Comparison of the effectiveness of bone marrow and spleen stem cells for platelet repopulation in lethally irradiated mice

To determine whether stem cells from spleen differ from those of bone marrow in their ability to support platelet repopulation after lethal irradiation we compared normal, splenectomized and plethoric (3 weeks exposure to CO) spleen cell recipients with similar groups of mice reconstituted with comparable numbers of bone marrow stem cells. Female LAF1 mice were given 825-950 R from a 60Co source and were transplanted with 0, 0.5, 1, 2, or 4 x 10(6) bone marrow or 10-64 x 10(6) spleen cells. CFUs content of the transplant was calculated from day 9 splenic nodules in mice receiving 5 x 10(4) bone marrow or 1.0-1.6 x 10(6) spleen cells. Blood volumes, platelet counts and hematocrits were determined on day 12. Total circulating platelets increased with increasing stem cell dose (range 50-1000 CFUs) after both types of transplant. Plethoric mice always had lower platelet levels than controls even when corrected for expanded blood volume. There was little difference between values for normal and splenectomized mice. Platelet production per CFUs in all groups except splenectomized spleen cell recipients declined with large transplants presumably because of feedback inhibition. Although the number of splenic megakaryocytes in spleen cell recipients was approximately 2.5 times as great as in bone marrow recipients, platelet levels were significantly higher only in normal and splenectomized mice receiving more than 300 CFUs. No differences were found between the two types of transplant when plethoric hosts were used.     

Short- and long-term repopulation of lethally irradiated mice by bone marrow stem cells enriched on the basis of light scatter and Hoechst 33342 fluorescence

Murine bone marrow subpopulations enriched in hemopoietic stem cells were transfused into lethally irradiated hosts to determine the contribution of host cells and two types of donor cells to marrow repopulation. Donor cell suspensions were a mixture of marrows from two congenic lines of mice containing electrophoretically distinguishable alloenzymes of phosphoglycerate kinase (PGK-A and PGK-B). The donor cells were sorted by high forward light scatter, low-to-intermediate perpendicular light scatter, and low Hoechst 33342 fluorescence intensity. The congenic hosts contained a third distinct marker, glucose phosphate isomerase (GPI-A). The two markers in the donor cells allowed determination of the clones generated by the seeded cells over a 36-week period of observation. The clone number declined rapidly during the first 12 weeks following transplantation and reached stable levels at 20 weeks, indicating the number of long-term repopulating cells (LTRC). The sorted subpopulation was enriched 170-fold for day-13 spleen colony-forming units (CFU-S), 235-fold for cells providing a 30-day survival, and 136- to 160-fold for LTRC. Survival for the 36-week observation period was 40%-100% for groups of hosts receiving 100-3000 sorted cells and 80% for controls receiving 2 x 10(5) unsorted cells. In all groups, similar distribution of phenotypes among peripheral blood erythrocytes, platelets, and lymphocytes at 36 weeks suggested that the repopulating donor stem cells were pluripotential. Transfusion of 3000 sorted cells, containing about 5 LTRC and 60 CFU-S, assured continuous repopulation with 95%-100% donor cells 4 to 36 weeks after transplantation, whereas significant numbers of host cells re-emerged temporarily or permanently when lower numbers of LTRC and CFU-S were transfused. The data indicate that both the quality and quantity of pluripotential stem cells in sorted bone marrow are important for complete long-term marrow reconstitution.     

Soluble factor(s) from bone marrow cells can rescue lethally irradiated mice by protecting endogenous hematopoietic stem cells

OBJECTIVE: Ionizing radiation-induced myeloablation can be rescued via bone marrow transplantation (BMT) or administration of cytokines if given within 2 hours after radiation exposure. There is no evidence for the existence of soluble factors that can rescue an animal after a lethal dose of radiation when administered several hours postradiation. We established a system that could test the possibility for the existence of soluble factors that could be used more than 2 hours postirradiation to rescue animals. MATERIALS AND METHODS: Animals with an implanted TheraCyte immunoisolation device (TID) received lethal-dose radiation and then normal bone marrow Lin- cells were loaded into the device (thereby preventing direct interaction between donor and recipient cells). Animal survival was evaluated and stem cell activity was tested with secondary bone marrow transplantation and flow cytometry analysis. Donor cell gene expression of five antiapoptotic cytokines was examined. RESULTS: Bone marrow Lin- cells rescued lethally irradiated animals via soluble factor(s). Bone marrow cells from the rescued animals can rescue and repopulate secondary lethally irradiated animals. Within the first 6 hours post-lethal-dose radiation, there is no significant change of gene expression of the known radioprotective factors TPO, SCF, IL-3, Flt-3 ligand, and SDF-1. CONCLUSION: Hematopoietic stem cells can be protected in lethally irradiated animals by soluble factors produced by bone marrow Lin- cells.     

