimental Hematology 29 (2001) 786–794

Lifelong hematopoiesis in both reconstituted and sublethally irradiated mice is provided by multiple sequentially recruited stem cells Nina J. Drize, Yulia V. Olshanskaya, Ludmila P. Gerasimova, Tatiana E. Manakova, Nina L. Samoylina, Tamara V. Todria, and Joseph L. Chertkov Hematological Scientific Center, Moscow, Russia (Received 10 May 2000; revised 2 February 2001; accepted 19 February 2001)

Objective. To evaluate the dynamics of stem cell production to hematopoiesis, the number of active stem cell clones and the lifespan of individual clones were studied. Materials and Methods. The clonal contribution of primitive hematopoietic stem cells (HSC) responsible for long-term hematopoiesis was determined using two approaches. In one model, irradiated female mice were reconstituted with retrovirally marked male hematopoietic cells. In the second model, mice were irradiated sublethally without hematopoietic cell transplantation. In both models, bone marrow cells were serially sampled from the same mouse throughout a 12- to 20-month period and injected into irradiated recipients for analysis of day 10 colony-forming unit-spleen (CFU-S). The donor origin of CFU-S was determined by the presence of retrovirally marked cells or cells with chromosomal aberrations. Results. The results of the two essentially different models show that 1) hematopoiesis is mainly the product of small clones of hematopoietic cells; 2) the lifespan of the majority of clones is only 1 to 2 months; 3) the clones usually function locally; and 4) the vast majority of the clones replace one another sequentially. Primitive HSCs capable of producing long-lived clones (about 10% among all clones), which exist during the entire life of a mouse, were detected by the radiation-marker technique only. Conclusion. Multiple short-living clones (at least on the level of CFU-S production) comprise the vast majority of the active stem cells in transplanted recipients or after endogenous recovery from sublethal irradiation. © 2001 International Society for Experimental Hematology. Published by Elsevier Science Inc.

The hematopoietic system throughout a lifetime produces a huge number of blood cells. It has been assumed that intensive cell production is provided by hematopoietic stem cells (HSC), which have the unique ability for self-maintenance. However, the attempts to demonstrate self-renewal of HSCs by repeated passages through irradiated recipients showed that hematopoiesis was exhausted after only 3 to 4 passages [1,2]. Nevertheless, that one single cell clone can function throughout the lifetime and can restore hematopoiesis after 1 or 2 passages may serve as indirect evidence of self-renewal [3,4]. Alternatively, it is possible that these data characterize the lifespan of an individual HSC, rather than its self-renewal [5]. Furthermore, there is indirect evidence of a restricted proliferative (and/or self-renewal) potential of HSCs. For exOffprint requests to: Joseph Chertkov, M.D., Ph.D., Laboratory for Physiology of Hematopoiesis, Hematological Scientific Center, Novozykovsky proesd, 4A, 125167 Moscow, Russia; E-mail: [email protected]

ample, just one passage essentially reduces the competitive repopulation ability of HSC compared with normal hematopoietic cells [2,6]. Quiescent fetal HSCs after several divisions in culture can reenter G0; such HSCs have reduced potential and are not capable of competing with those HSCs that never left G0 [7,8]. If HSCs have restricted self-renewal/ proliferative potential, one must assume that hematopoiesis is supported by multiple clones that sequentially replace each other [9]. The fate of the individual colony-forming unit-spleen (CFU-S) clones has been studied in mice reconstituted with retrovirally marked hematopoietic cells [10,11]. These studies showed that, throughout the lifetime of reconstituted mice, several dozen short-lived, locally functioning CFU-S sequentially succeeded one another. However, these studies had a number of limitations. First, the examination of clones was carried out only every 3 to 4 months. From the subsequent analysis, it became clear that the life expectancy of an

0301-472X/01 $–see front matter. Copyright © 2001 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(01)0 0 6 3 4 - 8

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

individual clone does not exceed 3 months because the overwhelming majority of clones were revealed only once. However, the actual lifespan of clones could be even shorter. Such information is essential for estimation of the total number of HSCs in the bone marrow (BM) and the number of HSCs used during a lifetime. That there were no data showing the relationship between the number of functioning clones of HSCs and the total number of transplanted cells limited the method and the determination of the actual number of HSCs. The third drawback of such investigations was that the observations were conducted on reconstituted animals, which makes it impossible to distinguish between the inability of HSCs to selfrenew and the effects of the transplantation procedure [12,13]. Finally, the major drawback of the studies was the use of a retrovirally transferred gene marker, which targets dividing cells only [14]. Therefore, HSCs that remain in G0 during the gene transfer procedure would not be detected. In the present study, we attempted to eliminate these drawbacks. The life expectancy of individual retrovirally marked clones was investigated by monthly testing, and a shorter clonal lifespan was found. Using different doses of transplanted cells, the relationship between the number of clones and the total number of injected hematopoietic cells was revealed. On the other hand, we applied an essentially different method of labeling HSCs, using radiation markers. For sublethally irradiated (5.5 Gy) mice, individual chromosomal markers were observed in a large proportion of HSCs, which were marked independent of cell cycle. This method allowed us to bypass all of the limitations associated with retroviral transduction and transplantation and to follow the fate of clones produced by HSCs, including those that were in G0 at the time of irradiation.

