Haematological Scientific Center Russian Academy of Medical Science Novosykovsky pr. 4a 125167 Moscow, Russia

The hematopoietic system is created by hierarchical progression of commitment events starting from the primitive hematopoietic stem cell (HSC). By strict definition, the primitive HSC is the founder cell for the entire hematopoietic system, i.e. it is capable of long-term support of hematopoiesis both in vivo and in vitro. It is generally assumed that the primitive HSC has the capacity both for differentiation and self-renewal (Chertkov, Gurevitch, 1984) Nevertheless, crucial question about capacity of HSC for inexhaustible proliferation remains unanswered Many attempts to prove unlimited proliferative capacity of HSC by bone marrow cell passages always demonstrated the exhaustion of HSC potential after 3-5 transfers (Micklem et al, 1987). However, the technical reservation prevented from getting the conclusive evidence of inability of HSC to maintain itself. There is the other way to answer this intriguing question. if the founder cell has the capacity for self-renewal, hematopoiesis would be stochastic and many primitive HSCs would be functioning simultaneously. On the other hand, if the primitive HSC has no capacity to self-renew, hematopoiesis would be characterised by clonal succession. According to this model, primitive HSCs are used sequentially, one after the other, producing clones which replace each other ( Rosendaal, Adam, 1984). Reconstitution of irradiated mice with bone marrow cells carrying specific markers (HSC marked by retroviral integration) was used for characterisation of clonal succession (Williams et al, 1984; Lemischka et al, 1986; Snodgrass, Keller, 1987; Capel et al, 1989; Jordan, Lemischka, 1990). As a rule, after several months of " clonal fluctuations" , monooligoclonal hematopoiesis occurs and longevity of clone functioning can be as long as 2 5 years. In contrast, the results obtained with HSC without transduced DNA sequences suggested that hematopoiesis is polyclonal both in normal and reconstituted animals (Micklem et a11987; Harrisson et al, 1978 ). Differences in experimental results concerning the clonality of HSC may be due to the prestimulation of HSC during gene transfer procedure with pharmacological concentrations of hematopoietic growth factors. It is possible that all primitive HSC may be induced to proliferate by cytokines. Furthermore, progenitors with selective proliferative advantage could gradually replace all others, and after a time, only this one or several cell clones will be preserved (mono-oligoclonal hematopoiesis). In order to verify this hypothesis and to study the possible clonal succession of HSC, the next conditions must be realized: clonal composition of the hematopoietic system must be studied sequentially in the same reconstituted mouse; the method used must be very sensitive to detect a great number of simultaneously functioning clones during clonal selection; and the method of gene transfer without prestimulation with pharmacological concentration of cytokines should also be used. Here all these conditions were realized and hematopoietic cell clonal succession has been demonstrated conclusively. The proliferative potential and proliferative activity of more mature HSC, CFU-S-11 , in animals reconstituted both with treated by exogenous cytokines and untreated hematopoietic cells were characterized.

Materials and Methods

Mice. Male and female (C57BI/6 x DBA/2) F 1mice, 3-5 months-old were used. Long-term hone marrow culture. Femoral bone marrow was cultured in Fisher's medium supplemented with 20% of 2: 1 mixture of horse and fetal bovine serum, 10high -6 M hydrocortisone , L-glutamine and antibiotics as described (Chertkov et al, 1983). CFU-S-ll assay, spleen repopulating ability and Hydroxyurea "suicide '1 of CFU-S-ll Recipients in the spleen colony-forming (CFU-S) assay were exposed to 1200 cGy 137 Cs irradiation (the dose rate 18 cGy/min, IPK irradiator). The dose was delivered into two equal fractions, given 3 h apart Irradiated female mice were injected i. v. with 1 to 4 x 10high 5 marrow from reconstituted mice. Individual macroscopic colonies were isolated under dissection microscope 11 days later and used both for DNA analysis and for determination of the selfrenewal capacity (spleen repopulating ability /SRA/. number of daughter CFU-S-8 per 11-day colony). The colony cells (0.2 colony equivalent) were injected into irradiated secondary recipients. The number of daughter colonies were counted 8 days later and the number of colonies generated per one CFU-S-11 was calculated. Proliferative activity of CFU-S-11 was determined by hydroxyurea (HU) suicide. Bone marrow cells from reconstituted mice were incubated in RPMI 1640 media supplemented with 5% fetal bovine serum for 2 hours at 37oC with and without 1 mg/ml HU, washed trice with Hanks' balanced salt solution (HESS) and injected ( 1-2 x 10 high 5) into irradiated recipients Number of spleen colonies was determined following 8 days. Suicide (Su) was calculated by next equation:

Retroviral vector and producing cellline. GP+E86 cell line producing a retrovirus containing human ADA cDNA was a generous gift ofD.A. Williams. The details of the vector have been previously reported (Luskey et al, 1992). This cell line produces virus with a titer 1 x 10 high 6 CFU/ml. One day before use for infection of HSC, flasks with confluent producer cells were treated with 5 µg/ml of mitomycin C at 37°C for 2.5 hours, repeatedly washed with HBSS, detached by trypsin-EDT A and split 1 :2 in aMEM with 20% fetal bovine serum. Bone marrow transfection. Two protocols of transfection were used. For prestimulation, a suspension of bone marrow cells from six 5-FU treated mice (150 mg/kg i.v., two days before cell collection), were cultured in 0.1% gelatine-treated T-25 flasks, at 37C in 10 ml alfa MEM with 20% fetal bovine serum and rhIL-6 ( 50-100 µg/ml) and rhSCF ( 50 µg/ml) (Protocol 1 ). In protocol 2 the similar cells were incubated over adherent celllayer of 15 Gy irradiated 3-4 weeks old long-term bone marrow cultures without exogenous cytokines. Following two days prestimulated bone marrow cells were transferred to a monolayer of producer cells in media with 4 µg/ml polybrene and with (protocol 1) or without (protocol 2) exogenous cytokines After 2 days of infection, the hematopoietic cells were extensively washed from producer cells and used for reconstitution of irradiated mice. The doses used and the number of reconstituted mice are shown in table 1.

Table 1 Characteristics of mice reconstituted with retrovirally marked bone marrow cells

Collection of hematopoietic cells for DNA analysis. Bone marrow samples were obtained at different time intervals from multiple groups of individual anaesthetized living mice in such a way to allow recovery between sample collection. Femoral bone marrow was aspirated by puncture through the knee joint by a syringe with a 22-gauge needle. Usually it was possible to obtain 5-15 x 10 high 6 bone marrow cells from a living mouse. Bone marrow cells from each transplanted mouse were injected into 6 irradiated female recipients for CFU-S and DNA analysis. Southern blot analysis. DNA from bone marrow and individual spleen colonies was extracted, digested with EcoR 1, electrophoresed through a 1% agarose gel, transferred to Hybond N filter and hybridized with ADA cDNA probe, prepared from 1.2 Kb PGK-ADA fragment. Digestion with EcoR 1 permitted analysis of individual clones of hematopoietic cells, since only one EcoRl restriction site is present within the LTRs of the vector used. Determination of the origin of CFU-S in reconstituted mice. Only male mice were used as donors of bone marrow cells, recipients were always female For identification of CFU-S origin a Y-chromosome probe (mouse Zfy-2) was used (a generous gift from Dr. L.G. Brown) Filters were rehybridized with this probe after stripping the ADA probe.