Early induction of immune resistance against leukemia in lethally total body irradiated mice reconstituted with syngeneic bone marrow cells obtained from previously immunized donor mice

BALB/c x DBA/2 F1 (CD2F1) mice were lethally irradiated and reconstituted with syngeneic bone marrow cells (SBMC) obtained from normal or previously immunized (against L1210 lymphatic leukemia) donors. These recipient mice are called TBI + SBMT or TBI + Imm-SBMT mice, respectively. TBI + Imm-SBMT, but not TBI + SBMT mice, were able to develop strong immune resistance against L1210 leukemia, but not against MOPC 104E plasmacytoma, if the immunization procedure (four i.p. injections at weekly intervals of immunogenic L1210 cells) was started as early as 7 days posttransplantation. Incubation of Imm-SBMC with mafosfamide (ASTA Z7654) before grafting abrogated the ability of the recipient mice to develop early resistance against the leukemia. Treatment of Imm-SBMC with monoclonal or polyclonal antibodies plus complement showed that two or three subpopulations of Imm-SBMC were necessary for the transfer of immune information against leukemia: T lymphocytes with phenotype Thy 1.2+, Lyt 1+2-, I-Ad-, macrophages with phenotype Mac-1+, I-Ad-, and probably asialo-GM 1+ cells. Recipient mice immunized against L1210 leukemia before TBI + SBMT do not develop early resistance to the leukemia.     

Physicochemical properties of circulating red blood cells of lethally x-irradiated mice treated with rat bone marrow.

1. Two months after injection of rat bone marrow into lethally X-irradiatedmice (950 r-RBM mice), 100% of the circulating RBC were serologically of therat type, indicating that the surface molecular configuration of RBC from theseexperimental mice are of the rat type.

2. The hemoglobin was found to be also very much like the rat type in itsease in crystallization, its alkali denaturation property, its electrophoretic property, and its tendency to form a paracrystalline state at low temperature.

3. These cells possessed dual osmotic properties; the relative hemoglobinconcentration released when cells were lysed in water was more comparable tothe rat type, but its temperature dependency was more comparable to the mousetype.

Submitted on April 9, 1957 Accepted on May 15, 1957     

Emperipolesis of marrow cells within megakaryocytes in the bone marrow of sublethally irradiated mice

The incidence of megakaryocytic emperipolesis was studied in the bone marrow of normal and X-irradiated mice. Two groups of mice received total body irradiation with a single dose of 5 Gy and one of the two groups had been treated with a radioprotective drug, ethiofos (WR-2721), before irradiation. Mice from a third group remained unexposed to irradiation and served as controls. The Wright-Giemsa stained bone marrow smears were analyzed every 5 days during a 30-day period, starting 1 day after irradiation. The number of megakaryocytes exhibiting the phenomenon was determined and expressed as an average value for every experimental group. The frequency of megakaryocytic emperipolesis was less than 15% of megakaryocytes from control smears but increased to 34% in mice that had only been irradiated and to 43% when mice were treated with WR-2721 before irradiation. In the last case, i.e., irradiation and treatment with a radioprotective drug, a positive correlation between the macrocytic megakaryocytes and elevated emperipolesis was noted. Under light microscopy, there were no signs of phagocytosis; engulfed cells remained unaltered with their normal structure intact. Granulocytic, erythroid, and lymphoid cells appeared to be the most frequent marrow cells engulfed by mature megakaryocytes. The number of incorporated cells in one megakaryocyte ranged from 1 to 3, though occasionally more than 6 were seen in macrocytic megakaryocytes. Based on our findings and on a review of the associated literature, we believe emperipolesis is an interesting cellular phenomenon related to the fast passage of marrow cells across the marrow-blood barrier, especially through the cytoplasm of megakaryocytes in response to an increased demand for cell delivery. The high demand for cell delivery which occurs after irradiation may cause certain mature bone marrow cells to take a transmegakaryocyte path to enter the circulation of the blood. Irradiation seems to have an immediate effect (observed after 24 h) on emperipolesis, suggesting that a humoral factor is involved in the pathogenesis.     