Materials and methods Mice For reconstitution experiments, 12- to 25-week-old male and female CBF1 (C57Bl/6xCBA) F1 mice were used as donors and recipients, respectively. Recipient mice were exposed to 1,000 cGy 137 Cs irradiation (the dose rate 18 cGy/min, IPK irradiator). The dose was divided into two equal exposures given 3 hours apart. Male donors of BM cells were injected either with a single IV dose of 5-fluorouracil (5-FU; Sigma Chemical Corp., St. Louis, MO, USA; 150 mg/kg body weight) or with six IP injections of hydroxyurea (HU; 1g/kg) every 6 hours. Two days after 5-FU injection or 2 hours after the last HU treatment, femoral BM was flushed out and resuspended in -MEM. For experiments with sublethal irradiation, female CBF1 (C57Bl/6 x CBA) F1 mice were exposed to 550 cGy 137Cs irradiation and used for longitudinal karyotypic study of clone kinetics. Retroviral vector-producing cell line GPE-86 cells producing a retrovirus containing human ADA cDNA expressed internally from the human phosphoglycerate kinase (PGK) promoter (PGK-hADA) was a generous gift of Dr. David A. Williams from Indiana University (Indianapolis, IN,

787

USA). The details of the vector have been previously reported [15]. The PGK-hADA cell line produces the virus with a titer 1  106 CFU/mL. The supernatants from this cell line were shown to be free of helper virus by 3T3BAG mobilization assay [16]. To rule out the presence of helper recombinant retroviruses in reconstituted mice, serum collected from sacrificed animals was routinely assayed by this method. One day before transduction, flasks with confluent PGK-hADA cells were irradiated with 40 Gy, detached by trypsin-EDTA, and split 1:2 in -MEM with 20% fetal calf serum (FCS). Transduction of hematopoietic cells with recombinant virus BM cells (5.2  106 cells/10 mL) were cultured for 2 days in 0.1% gelatin-treated T-25 flasks at 37C in -MEM medium supplemented with 20% FCS, recombinant rat stem cell factor (50 ng/ mL; Amgen, Thousand Oaks, CA, USA), and recombinant human interleukin-6 (50 U/mL; Amgen). Prestimulated BM cells were transferred onto the monolayer of PGK-hADA cells (2.2  106 BM cells/flask) in the media containing 4 g/mL polybrene and the same cytokines for 48 hours. The hematopoietic cells were recovered from PGK-hADA cell line (6.0  105 cells/flask) and used for reconstitution of irradiated mice (5  105 cells per mouse). To study the relationship between the size of cellular inocula and the number of functioning clones, the irradiated mice were reconstituted with 8  105 (small dose) or 8  106 (large dose) cells. Analysis of recipient animals BM samples were collected under ether anesthesia from the femur of individual reconstituted mice 4, 5, 6, 7, 8, 9, 10, and 12 months after transplantation, as described previously[10]. In brief, BM was aspirated repeatedly from the left and the right femur alternately by puncture through the knee joint with a 22-gauge needle. Usually it was possible to collect 5  106 to 15  106 BM cells from the femur of the living mouse. Aliquots of BM from each mouse were injected into six irradiated female recipients for CFUS analysis, and the rest of BM cells were used for DNA isolation. Lethally irradiated female secondary recipients were injected IV with 2  105 to 3  105 BM cells from reconstituted mice (low efficiency of spleen colony seeding with hematopoietic cells of reconstituted mice has been previously demonstrated [10]). Individual macroscopic spleen colonies were isolated under a dissection microscope 10 days later and used for DNA analysis. Determination of donor CFU-S origin by polymerase chain reaction Male mice were used as donors of BM cells; recipients were always female. For identification of CFU-S origin, polymerase chain reaction (PCR) analysis of DNA obtained from individual spleen colonies was used. The primers were chosen in the C-terminal domain of the sex-determining region [17,18] of the mouse Y-chromosome (5-CTCCTGATGGACAAACTTTACG-3 sense and 5-TGAGTGCTGATGGGTGACGG-3 antisense), and a 444-bp fragment of genome was amplified. Thirty cycles of PCR amplification were used under the following conditions: denaturation 45 seconds at 94C; annealing 60 seconds at 60C; and extension 45 seconds at 72C. PCR and Southern blot analysis of retrovirally marked donor cells and CFU-S Standard procedures were used for preparation of high-molecularweight genomic DNA samples [19]. DNA from the total BM and