Results

The efficiency of gene transfer into CFU-S-II was high, 52-83% of CFU-S derived colonies contained the human ADA cDNA (protocol I ). Without exogenous cytokines the efficiency of transduction was lower, only 18-32% of CFU-S derived colonies were marked (protocol 2). In many instances, several proviral copies per genome were integrated as demonstrated by Southern analysis (Fig. I). Transduction efficiency into primitive HSC also appeared to be good since, as seen in table 2, marked clones were still present in CFU-S-II from mice up to I year after reconstitution. Analysis of clonal composition of mice reconstituted with cytokine stimulated HSC revealed that 2 and 4 months after reconstitution the integration patterns varied between different colonies of the same mouse, implying that the progeny of many different clones were functioning simultaneously. From 23 colonies studied in first 4 months after reconstitution 22 had different unique markers (Table 2).

Table 2. Clonal composition of the hematopoietic system at different times after reconstitution
with bone marrow cells prestimulated with exogenous cytokines.

Fig. 1 Human ADA cDNA sequence integration pattern of primary day 11 CFU-S colonies derived from bone marrow cells prestimulated with exogenous cytokines or with adherent cell layer. Lanes 1-7, DNA from colonies, produced by cy1okine prestimulated cells; lanes 8-11, DNA from colonies, produced by adherent cell layer prestimulated cells; lane 12, DNA from producer cell line GP+E86 (hADA).

Table 3 Clonal composition of hematopoietic system in different time after reconstitution
with non-stimulated by exogenous growth factors bone marrow cells.

Fig 2 Detection of y -chromosome sequences in CFU-S-derived spleen colonies from female mice reconstituted with transduced male hematopoietic cells Lanes 1-4, 6 months after reconstitution; lanes 5- 7, 12 months after reconstitution

Proliferative activity of CFU-S was studied in mice 6 months following reconstitution (Fig 3) The results were essentially the same in groups reconstituted by bone marrow cells transduced both with and without exogenous cytokines prestimulation. In mice reconstituted with large cell inoculum suicide rate was about 10%, ie CFU-S were mainly out of cell cycle as during steady-state hematopoiesis in normal animals Mice reconstituted with low cell dose demonstrated significant increase of suicide which was about 20% CFU-S with integrated human ADA cDNA had 2-fold higher suicide rate than unmarked CFU-S. However, because of low amount of studied marked colonies (9 from 42) the difference is insignificant.

Fig 3 CFU-S suicide in mice 6 months following reconstitution. 1,2, reconstitution according protocol 1, low and large dose of bone marrow cells, respectively; 3,4, reconstitution according protocol 2, low and large dose of bone marrow cells, respectively.

Spleen repopulating ability (SRA) of CFU-S-ll 6 and 12 months after reconstitution was significantly decreased in all experimental groups of mice (Fig. 4). The average number of daughter CFU-S per 11 day colony was 20-60 as compared with 150 for normal bone marrow derived colony. In all mice reconstituted with cytokine prestimulated cells SRA gradually decreased during 12 months. On the other hand, two of five mice reconstituted with transduced cells without cytokine prestimulation had higher SRA I year after reconstitution than that after 6 months

Fig 4 Spleen repopulating ability of CFU-S-ll in individual reconstituted mice 6 (full bars) and 12 (open bars) months after reconstitution A, B, reconstitution according protocol 1, low and large dose of bone marrow cells, respectively; C, D, reconstitution according protocol 2, low and large dose ofbone marrow cells, respectively