Development of a lethal mast cell disease in mice reconstituted with bone marrow cells expressing the v-erbB oncogene.

An animal model for malignant mastocytosis is described in mice reconstituted with bone marrow cells expressing the v-erbB oncogene. The lethal mast cell disease is characterized by massive infiltration of bone marrow, spleen, and several other visceral organs by connective tissue mast cells, which normally reside in the skin and the peritoneal cavity. As is frequently found in malignant mastocytosis, the v-erbB-induced mast cell disease was accompanied in some primary recipients by an acute myelogenous leukemia (AML) that killed all secondary recipients regardless of whether the AML was already evident in the primary host. The infiltrating mast cells stained strongly positive with berberine sulfate, suggesting that they were terminally differentiated and in vitro they showed only a weak proliferative capacity. The leukemias were clonal but apparently of different origin than the malignant mast cells, implying the transformation of two independent cell populations. Leukemic cells expressed various myeloid-specific markers as well as the B220 antigen, normally associated with the B-cell lineage. However, the Ig heavy chain genes were still in germ line configuration. In culture, these cells proliferated in the absence of exogenous growth factors and had the capacity to differentiate into mature myeloid cells. Preliminary experiments suggest that v-erbB may use parts of a signal transduction pathway normally coupled to the c-kit receptor. The v-erbB-induced malignant mast cell disease should provide a useful animal model for elucidating the cause for malignant mastocytosis in humans and to explore possible therapeutic strategies.     

Cultivation of leukemic human bone marrow cells in diffusion chambers implanted into normal and irradiated mice.

In order to elucidate the question of whether the maturation defect in vivo in acute leukemia is due to environmental or cellular factors, we have cultured human leukemic cells in a nonleukemic milieu, i.e., diffusion chambers implanted into the abdominal cavity of normal and irradiated mice. For each harvest, the cell count was measured and differential counts and the number of peroxidase-positive cells determined. The cell number decreased with time, without significant difference between culture in irradiated (500 rads) and normal mice. The blast cells succeeded only in developing distorted promyelocytes and myelocytes. There was a general pattern of increase in the number of peroxidase-positive cells. The study supports the concept that acute myeloid leukemia (AML) is a disturbance of cellular maturation due to cellular rather than environmental defects.     

Transient dyserythropoiesis in repopulated human bone marrow following transplantation: an ultrastructural study

S ummary . Transmission electron microscopy was used to examine marrow samples from 15 patients with aplastic anaemia or acute leukaemia who had been treated with bone marrow transplantation. There were 11 allogeneic, three syngeneic and one autologous graft. The purpose was to estimate the frequency, type and extent of dyserythropoietic change. Transient dyserythropoietic features were substantiated in all cases. Nuclear changes were present in 12 cases, iron laden mitochondria (sideroachrestic phenomena) in 10 and cytoplasmic contacts and/or connections between red cell precursors in 10. Dyserythropoiesis was most conspicuous in the majority of cases between 14 and 28 d after transplantation but it may persist for over 100 d. No deficit in red cell production was noted and it is proposed that dyserythropoiesis in this circumstance is a physiological rather than a pathological phenomenon.     

Ex vivo expansion of autologous bone marrow CD34(+) cells with porcine microvascular endothelial cells results in a graft capable of rescuing lethally irradiated baboons.