788

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

individual spleen colonies was extracted, and PGK-hADA provirus was detected by PCR. Primers in ADA coding sequences (5-GACAAGCCCAAAGTAGAACTGC-3 sense and 5-TGACCCCGAAGTCTCGCTCC-3 antisense) amplified a 418-bp fragment of the proviral genome. Thirty cycles of PCR amplification were performed under the following conditions: denaturation 30 seconds at 94C; annealing 30 seconds at 60C; and extension 30 seconds at 72C. DNA samples proven to be positive for hADA by PCR were used for standard Southern blot technique [19]. Restriction digestion with EcoR1, electrophoresis through a 1% agarose gel, transfer to Hybond N filter, and hybridization with ADA cDNA probe, prepared from the PCR-amplified 418-bp fragment of ADA gene, were performed as described previously [19]. Digestion with EcoR1 permitted analysis of individual clones of hematopoietic cells, because only one EcoR1 restriction site is present within the vector used. The alignment of bands was based on molecular weight standards (phage  DNA digested with HindIII). A total of 1,639 spleen colonies and 80 samples of BM DNA were studied. Analysis of sublethally irradiated animals BM samples were aspirated from individual sublethally irradiated mice 4, 9, 12, 16, and 20 months after irradiation. For CFU-S analysis, BM cells from each mouse were injected IV into six lethally irradiated female recipients (1 to 4  105 cells/mouse), and the rest of the BM cells were used for chromosome analysis. Chromosome analysis BM cells from sublethally irradiated mice were suspended in 1 mL of phosphate-buffered saline supplemented with 1% FCS and treated for 50 to 60 minutes with colchicine (10 g/mL). For karyotypic analysis of CFU-S at day 10 after transplantation, recipient mice were injected IP with 0.2 mL of colchicine (10 g/mL) and sacrificed 1 hour later. Individual spleen colonies were isolated under a dissection microscope and resuspended in 0.5 mL of phosphate-buffered saline supplemented with 1% FCS. Standard method was used for slide preparation. The technique for G-bands using trypsin and Wright stain was applied [20]. Chromosome analysis was performed according to the methods of Nesbitt and Francke [21]. Many individual colonies were marked by unique chromosomal abnormalities. The marker was considered as clonal in the case of detection of at least two mitoses with identical chromosomal changes. A total of 996 spleen colonies and 31 BM samples were studied. In each colony, 2 to 25 metaphases were analyzed, and more than 9,000 metaphases were analyzed. Statistics Statistical analysis was performed using the Student’s t-test. Calculation of the number of CFU-S clones was performed using Kulikov’s model [10].

Results Treatment of BM donors BM donors were treated with 5-FU to eliminate the proliferating cells and to increase the efficiency of retroviral gene transfer. To determine the mechanism by which 5-FU affects transduction efficiency, BM donors were treated with either 5-FU or six consecutive injections of HU, which is

also known to be a potent anti-proliferative drug [22,23]. Thus, six injections of HU were equally as effective as a single dose of 5-FU. After such treatment, less than 1% of CFU-S survived. Both the transduction efficiency of CFU-S (99% and 96% after 5-FU and HU treatment, respectively) and the fate of individual HSCs after transplantation (see later) appeared to be identical for BM grafts from both groups of donors. Hence, all data were analyzed together. These results suggest that the increase in transduction efficiency after 5-FU treatment is due to killing of proliferating cells and induction of proliferation of dormant progenitors, as previously reported [22]. Hematopoiesis in lethally irradiated mice reconstituted with retrovirally transduced BM cells Retrovirally transduced BM cells from the male donors were transplanted into irradiated female recipient mice. Donor hematopoietic cells were found in the reconstituted mice for at least 1 year after transplantation and the proportion of Y-positive (donor) CFU-S colonies was high (83— 100%, data not shown). The transduction efficiency of HSCs was high. Thus, the percentage of marked CFU-S derived from transduced HSC varied from 0% to 100% during the experiment. Assessment of the BM of individual mice exhibited drastic variations in the percentage of marked cells sampled serially. For instance, BM of mouse 5 contained 20% transduced CFU-S at 8 months, 72% at 9 months, and no marked CFU-S at 10 months. BM of mouse 6 at 7 months contained 67% marked CFU-S, 0% at 8 months, and 25% at 9 months. One explanation for such variability could be fluctuation of the clonal composition of the hematopoietic system. Clonal kinetics in reconstituted mice To determine the temporal in vivo contribution of stem cell–derived clones to BM hematopoiesis, we studied 590 marked individual CFU-S-derived colonies in 12 recipients that received retrovirally transduced BM cells. In total, 135 individually marked clones were revealed. Data from four mice studied for 1 year are shown in Figure 1. Representative blots of DNA from CFU-S-derived marked colonies from these mice are shown in Figure 2. Hematopoiesis was polyclonal, with about 20 to 30 clones simultaneously functioning at any given time point. The lifetime of an individual clone was short, and the overwhelming majority of clones appeared no more than once, i.e., each clone functioned for 1 month. Of 101 clones, only six were functioning for 2 to 3 months (mouse 5: clones 3 and 13; mouse 6: clone 8; mouse 8: clones 1 and 5; and mouse 9: clone 5). Mouse 8 was the only exception. Despite the 20 clones that were present, hematopoiesis, in general, was oligoclonal. In this recipient, it was maintained by clone 1 (of 62 marked colonies, 58 were the derivatives of clone 1) within 4 to 7 months after transplantation and by clone 5 within 5 to 12 months (of 106 marked colonies, 88 were derivatives of