Discussion

The results presented here demonstrate that the difference in the developmental fate of primitive HSC, measured with traditional methods (competitive repopulation assay, serial bone marrow passages et al), as compared with methods of retroviral gene transfer, could be explained by use of pharmacological doses of exogenous hematopoietic growth factors. A combination of cytokines (IL-6 + SCF) induced high efficiency gene transfer into HSC, what suggests that quiescent HSC had been stimulated for proliferation. Efficiency of gene transfer after more physiological prestimulation of HSC on irradiated ACL of long-term bone marrow culture was significantly lower, possibly because of low proportion of HSC that had been triggered in cell cycle In mice reconstituted with bone marrow cells prestimulated with exogenous growth factors, many marked clones appeared simultaneously. Two to 4 months after reconstitution, essentially each marked CFU-S represents an unique clone (22 of 23 colonies were represented by unique marker). Statistically, the results indicate that in the first months after transplantation 200-400 clones are functioning simultaneously (confidence interval, 95%) CFU-S is a progenitor with relatively low proliferative potential and it can produce differentiated progeny not more than 4 -5 weeks both in vivo (Bodine et al, 1991 ; Bodine et al, 1992) and in vitro (Chertkov et al, 1987) Therefore, during 2 -4 months after reconstitution the revealed marked clones are progeny of primitive HSC, rather than initial CFU-S Hence, many if not all primitive HSC are proliferative at least 4 months after transplantation During this time, selection of clonogenic progenitor(s) with proliferative advantages is possible, and this clone(s) can replace all others. As a result of such selection, all marked cells would have the same unique marker Indeed, 6 -12 months after reconstitution, hematopoiesis appeared to be oligo-monoclonal as shown here and by others (Rosendaal, Adam, 1984;Capelet al, 1989;Limet al, 1989) Essentially, different results were obtained after transplantation of cells not stimulated by exogenous cytokines The efficiency of gene transfer was significantly lower, however the system was suitable for the study of the hematopoietic clonal composition after reconstitution. No phase of polyclonality was observed in this model with use of an ACL for prestimulation. From the first months after transplantation only lor 2 marked clones have been observed. The longevity of marked clones existence was shorter than 2 -3 months because they were revealed only once during one year; analysis was performed each 2 -3 months. It appears that no stem cell selection took place and hematopoiesis was maintained by clonal succession. If appear simultaneously. Hence, the cause of peculiarity of results obtained with gene transfer technique is the use of pharmacological concentration of exogenous cytokines. If it is possible to avoid such stimulation of HSC and to use a more physiological gene transfer, the same clonal succession and limited proliferative potential of primitive HSC would be observed. The other evidence, that transduced foreign sequences do not change the proliferative potential of HSC, is the same SRA in marked and non-marked colonies The clonal succession of hematopoietic cells was also supported by the data about increase of SRA in 12 months as compared with 6 months in the group of mice reconstituted with the physiologically prestimulated cells (protocol 2). It is difficult to understand how can SRA increase in the progeny of the same primitive HSC In the case of clonal succession this effect could be explained by replacing the clones with low proliferating potential by clones which were less exhausted. The incapacity of primitive HSC to maintain itself is supported by the data obtained here and earlier (Chertkov et al, 1993 ), that after reconstitution with large cell innoculum the proliferative potential of progenitors, as measured by SRA, is higher than after reconstitution with low cell dose. Of course, this could not happen if the primitive HSC has unlimited proliferative potential Proliferative activity of CFU-S, as measured by HU suicide, was increased only in groups reconstituted with low dose of hematopoietic cells It is known, that clone size is bigger in animals transplanted with low doses of bone marrow cells ( Micklem et al, 1987). The data suggest that this increase of clone size is due at least partially by expansion of CFU-S at the expense of increased proliferative rate On the whole, the results support the clonal succession model of hematopoiesis and mortal nature of primitive HSC. The data demonstrate the possibility of using this model for the sequential study of the clonal composition of the hematopoietic system and primitive HSC developmental potential

Acknowledgements

The authors wish to thank Drs D. A. Williams and L.G.Brown for generous gift of the GP+E86 (hADA) cell line and the y probe.
This study was supported in part by ISF (Soros fund) and RFFR (Russian Fund for Fundamental Research)

References

Bodine DM, Crosier PS, Clark SC (1991) Effects of hematopoietic growth factors on the survival of primitive stem cells in liquid suspension culture. Blood 78:914-920

Bodine DM, Orlic a, Birkett NC, Seidel NE, Zsebo KM (1992) Stem cell factor increases coJony-forming unit-spleen number in vitro in synergy with interleukin-6, and in vivo in SI/Sld mice as a single factor. Blood 79:913-920