Hematopoietic stem cell (HSC) self-renewal in vitro has been reported to result in a diminished proliferative capacity or acquisition of a homing defect that might compromise marrow repopulation. Our group has demonstrated that human HSC expanded ex vivo in the presence of porcine microvascular endothelial cells (PMVEC) retain the capacity to competitively repopulate human bone fragments implanted in severe combined immunodeficiency (SCID) mice. To further test the marrow repopulating capacity of expanded stem cells, our laboratory has established a myeloablative, fractionated total body irradiation conditioning protocol for autologous marrow transplantation in baboons. A control animal, which received no transplant, as well as two animals, which received a suboptimal number of marrow mononuclear cells, died 37, 43, and 59 days postirradiation, respectively. Immunomagnetically selected CD34(+) marrow cells from two baboons were placed in PMVEC coculture with exogenous human cytokines. After 10 days of expansion, the grafts represented a 14-fold to 22-fold increase in cell number, a 4-fold to 5-fold expansion of CD34(+) cells, a 3-fold to 4-fold increase of colony-forming unit-granulocyte-macrophage (CFU-GM), and a 12-fold to 17-fold increase of cobblestone area-forming cells (CAFC) over input. Both baboons became transfusion independent by day 23 posttransplant and achieved absolute neutrophil count (ANC) >500/microL by day 25 +/- 1 and platelets >20,000/microL by day 29 +/- 2. This hematopoietic recovery was delayed in comparison to two animals that received either a graft consisting of freshly isolated, unexpanded CD34(+) cells or 175 x 10(6)/kg unfractionated marrow mononuclear cells. Analysis of the proliferative status of cells in PMVEC expansion cultures demonstrated that by 10 days, 99.8% of CD34(+) cells present in the cultures had undergone cycling, and that the population of cells expressing a CD34(+) CD38(-) phenotype in the cultures was also the result of active cell division. These data indicate that isolated bone marrow CD34(+) cells may undergo cell division during ex vivo expansion in the presence of endothelial cells to provide a graft capable of rescuing a myeloablated autologous host.     

Immunologic mechanisms in heavily irradiated mice treated with bone marrow

1. Humoral antibody production has been studied in severely irradiatedmice treated with isologous (same strain) or homologous (different strain)bone marrow.

2. The two methods of study involved functional end points of humoralantibody production as evidenced by in vivo lysis of rat erythrocytes or byregression of mouse leukosis E.L. 4 in histoincompatible mouse recipients.

3. Humoral antibody production was lost after irradiation and isologousmarrow treatment, but recovered partially in two weeks and almost completely in four weeks.

4. Established immunity was not abruptly terminated after irradiation andtreatment with either isologous or homologous marrow, although there waspremature loss of immunity to rat erythrocytes by the irradiated, isologousmarrow-treated mouse.

5. Permanent immunity could not be transferred by isologous marrow orspleen from immunized donors to irradiated recipients.

6. Treatment of mice histoincompatible to E.L. 4 leukosis with histocompatible donor bone marrow failed to establish rejection of the tumor.

7. These studies support the concept that humoral antibody production inirradiated, marrow-treated mice remains a function of the host rather thanof the transplanted tissues.

8. These studies failed to clarify the conflicting evidence concerning themechanism of the late illness that occurs after treatment of the irradiatedmouse with bone marrow from a different strain or species.

Submitted on June 25, 1958 Accepted on November 1, 1958     

NIH conference. Molecular genetic analysis of human lymphoid neoplasms. Immunoglobulin genes and the c-myc oncogene

Immunoglobulin genes responsible for individual antibodies are organized as discontinuous DNA segments in their germline form. As an uncommitted stem cell develops into an antibody-synthesizing plasma cell, rearrangements of these immunoglobulin gene segments serve to activate the genes and to generate the virtually unlimited capacity to synthesize antibodies that recognize potential antigens. The analysis of immunoglobulin gene structure and arrangement has been of immense value in the study of human lymphoid neoplasms. Recombinant DNA technology involving analysis of immunoglobulin gene arrangement has been used to classify neoplasms of previously uncertain lineage, aid in the diagnosis of neoplasms of the B-cell series, and define the state of differentiation of neoplastic B-cell precursors. Furthermore, the demonstration of translocation of a particular transforming gene, the c-myc oncogene, into the immunoglobulin gene locus in Burkitt's lymphoma has provided a major insight into the cause of malignant transformation of these lymphoid cells.     

Marrow repopulation in mice treated with busulphan or isopropyl methane sulphonate and bone marrow

By using karyotypic analysis of female mice treated with busulphan or isopropyl methane sulphonate (IMS), and injected with male bone marrow the donor contribution to both total marrow cellularity and spleen colony forming cells (CFU-S) was assessed for up to 6 months after transplant. In the mice treated with busulphan the marrow cells yielded metaphases of which between 40% and 83% were of donor type. Between 60% and 97% of metaphases in spleen colonies formed in irradiated mice were of donor type during the 24-week study period. In contrast, mice prepared for the transplant with IMS showed no cells of donor type at any time after transplant, neither did they possess CFU-S of donor type. We were therefore led to conclude that the donor cells made no contribution to longterm engraftment in mice prepared with IMS, whilst in those prepared with busulphan they were the predominantly active haemopoietic cells. These results are consistent with a model of haemopoiesis in which the most primitive cells reside in a 'niche' where they are resistant to the effects of IMS but susceptible to the action of busulphan. Busulphan may vacate some niches to allow engraftment by transplanted marrow, whilst IMS yields no unoccupied niches for grafted cells to occupy, and cannot therefore lead to a stable chimaerism.     