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

789

Figure 1. Temporal dynamics of clonal fluctuations in four mice reconstituted with retrovirally marked BM cells. Each rectangle represents individual clones. Figures in rectangles indicate the unique clone identification number. Numbers in parentheses represent the number of CFU-S-derived colonies with the same unique integration site that belong to this clone. Rectangles with symbols indicate the persistent clones. (A) Mouse 5 and 6. (B) Mouse 8 and 9.

clone 5). The second characteristic of the observed clones was their small size (often 1—2 CFU-S). At 12 months, the mice were sacrificed and the clonal composition of the BM from four bones was studied. Only in mouse 8 was the unique hematopoietic clone (number 5) present in all organs; in the remaining mice, the clonal composition of different bones differed greatly. The great majority of the clones were present only in one bone, and only a few clones were revealed in two bones (Fig. 1). For instance, in mouse 5, 10 clones were detected in the left tibia and only one of them was present in the right tibia, whereas the remaining clones did not appear in other sites. Thus, this experiment confirmed that hematopoiesis in the reconstituted mice is the result of many small locally functioning short-lived clones. Relationship between size of cellular inocula and number of individual clones Taking into the account the number of observed individual CFU-S clones and the number of BM cells used for reconstitution, it is possible to calculate the probable number of primitive HSCs present in the BM using Kulikov’s model

[9]. However, for such a calculation, it is important to know that there is a dose-response effect between the number of detected marked HSC that give rise to clonogenic progenitors and the size of BM inocula. To address this question, we studied the relationship between the number of detected clones and the dose of injected cells. Two groups of recipients were reconstituted with 8  105 (small dose) and 8  106 (large dose) transduced BM cells. The survival of mice was approximately identical in both groups (Fig. 3A), whereas the reversion to recipient hematopoiesis was much higher in the group receiving the small dose (Fig. 3B). The percentage of marked colonies among Y-positive (donor) colonies was approximately the same in both groups of mice (Fig. 3C). These data suggest that local dispersion of hematopoietic clones is not the result of limiting dilution hematopoiesis. The kinetics of the clones in this experiment corresponded to those in previous studies [10,11]. Polyclonal hematopoiesis was observed in both groups of recipient mice. The calculated number of clones in the large-dose group (50 11) was approximately four times higher than in the small dose group (12 5). Hence, there is a correlation between the number of functioning clones and the num-

790

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

Figure 2. Long-term clonal behavior of engrafted BM cells. Analysis of four mice reconstituted with retrovirally marked BM cells. Total BM DNA at 12 months after transplantation and DNA from CFU-S-derived spleen colonies at various times after transplantation were cleaved with EcoR1 and analyzed by Southern blot. Mouse 5, lane 1: total BM DNA, lanes 2–6: CFU-S (clones 1– 3, Fig. 1A) 4 months after reconstitution; lane 7: total BM DNA, lanes 8–11: CFU-S (clones 3–6, Fig. 1A) 6 months; lane 12: total BM DNA, lanes 13–26: CFU-S (clones 12–14, Fig. 1A) 9 months. Mouse 6, lane 1: total BM DNA, lanes 2–4: CFU-S (clone 1, Fig. 1A) 4 months after reconstitution; lane 5: total BM DNA, lanes 6–14: CFU-S (clones 3–8, Fig. 1A) 5 months, lanes 15–20: CFU-S (clones 8–10, Fig. 1A) 6 months; lanes 21–26: CFU-S (clones 16–20, Fig. 1A) 7 and 9 months. Mouse 9, lane 1: total BM DNA, lanes 2–8: CFU-S (clones 1–3, Fig. 1B) 4 months after reconstitution; lane 9: total BM DNA, lanes 10–13: CFU-S (clones 4 and 5, Fig. 1B) 5 months; lane 14: total BM DNA, lanes 15–18: CFU-S (clones 5–7, Fig. 1B) 6 months, lanes 19–24: CFU-S (clones 8 and 9, Fig 1B) 8 months, lanes 25–27: CFU-S (clones 12 and 13, Fig. 1B) 9 months. Mouse 12, lane 1: total BM DNA, lanes 2–1: CFU-S (clones 1–5) 4 months after reconstitution; lane 12: total BM DNA, lanes 13– 23: CFU-S (clones 6–9) 5 months; lane 24: total BM DNA, lanes 25–27: CFU-S (clone 6) 6 months.

ber of injected cells, but it is not strictly proportional. This could be due to either low accuracy of the calculations or higher efficiency of clone formation by HSC their content in the body is decreased.

