Medical Department Brookhaven National Laboratory Upton, New York 11973
* Research supported by the u. s. Energy Research and Development Administration.

The intent of this presentation is to present a notion that the in vitro bone marrow and blood culture system does not truly measure the dimensions of the committed stem cell pool. The measure of stem cell reserve -preferably the pluripotent and the committed stem cell pools is not a trivial, clinical problem. In the course of disease, chemotherapy, radiotherapy, one could much better guide management if there were a reliable reproducible method to assay for the stem cell reserve. In addition, a stem cell assay for man is critical to an understanding of normal hematopoiesis and its full characterization. At first look, the in vitro bone marrow culture combined with tritiated thymidine suicide appeared to be an answer to the problem. This can now be questioned. The diffusion chamber culture technique measures both pluripotent and committed stem cells but it also is capricious and poorly reproducible although some technical progress is being made (1,2). The concept of stem cells in hemopoiesis is several decades old. Maximow and Bloom (3) have expressed their thoughts clearly. They divided hemopoiesis into homoplastic and heteroplastic hemopoiesis. Homoplastic hemopoiesis is defined as the production of mature cells by young elements of the same type. They believe that under physiological conditions the needs of the adult organism are supplied generally by homoplastic hemopoiesis. When requirements for blood cells were increased, as after hemorrhage, during infection, during regeneration from injury, homoplastic hemopoiesis was insufficient and new erythroblasts and myelocytes then developed from a pluripotent stem cell and this they called heteroplastic hemopoiesis. Modern parlance divides the stem cells into the pluripotent and the committed stem cells (4, 5). Pluripotent hemopoietic stem cell can only be studied directly in the mouse. It is measured by the capability of hemopoietic cell suspensions to produce splenic colonies. The cell that produces the colony is called the CFU-S. The CFU-S derived from the bone marrow and spleen of the mouse has a capabilityof producing erythrocytic, granulocytic, and megakaryocytic lines, respectively (6). Becker et al. (7) have clearly shown that colonies are formed from a single CFU-S, demonstrating the clonal nature of the spleen colonies. Till et al. (8) in studying the growth rates of splenic colonies developed a stochastic model for stem cell proliferation, in which they feel that the birth process or replication of the stem cell and the death process ( differentiation into a cell no longer able to replicate the CFU-S) appear at random in the population of the colony forming cells proliferating in the splenic colony. Becker et al. (9) have determined the fraction of the CFU-S that are in DNA synthesis by treating suspensions of stem cells with high concentrations of ³HTdR to kill the cells in DNA synthesis. When the hemopoietic system is expanding a large fraction of the CFU-S (as much as 60-700/0) may be in DNA synthesis. In the steady state proliferation of the adult marrow and spleen the fraction of CFU-S in DNA synthesis is barely perceptible. Vassort et al. (10) have shown that the CFU-S has a ³HTdR suicide varying from 9-200/0 depending upon the strain of the adult mouse. Onlya fraction of the CFU-S produce splenic colonies upon transplantation. The fraction of these that produce splenic colonies is called the f-factor (11) ; this is roughly 0.17. There is considerable variation in the f-factor depending upon the period of time that the cell will circulate in the blood and many other factors of biological and statistical nature that may operate at any given time. As of the moment there is no way of detecting the pluripotent stem cell in man or mammals other than the mouse, and to a lesser extent, the rat. Pluripotent stem cells migrate through the blood. Two clear-cut experiments showed this years ago. Brecher and Cronkite (12) showed that when one member of a parabiotic pair is shielded while the other is receiving fatal irradiation, the irradiated one is protected from radiation lethality, thus showing the migration of the pluripotent hemopoietic stem cell from the nonirradiated into the irradiated twin. Swifl: et al. (13) showed the protection against radiation is conferred if one half the body only is exposed, followed in a matter of minutes by exposure of the remaining half and shielding of the previously exposed portion, thus showing migration of the pluripotent hemopoietic stem cell during this interval. Goodman and Hodgson (14) and Trobaugh and Lewis (15) showed that the PHSC circulate in the peripheral blood under normal steady state conditions. In mice the concentration of the CFU-S is 10-30 cells/mI (16). Their half-time in the blood, however, is reported to be only about 6 minutes (17). Accordingly, one can estimate in the 30 g mouse that the PHSC daily turnover rate is equal to:

If the human being has the same concentration in the blood as the mouse and similar rate of clearance (T 1/2, 6 minutes) the turnover rate in the standard man will be 1.7 x 107/ da y or 2 x 105 /kg/ da y .Thus, this line of logic leads one to believe that a number of stem cells pass through the blood per day equivalent to about 1/10 the number of PHSC estimated in the bone marrow. It is presumed that a dynamic equilibrium between blood and marrow exists. Of considerable interest is the flow of pluripotent stem cells through the alveolar capillaries. The number passing through the capillaries is equal to the concentration in the blood times the cardiac output. In man this amounts to -10 high 5/min. Thus, pluripotent hemopoietic stem cells are brought to within one to a very few micrometers of the gaseous external environment. This is of potential importance in pulmonary toxicology in considering the hazards of inhaled toxic gases such as ozone, the nitrogen oxides, etc. thus submitting pluripotent hemopoietic to potential in haled leukemogenic agents. Committed stem cell pool In man the abundance of the committed stem cells is generally considered to be measured by the in vitro bone marrow culture of colony forming units-culture (CFU-C). In man their abundance is about 0.1 to 11104 blood cells and 0.1 to 11103 bone marrow cells (18). Their thymidine suicide is of the order of 0.35, represent ing the fraction in DNA synthesis (19). A study of the human stem cell is very difficult. Human cells do not produce spleen colonies in mice. It has been suggested by several investigators that the colonies formed in culture of human peripheral blood or bone marrow by CFU-C are the committed stem cells since they produce differentiated cells ( erythrocytic, granulocytic, and macrophagic colonies). Since one cannot measure human stem cells satisfactorily, an approach to the study of the human stem cell pool is to start from the peripheral blood, where the turnover rates of erythrocytes and granulocytes are well known. From the sum of these turnover rates, the structure of human bone marrow (amplification from stem cell to nondividing stem cell), absolute cellularity, DNA synthesis time, and the fraction of cells in DNA synthesis, the minimum flux of committed stem cells into the erythrocytic and granulocytic differentiated pools of the marrow can be estimated. The data from which the calculations are made have been reviewed {20) and are summarized as follows :

1. Erythrocyte average life span = 120 days
2. Erythrocyte turnover rate (RTR) = 12 x 10 high 7/kg/hr (288 x 10 high 7/day)
3. Granulocyte life span (random loss) half time of 6.8 hrs. (21)
4. Granulocyte turnover rate (GTR) = 6.8 x 10 high 7/kg/hr
5. RTR + GTR = 18.8 x 10 high 7/kg/hr
6. Erythroid marrow cellularity (NE) = 536 x 10 high 7/kg (22,23 )
7. Granulocytis marrow (NG) = 1140 x 10 high 7/kg (22,23)
8. NE + NG = 1,676 x 10 high 7/kg
9. DNA synthesis time in human bone marrow is about 12 hours for erythrocy tic and granulocytic proliferating pool. (24 )
10. Amplification from the committed stem cell to the nondividing erythrocytic and granulocytic cell averages 16.

This is shown schematically in Figure 1.

STRUCTURE OF HUMAN BONE MARROW HEMOPOIETIC CELL PROLIFERATION

Fig. 1*: Schematic presentation of human bone possible quantitation of the stem cell compartments. Marrow cellularity from Donohue et al (23) Amplification from Cronkite and Vincent (20) DNA synthesis time from Stryckmans et al (24) It is assumed that 1 per 1000 bone marrow cells DNA synthesis. * This figure published by courtesy of Appleton-Century-Crofts.