In vivo radioprotective effects of recombinant human granulocyte colony-stimulating factor in lethally irradiated mice.

The purpose of this study was to investigate the in vivo radioprotective effects of recombinant human granulocyte colony stimulating factor (rhG-CSF) in lethally irradiated BALB/c mice. We initially analyzed the effects of increasing doses of rhG-CSF on survival of mice receiving 700 cGy (LD100/30) single dose total body irradiation (TBI). While 1 microgram/kg to 100 micrograms/kg doses of rhG-CSF were not radioprotective, a dose-dependent radioprotection was observed at 200 micrograms/kg to 4,000 micrograms/kg rhG-CSF. We next compared four different rhG-CSF treatment regimens side by side for their radioprotective effects in LD100/30 irradiated mice. One hundred percent of control mice receiving phosphate buffered saline died within 21 days after TBI with a median survival of 14 days. The median survival was prolonged to 20 days and the actuarial 60-day survival rate was increased to 27% when mice received 2,000 micrograms/kg rhG-CSF 24 hours before TBI (P = .0002; Mantel-Peto-Cox). Similarly, the median survival time was prolonged to 24 days and the actuarial 60-day survival rate was increased to 33%, when mice were given 2,000 micrograms/kg rhG-CSF 30 minutes before TBI. Optimal radioprotection was achieved when 2,000 micrograms/kg rhG-CSF was administered in two divided doses of 1,000 micrograms/kg given 24 hours before and 1,000 micrograms/kg given 30 minutes before TBI. This regimen prolonged the median survival time of LD100/30 irradiated mice to more than 60 days and increased the actuarial 60-day survival rate to 62% (P = .0001; Mantel-Peto-Cox). By comparison, no survival advantage was observed when mice received rhG-CSF 24 hours post-TBI. Similar radioprotective effects were observed when mice were irradiated with 650 cGy (LD80/30). The presented findings provide conclusive evidence that rhG-CSF has significant in vivo radioprotective effects for mice receiving LD100/30 or LD80/30 TBI.     

Bone marrow cells other than stem cells seed the bone marrow after rescue transfusion of fatally irradiated mice

In a previous publication, iodinated deoxyuridine (125IUdR) incorporation data were interpreted as indicating that spleen colony-forming units (CFU-S) in DNA synthesis preferentially seeded bone marrow. In the present studies, the CFU-S content of marrow from irradiated, bone-marrow transfused mice was directly determined. Pretreatment of the transfused cells with cytocidal tritiated thymidine resulted in an insignificant diminution in CFU-S content when compared with nontritiated thymidine pretreatment, implying that there is no preferential seeding. The 125IUdR incorporation data have been reinterpreted as being a result of the proliferation of other progenitor cells present that have seeded the bone marrow.     

Liver repopulation with bone marrow derived cells improves the metabolic disorder in the Gunn rat

Background

Reversible ischaemia/reperfusion (I/R) liver injury has been used to induce engraftment and hepatic parenchymal differentiation of exogenous β2‐microglubulin⁻/Thy1⁺ bone marrow derived cells.

Aim

To test the ability of this method of hepatic parenchymal repopulation, theoretically applicable to clinical practice, to correct the metabolic disorder in a rat model of congenital hyperbilirubinaemia.

Methods and results

Analysis by confocal laser microscopy of fluorescence labelled cells and by immunohistochemistry for β2‐microglubulin, 72 hours after intraportal delivery, showed engraftment of infused cells in liver parenchyma of rats with I/R, but not in control animals with non‐injured liver. Transplantation of bone marrow derived cells obtained from GFP‐transgenic rats into Lewis rats resulted in the presence of up to 20% of GFP positive hepatocytes in I/R liver lobes after one month. The repopulation rate was proportional to the number of transplanted cells. Infusion of GFP negative bone marrow derived cells into GFP positive transgenic rats resulted in the appearance of GFP negative hepatocytes, suggesting that the main mechanism underlying parenchymal repopulation was differentiation rather than cell fusion. Transplantation of wild type bone marrow derived cells into hyperbilirubinaemic Gunn rats with deficient bilirubin conjugation after I/R damage resulted in 30% decrease in serum bilirubin, the appearance of bilirubin conjugates in bile, and the expression of normal UDP‐glucuronyltransferase enzyme evaluated by polymerase chain reaction.