Figure 3. (A) Survival, (B) proportion of donor (Y-positive) CFU-S, and (C) proportion of marked CFU-S among donor CFU-S in mice reconstituted with large and small inocula of BM cells. DNA from each isolated spleen colony was analyzed by PCR for Y and hADA sequences. Each column with error bar represents mean SE. x-axis time after reconstitution; y-axis percent of survival (A), percent of Y-positive CFU-S (B), and percent of hADA-positive CFU-S (C).

Hematopoiesis in sublethally irradiated mice Sublethal irradiation dose can induce the appearance of the unique chromosomal markers in noncycling cells, providing an alternative method of tracking the in vivo behavior of stem cells. The kinetics of clones with unique radiation markers was studied for 20 months after exposure to 5.5-Gy irradiation in 13 mice. This irradiation dose appeared to be effective in causing chromosomal changes and labeling a considerable proportion of HSCs. Figure 4 shows an example of a unique radiation marker in the cells of an individual spleen colony produced by BM of a mouse irradiated 9

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

791

Figure 4. Karyotype of individual spleen colony derived from the BM of sublethally irradiated mouse 24 (clone 1 {40,t(1,8)}, 9 months after irradiation).

months earlier. Throughout the entire experiment, the proportion of marked colonies varied from 10% to 94% (average 52.6% 26.4%) (Table 1). The chromosomal aberrations in the marked clones included deletions, inversions, insertions, duplications, reciprocal translocations, and Robertson translocations. Sometimes, there appeared to be additional chromosomes (in one case, a karyotype was 41, 19) as well as nonidentifiable chromosome markers. More complex chromosomal modifications involving three or more chromosomes (in 30%–50% of marked colonies) also were observed, a result that is not unusual for the applied radiation dose. Representative karyotypes of marked clones are listed in Table 2. The most frequent modifications affected large chromosomes 1, 2, 3, 4, and X, as well as chromosomes 11, 14, 17, and 18. In normal BM cells, no chromosome aberrations were revealed; only rare gaps were observed (data not shown). Table 1. Proportion (%) of colonies with chromosomal aberrations in sublethally irradiated mice Time after irradiation (mo) Mouse no. 21 22 23 24 25 26 29 30 33

4

25 (1/4) 38 (5/13) 46 (13/28) 36 (20/55) 10 (5/48)

9

89 (24/27) 31 (18/58) 77 (27/35) 58 (11/19) 80 (20/25) 11 (2/18) 92 (13/14) 29 (10/35)

12

16

20

83 (5/6) 92 (13/14)

47 (34/73)

60 (46/78)

37 (21/57)

58 (18/31)

54 (27/50) 92 (33/36) 17 (1/6) 39 (7/18) 29 (16/55)

51 (22/43)

92 (13/14) 30 (14/47) 64 (21/33) 29 (9/31) 94 (34/36) 44 (12/27)

Numbers in parentheses indicate number of marked colonies/number of studied colonies.

In 13 experimental mice, 195 marked HSC-derived clones (CFU-S) were obtained. The mean number of marked clones during the entire experiment was 16.1 1.2 (12–22 per mouse). The clonal kinetics is presented in Figure 5. In this ex-

Table 2. Chromosomal markers detected in CFU-S–derived colonies in a sublethally irradiated mouse (mouse 23)

Clone no. 1 2 3

4 5 6 7 8 9 10 11 12 13 14 15 16 17

Karyotype 40, t(2;5), t(7;15), t(11;16), del(8), del(9), del(18) 40, t(2;3;7)?, del(8)? 40, t(1;5), t(3;6), del(5), t(7;14), t(10;X), der(11), der(18)

40, t(X;17) 40, t(1;11), t(3;17), t(4;9) 40, inv(19) 40, der(1), t(2;X), del(5) 40, t(4;14), del(5) Nonidentifiable* 40, t(4;10), del(19) 40, del(1), t(3;16) 40, t(4;13), del(18)? 40, del(3) 40, t(6;14) 40, t(1;7) 40, der(1), der(4), del(6) 40, 1, 2, 6, 15, +4mar

Number of Time of CFU-S carrying detection after the identical irradiation (mo) clonal marker 4

6

4 4 9 12 16 4 4 4 12 12 12 12 12 12 12 12 12 12 16

3 1 20 11 32 1 1 1 1 1 1 1 1 2 1 1 1 1 1

*In the case of nonidentifiable markers, it was impossible to obtain accurate karyotypic analysis of the metaphases, but it still was possible to identify a unique marker.