Capel B, Hawley R, Covarrusias L, Hawley T, Mintz B (1989) Clonal contribution of small numbers of retrovirally marked hematopoietic stem cells engrafted in unirradiated neonatal W/W mice. Proc Natl Avcad Sci USA 86:4564-4567

Chertkov JL, Drize NI, Gurevitch OA, Udalov GA (1983) Hemopoietic stromal precursors in long-term culture of bone marrow I. Precursors characteristics, kinetics in culture, and dependence on quality of donor hemopoietic cells in chimeras. Exp Hematol 11:231-242

Chertkov JL, Gurevitch OA (1984) Hematopoietic stem cell and its microenvironment. Moscow. Meditzina

Chertkov JL, arize NI, aeryugina El, Udalov GA (1987) Individual clones of hemopoietic cells in murine long-term bone marrow cultures Leukemia 1: 491-496

Chertkov II, Iiang S, Lutton JD, Harrison I, Levere Ra, Tiefenthaler M, Abracham NG (1993) The hematopoietic stromal microenvironment promotes retrovirus-mediated gene transfer into hematopoietic stem cells. Stem Cells II 218-227

Jourdan CT, Lemischka IR ( 1990) Clonal and systemic analysis of long-term hematopoiesis in the mouse. Genes Dev 4.220

Harrison DE, Astle CM, aelaittre IA (1978) Loss of proliferative capacity in immunohemopoietic stem cells cause by serial transplantation rather then ageing. I Exp Med 147:1526-1531

Lemischka IR, Raulet OH, Mulligan RC ( 1986) Developmental potential and dynamic behaviour of hematopoietic stem cells. Cell 45.917-923

Lim B, Apperley JF, Orkin SH, Williams DA ( 1989) Long-term expression of adenosine deaminase in mice transplanted with retrovirus-infected hematopoietic stem cells Proc Nat Acad Sci USA 86 8892-8895

Luskey BD, Rosenblatt M, Zsebo K, Williams DA (1992) Stem cell factor, interleukin3, and interleukin-6 promote retroviral-mediated gene transfer into murine hematopoietic stem cells Blood 80:396-405

Micklem HS, Lennon JE, Amsell 10, Gray RA (1987) Minbers and dispersion of repopulating cell clones in radiation chimeras as a function of injected cell dose. Exp Hematol 15:251-257

Rosendaal M, Adam J (1984) Haemopoiesis by clonal succession? Blood Cells 10:473- 485

Snodgrass R, Keller G ( 1987) Clonal fluctuation within the hematopoietic system in mice reconstituted with retrovirus-infected cells EMBO J 6:3955-3961

Williams DA, Lemischka IR Nathan DG, Mulligan RC (1984) Introduction of new genetic material into pluripotent stem cells of the mouse Nature 310:476-482

Abstract

Developmental fate of retrovirally marked hematopoietic stem cells (HSC) critically depends on the use of pharmacological concentration of growth factors during gene transfer procedure Cytokines (interleukin-6 plus stem cell factor) stimulated HSCs proliferate at least 4 months after reconstitution. At this time hundred of hematopoietic clones are exist simultaneously. Later as a result of clonal selection only oligo-clonal hematopoiesis was detected. Vice verse, after more physiological prestimulation of HSC on irradiated adherent cell layer of long-term bone marrow culture no phase of polyclonal hematopoiesis has been observed Hematopoiesis was according to clonal succession model and replacing each other hematopoietic cell clones were revealed The longevity of clone existence was not longer than 2-3 months The proliferative activity of CFU-S was higher in reconstituted with small dose of bone marrow cells mice That could be connected with enlargement of individual clone size in these animals The results support the clonal succession model of hematopoiesis and incapacity of primitive HSC to maintain itself The data demonstrate the possibility of using this model for the sequential study of the clonal composition of the hematopoietic system and primitive HSC developmental potential