Conclusions

I/R injury induced hepatic parenchymal engraftment and differentiation into hepatocyte‐like cells of bone marrow derived cells. Transplantation of bone marrow derived cells from non‐affected animals resulted in the partial correction of hyperbilirubinaemia in the Gunn rat.     

Protection of lethally irradiated mice with allogeneic fetal liver cells: influence of irradiation dose on immunologic reconstitution.

After lethal irradiation long-lived, immunologically vigorous C3Hf mice were produced by treatment with syngeneic fetal liver cells or syngeneic newborn or adult spleen cells. Treatment of lethally irradiated mice with syngeneic or allogeneic newborn thymus cells or allogeneic newborn or adult spleen cells regularly led to fatal secondary disease or graft-versus-host reactions. Treatment of the lethally irradiated mice with fetal liver cells regularly yielded long-lived, immunologically vigorous chimeras. The introduction of the fetal liver cells into the irradiated mice appeared to be followed by development of immunological tolerance of the donor cells. The findings suggest that T-cells at an early stage of differentiation are more susceptible to tolerance induction than are T-lymphocytes at later stages of differentiation. These investigations turned up a perplexing paradox which suggests that high doses of irradiation may injure the thymic stroma, rendering it less capable of supporting certain T-cell populations in the peripheral lymphoid tissue. Alternatively, the higher and not the lower dose of irradiation may have eliminated a host cell not readily derived from fetal liver precursors which represents an important helper cell in certain cell-mediated immune functions, e.g., graft-versus-host reactions, but which is not important in others, e.g., allograft rejections. The higher dose of lethal irradiation did not permit development or maintenance of a population of spleen cells that could initiate graft-versus-host reactions but did permit the development of a population of donor cells capable of achieving vigorous allograft rejection. These observations contribute to understanding of some of the persisting immunodeficiencies that are observed in man after fatal irradiation and bone marrow transplantation. These results should suggest better approaches to more effective cellular engineering for correction of immunodeficiency diseases and for treatment of immunodeficiency diseases and of leukemias and malignancies of man.     

Enhanced antibody affinity in sublethally irradiated mice and bone marrow chimeras.

Sublethally irradiated mice primed with dinitrophenyl (Dnp)-keyhole limpet hemocyanin immediately after irradiation or 30 days later and subsequently boosted with a second injection of antigen displayed a secondary response to Dnp characterized by antibody affinity greater than that in unirradiated controls. Also, in radiation chimeras primed with Dnp-keyhole limpet hemocyanin 120 days after syngeneic or allogeneic bone marrow transplantation the antibodies against Dnp produced after boosting were of higher affinity than the antibodies raised in normal mice. These findings are tentatively attributed to lack of suppressor thymus-derived lymphocytes (T cells) in sublethally irradiated mice and bone marrow chimeras, in which the enhanced ability to produce antibodies of high affinity may compensate for quantitative defects of the immune system.     

Liver repopulation trial using bone marrow cells in a retrorsine-induced chronic hepatocellular injury model

OBJECTIVES: The aim of this study was to determine the potential of bone marrow derived cells to participate in liver repopulation. In this model, the injected cells had a "selective growth advantage" compared to the native hepatocytes whose proliferation was blocked by retrorsine. METHODS: Total bone marrow cells were isolated from male Fisher 344 rats not deficient in dipeptidyl peptidase activity (F344, DPP IV+). The animals were given an injection of retrorsine and were divided in 2 groups: 1/group R (N=13): female F344 rats received 4.106 male cells at day 0 (labeled by chromosome Y). 2/group RH (N=19): Male F344 DPP IV- rats received 4.106 male DPP IV+ cells after hepatectomy at day 0 (labelled by DPP IV activity). RESULTS: Group R: no male cell was detected by PCR at day 14, 28, 56 and 84. Group RH: isolated DPP IV+ transplanted cells were observed at days 14 and 28 in the periportal areas. Later, these cells were no longer visible. Liver regeneration occurred by proliferation of small clusters of hepatocytes. CONCLUSIONS: In this experimental model the capacity of transplanted bone marrow cells to repopulate the liver was tested against the same capacity of native liver stem cells. Liver regeneration occurred via native liver cells seen as small hepatocytes. In this model the small hepatocytes may be considered as hepatic stem cells.