792

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

Figure 5. Temporal dynamics of clones in nine sublethally irradiated mice. The height of rectangle indicates the number of CFU-S–derived colonies with the same unique radiation marker. Pattern of rectangle indicates the time after irradiation.

periment, as well as in the studies with retrovirally marked HSCs, mainly small short-lived clones, which substituted for each other sequentially, contributed to hematopoiesis. Hematopoiesis was never monoclonal. The majority of individual clones were observed only once. Rarely (17%), clones were detected twice in a sequential manner, i.e., the duration of functioning of such clones did not exceed 4 to 6 months. However, about 10% of clones were long-lived and functioned during the entire life of the mouse. Thus, in eight mice, which were examined at least three times, 13 of 118 investigated clones were long-lived (10.8% 1.6%). In experiments with retrovirally marked HSCs, we found only two long-lived clones of 659 studied (0.3% 0.2%) [10,11]. These data suggest that the transduced cells (cells that were in the S phase of the cell cycle during the time of gene transfer) rarely retain the high potential necessary for long-term repopulation. These findings are consistent with results from other studies [3,24,25].

Discussion To study the biology of primitive HSCs, the following requirements must be fulfilled. First, primitive HSCs should

be individually and uniquely marked, permitting the tracking of the fate of their progeny. Second, the methods should allow one to follow the fate of individual HSC-derived clones during the entire life of an animal. Third, to analyze all the functioning clones rather than only large ones, the sensitivity of the method should permit detection of the contribution of the individual primitive HSC. Fourth, only short-living descendant cells should be analyzed, to avoid invalidation of the results by long-living cells such as macrophages and some populations of lymphocytes. Finally, the methods should not alter the biologic properties of HSCs. The two techniques applied in this study mainly satisfy these requirements. Both retroviral gene transfer and exposure to irradiation marked cells individually, allowing the identification not only of initially marked cells, but also their progeny. The method of BM aspiration permits examination of the fate of individual HSC during the entire life of the mouse. Furthermore, analysis of the hematopoietic system in animals with marked HSCs by transplantation of BM cells to secondary irradiated recipients and analysis of the spleen colonies increase the sensitivity of the method up to the level of one marked cell (CFU-S). As CFU-S refers to a

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

category of short-lived (about 4 weeks) precursor cells, there are no distortions by long-lived cells [26,27]. It should be noted that the methods used have some drawbacks. Serial BM sampling may be responsible for the change in hematopoietic clonal composition. However, this possibility is highly unlikely, because essentially the same results (existence of multiple unique clones in different hematopoietic sites) were revealed in unaspirated sites such as the tibias and spleen of the same mouse (Fig. 1 and [10]). In addition, retroviral vector transduction is possible only in cells that proliferate during the procedure of gene transfer. It is not clear if there are changes in the properties of dormant HSCs after their in vitro culture and entry into the cell cycle. The method of clonal analysis using radiation-induced markers labels cells irrespective of their position in a cell cycle. However, a major drawback is associated with radiation itself. Sublethally irradiated HSCs may have a different fate in an organism and exhibit different properties in comparison with normal HSCs. Nevertheless, that the results obtained using the two essentially different methods were very similar suggests that these drawbacks were not limiting. Both methods of clonal analysis independently demonstrated that 1) hematopoiesis is the result of the contributions of small clones of hematopoietic cells; 2) the lifespan of a majority of such clones is only 1 to 2 months; and 3) the clones replace one another sequentially. The identification of primitive HSCs capable of producing long-lived clones, which exist during the entire life of a mouse, was detected by the radiation marker technique only. Long-lived clones constituted up to 10% of all hematopoietic clones during the life of the mouse. The possibility that the radiation-induced chromosome aberrations conferred altered growth parameters and caused clonal persistence is not likely, because all of the disparate aberrations would cause the same effect. These differences cannot be related to the fact that in mice reconstituted with retrovirally marked HSCs, the sampling interval was three times shorter than in sublethally irradiated mice, because essentially the same results were observed in reconstituted mice sampled every 3 to 4 months [10,11]. More probably, the cause for this discrepancy is that radiation marks HSCs independently of their position in the cell cycle, whereas retroviral marking is possible only in proliferating cells. Therefore, it is reasonable to assume that long-lived HSCs are in G0 and generally escape retroviral labeling. The reduction of HSC proliferative potential after their transit to a G0/G1 border supports this suggestion [7]. In addition, some HSCs can undergo a few cell divisions and then return to dormancy for indefinite periods of time [28,29]. Nevertheless, such HSCs are closer to the G0/G1 border and could be more quickly induced to in vitro colony formation by cytokines as compared with primitive HSCs [30]. In accordance with these studies, our results show that retrovirally marked HSCs (i.e., the cells that proliferated during transduction) can become dormant after transplanta-

793

tion, but then can start to proliferate again after variable time intervals during the lifespan of the recipient mouse. Earlier studies suggested that several months after reconstitution of mice with retrovirally marked HSCs, hematopoiesis was oligoclonal or monoclonal and one or two functioning clones maintained hematopoiesis after serial transplantation [3,31,32]. It was proposed that these data substantiate the self-renewal properties of HSCs. However, as discussed earlier, HSCs are able to return to G0, and there is a chance that their progeny would be found in the future analyses. It is important to consider that monoclonal hematopoiesis was clearly demonstrated only in studies with retrovirally marked HSCs [31]. When other techniques were used, polyclonal hematopoiesis was observed [33–36]. What is the cause of the significant differences between our results and those of previous studies using essentially the same technique of marking HSCs with retroviral vectors? It is impossible to exclude the differences in transduction protocols, mice used, and vector used. However, a more probable explanation is that our group was the first to examine the HSC clones from BM using the CFU-S assay throughout the entire life of the mouse. Previous studies used Southern blot analysis of total BM and therefore focused only on larger clones, which constituted 5% to 10% of all hematopoietic cells in the sample. Consequently, most of the smaller clones that were analyzed in our study potentially would not have been detected in previous studies and, thus, might have given the false impression of monoclonality or oligoclonality. In addition, longitudinal studies of the peripheral blood were focused on more mature and less primitive cells than the CFU-S [31]. Our data, on the other hand, are based on DNA analysis of primitive BM precursors. Although the CFU-S as well as any other hematopoietic progenitor is part of the hematopoietic cell hierarchy, the patterns of distribution within the mature cell clones may not necessarily be identical. There is a possibility that CFU-S may not function during steady-state hematopoiesis and are only active during stress conditions. In the future, it will be necessary to use sensitive techniques to study the CFU-S as well as the mature hematopoietic cell compartments to understand the clonal dynamics of normal hematopoiesis during the transition from primitive to mature cells.

Acknowledgments This work was supported by grants from the Russian Fund for Fundamental Research (96-04-48155) and CRDF (RN1-419). We thank Dr. Sergey Sokol and Dr. John Dick for critically reading the manuscript and Marina Drutskaya for help in preparing the manuscript. The authors gratefully acknowledge the help of Dr. Dimitry Kuprash in choosing primers for PCR analysis of hADA and sex-determining region of Y-chromosome. We express our appreciation to Dr. Sergey Nedospasov, Dr. Regina Turetskaya, and Dr. Sergey Sokol for helpful suggestions in molecular biology techniques, and

794

N.J. Drize et al./Experimental Hematology 29 (2001) 786–794

Dr. Sergey Kulikov for multinomial calculations. We also thank Dr. David A. Williams for providing hADA-producing cell line.

References 1. Harrison DE, Astle CM Delaittre JA (1978) Loss of proliferative capacity in immunohemopoietic stem cells caused by serial transplantation rather than ageing. J Exp Med 147:1526 2. Ogden DA, Micklem HS (1976) The fate of serially transplanted bone marrow cell populations from young and old donors. Transplantation 22:287 3. Capel B, Hawley RG, Mintz B (1990) Long- and short-lived murine hematopoietic stem cell clones individually identified with retroviral integration markers. Blood 75:2267 4. Osawa M, Hanada K, Hamada H, Nakauchi H (1996) Long-term lymphohematopoietic reconstitution by a single CD34 low/negative hematopoietic stem cell. Science 273:242 5. Brecher G, Bookstein N, Redfearn W, Necas E, Pallavicini MG, Cronkite EP (1993) Self-renewal of the long-term repopulating stem cell. Proc Natl Acad Sci U S A 90:6028 6. Hellman S, Botnick LE, Hannon EC, Vigneulle RM (1978) Proliferative capacity of murine hematopoietic stem cells. Proc Natl Acad Sci U S A 75:490 7. Gothot A, Pyatt R, McMahel J, Rice S, Srour EF (1997) Functional heterogeneity of human CD34() cells isolated in subcompartments of the G0/G1 phase of the cell cycle. Blood 90:4384 8. Traycoff CM, Orazi A, Ladd AC, Rice S, McMahel J, Srour EF (1998) Proliferation-induced decline of primitive hematopoietic progenitor cell activity is coupled with an increase in apoptosis of ex vivo expanded CD34 cells. Exp Hematol 26:53 9. Kay HEM (1965) How many cell-generations? Lancet 11:418 10. Drize NJ, Keller JR, Chertkov JL (1996) Local clonal analysis of the hematopoietic system shows that multiple small short-living clones maintain life-long hematopoiesis in reconstituted mice. Blood 88:2927 11. Drize N, Chertkov J, Sadovnikova E, Tiessen S, Zander A (1997) Long-term maintenance of hematopoiesis in irradiated mice by retrovirally transduced peripheral blood stem cells. Blood 89:1811 12. Morrison SJ, Wandycz AM, Akashi K, Globerson A, Weissman IL (1996) The aging of hematopoietic stem cells. Nat Med 2:1011 13. Harrison DE, Stone M, Astle CM (1990) Effects of transplantation on the primitive immunohematopoietic stem cell. J Exp Med 172:431 14. Weiss R, Teich N, Varmus H, Coffin J (1982) RNA tumor viruses. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory 15. Luskey BD, Rosenblatt M, Zsebo K, Williams DA (1992) Stem cell factor, interleukin-3, and interleukin-6 promote retroviral-mediated gene transfer into murine hematopoietic stem cells. Blood 80:396 16. Danos O (1991) Methods in molecular biology: viral vectors for gene expression. Clifton, NJ: Humana Press, p. 17. 17. Mardon G, Page DC (1989) The sex-determining region of the mouse Y chromosome encodes a protein with a highly acidic domain and 13 zinc fingers. Cell 56:765

18. Nagamine CM, Chan KM, Kozak CA, Lau YF (1989) Chromosome mapping and expression of a putative testis-determining gene in mouse. Science 243:80 19. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory 20. Seabright M (1971) A rapid banding technique for human chromosomes. Lancet 2:971 21. Nesbitt MN, Francke U (1973) A system of nomenclature for band patterns of mouse chromosomes. Chromosoma 41:145 22. Uchida N, Friera AM, He D, Reitsma MJ, Tsukamoto AS, Weissman IL (1997) Hydroxyurea can be used to increase mouse c-kitThy1.1(lo)Lin/loSca-1() hematopoietic cell number and frequency in cell cycle in vivo. Blood 90:4354 23. Randall TD, Weissman IL (1997) Phenotypic and functional changes induced at the clonal level in hematopoietic stem cells after 5-fluorouracil treatment. Blood 89:3596 24. Capel B, Hawley R, Covarrubias L, Hawley T, Mintz B (1989) Clonal contributions of small numbers of retrovirally marked hematopoietic stem cells engrafted in unirradiated neonatal W/Wv mice. Proc Natl Acad Sci U S A 86:4564 25. Lemischka IR, Raulet DH, Mulligan RC (1986) Developmental potential and dynamic behavior of hematopoietic stem cells. Cell 45:917 26. Li CL, Johnson GR (1992) Long-term hemopoietic repopulation by Thy-1lo, Lin, Ly6A/E cells. Exp Hematol 20:1309 27. van der Sluijs JP, de Jong JP, Brons NH, Ploemacher RE (1990) Marrow repopulating cells, but not CFU-S, establish long-term in vitro hemopoiesis on a marrow-derived stromal layer. Exp Hematol 18:893 28. Sato T, Laver JH, Ogawa M (1999) Reversible expression of CD34 by murine hematopoietic stem cells. Blood 94:2548 29. Huang S, Law P, Francis K, Palsson BO, Ho AD (1999) Symmetry of initial cell divisions among primitive hematopoietic progenitors is independent of ontogenic age and regulatory molecules. Blood 94:2595 30. Gothot A, Pyatt R, McMahel J, Rice S, Srour EF (1998) Assessment of proliferative and colony-forming capacity after successive in vitro divisions of single human CD34 cells initially isolated in G0. Exp Hematol 26:562 31. Jordan CT, Lemischka IR (1990) Clonal and systemic analysis of longterm hematopoiesis in the mouse. Genes Dev 4:220 32. Keller G, Snodgrass R (1990) Life span of multipotential hematopoietic stem cells in vivo. J Exp Med 171:1407 33. Harrison DE, Astle CM, Stone M (1989) Numbers and functions of transplantable primitive immunohematopoietic stem cells: effects of age. J Immunol 142:3833 34. Abkowitz JL, Ott RM, Holly RD, Adamson JW (1988) Clonal evolution following chemotherapy-induced stem cell depletion in cats heterozygous for glucose-6-phosphate dehydrogenase. Blood 71:1687 35. Abkowitz JL, Linenberger ML, Newton MA, Shelton GH, Ott RL, Guttorp P (1990) Evidence for the maintenance of hematopoiesis in a large animal by the sequential activation of stem-cell clones. Proc Natl Acad Sci U S A 87:9062 36. Fialkow PJ (1982) Cell lineages in hematopoietic neoplasia studied with glucose-6-phosphate dehydrogenase cell markers. J Cell Physiol Suppl 43:37