If it is assumed that detection of the most immature erythrocytic and granulocytic precursor establishes the cytologic boundary between the committed stem cell pool, an estimate can then be made of flux of committed stem cells into the differentiated pool (red and white) by dividing the sum of the red cell turnover rate and granulocyte turnover rate in the blood by the average amplification of 16. The total turnover rate in the blood is 18.8 x 10 high 7/kg/hr. Thus with an amplification of 16 in the bone marrow the input of committed stem cells into the cytologically differentiated bone marrow pool is 1.17 x 1 0 high 7 /kg/hr . In a stem cell pool the birth rate (KB) must be twice the exit from the stem cell pool in order to maintain its steady state size. Thus birth rate is 2.35 x 10 high 7/kg/hr. By assuming that the DNA synthesis time of committed stem cells is the same as that which has been measured in the differentiated pool of 12 hours, one can then simply calculate the number of cells that must be in DNA synthesis from the product of the DNA synthesis time and the birth rate. Thus, there are 28 x 10 high 7/kg in DNA synthesis. If one uses the ³HTdR suicide of 0.35 as an estimate of the fraction of the committed stem cells that are in DNA synthesis, then the quotient of the number of cells in DNA synthesis by the fraction in DNA synthesis gives the absolute number or 80 x 10 high 7/kg. Referring to Figure lone sees that the total number of differentiated erythrocytic and granulocytic cells is 1676 x 10 high 7/kg. The ratio of the committed stem cells to the differentiated stem cells is therefore 1 :21. Thus 4.8 0/0 of the total marrow is committed stem cells and one can on a priori grounds say that the in vitro marrow culture system does not measure the committed stem cell pool. One can ask whether this approach is valid. Does the thymidine suicide for the colony forming unit in culture apply to all the committed stem cell pool? If the fraction in D N A synthesis is smaller, the fraction of the total marrow occupied by committed stem cells becomes larger. If the time for DNA synthesis is overestimated, the number is overestimated. If the DNA synthesis time is really less than the measured time for differentiated marrow cells, the estimate of the number in DNA synthesis is too high. Even reducing time for DNA synthesis to a probably unrealistic low value of 1 hour for man reduces the number of committed stem cells to 6.6 x 1 0 high 7 /kg or still 1 in 300 marrow cells are committed stem cells. The above line of logic and assumption leads to a ratio of committed stem cells to total marrow cells that is completely incompatible with the in vitro bone marrow culture of 5/10 high 4 colony forming cells in human bone marrow. It can be argued that the culture technique does not have sufficient stimulatory power to switch all the colony stimulating cells into proliferation. This is an attractive thought since the ratio has tended to increase as more potent sources of colony stimulating factor are developed and culture techniques are thus "souped up." Let us take another line of argument and accept that the in vitro colony forming techniques do, in fact, accurately measure the number of committed stem cells in the bone marrow. If one then goes through the arithmetic as before and calculate on the basis of the DNA synthesis time of 12 hours, thymidine suicide of 0.35 and an absolute marrow cellularity of 1800 x 10 high 7/kg, it is simple to show that the flux of the committed stem cells into the differentiated pool is such that an amplification of 520 would be required. Such an amplification would require about 9 serial mitoses in the proliferating granulocytic pool alone. In the granulocytic series it is reasonably well established that the time from the myeloblast to the myelocyte is 130 hours (20). If there is equal time for each successive multiplicative cell cycle there would be a generation time of 14 hours. The mitotic time is 0.75 hours (25). The mitotic index would be about 5 0/0 or 5 times that observed by Killmann et al. (26 ). With a DNA synthesis time of 12 hours (24) the fraction of granulocytic proliferating cells in DNA synthesis would be 0.85 compared to the observations in man of 0.15-0.30 (26) and become incompatible with the number of myelocytes known to have a diploid DNA content. From these lines of reasoning and calculations based on experimentally determined data the notion of 9 serial mitoses in the granulocytic proliferating pool must be rejected along with the notion that the present in vitra methods of culturing bone marrow determines the fraction of committed stem cells in human bone marrow. In addition, one can argue that knowledge on the cell turnover rate in the peripheral blood is incorrect. However, this is unlikely. Other arguments are that calculations based on average values are misleading and that there is a small fraction of stem cells that are dividing very rapidly or that the proliferation rate in the differentiated pool is grossly underestimated by averages and a small fraction of rapidly dividing cells will enable on one to "balance the books". The answer is not evident and for the moment this autohor feels compelled to question whether the in vitra culture bone marrow methods really estimate the fraction of the bone marrow that consists of committed stem cells.

References

1. Boyum, A., Carsten, A. L., Laerum, 0. D. and Cronkite, E. P. Kinetics of celJ proliferation of murine bone marrow cells cultured in diffusion chambers: effect of hypoxia, bleeding, erythropoietin injections, polycythemia, and irradiation of the host. Blood 40:174, 1972.

2. Cronkite, E. P., Boecker, W., Carsten, A. L., Chikkappa, G., Joel, D., Laissue, J. and Öhl, S. The use of diffusion chamber cultures in the study of normal and leukemic cell proliferation in man. In Robinson, W. A. (ed.) Hemopoiesis in Culture, DHEW Publ. No. (NIH) 74-205, Bethesda, 1973, P. 185.

3. Maximow, A. A. and Bloom, W. (eds) Text Book of Histology, Philadelphia, W. B. Saunders, 1935.

4. Stohlman, F., Jr. The Kinetics of Cellular Proliferation, Grune and Stratton, New York, 1959.

5. Stohlman, F., Jr. (ed) Symposium on Hemopoietic Cellular Proliferation. St Elizabeth's Hospital Centennial 1869-1969, Boston, Massachusetts, Nov. 5-6 1969, Grune and Stratton, New York, 1970.

6. Till, J. E. and McCulloch, E. A. A direct measurement of the radiation sen sitivity of normal mouse bone marrow cells. Radiat. Res. 14:213, 1961.

7. Becker, A. J ., McCulloch, E. A. and Till, J. E. Cystological demonstration of the clonal nature of spleen colonies derived from transplanted mouse marrow cells. Nature 197:452, 1963.

8. Till, J. E., McCulloch, E. A. and Siminovitch, L. A. stochastic model of stem cell proliferation, based on the growth of spleen colony-forming cells. Proc. Nat. Acad. Sci. 51 :29, 1964.

9. Becker, A. J., McCulloch, E. A., Siminovitch, L. and Till, J. E. The effect of differing demands for blood cell production on DNA synthesis by hemopoietic colony-forming cells of mice. Blood 26:296, 1965.

10. Vassort, F., Winterholer, M., Frindel, E. and Tubiana, M. Kinetic parameters of bone marrow stem cells using in viva suicide by tritiated thymidine or by hydroxyurea. Blood 41 :789, 1973.

11. Till, J. E. and McCulloch, E. A. The "f-factor" of the spleen colony assay for hemopoietic stem cells. Series Hematology 5:15-21,1972.

12. Brecher, G. and Cronkite, E. P. Postradiation parabiosis in survival in rats. Proc. Soc. Exper. BioI. Med. 77:292, 1951.

13. Swift, M. N., Taketa, F. T. and Bond, V. P. Regionally fractionated x-irradia tion equivalent in dose to total body exposure. Rad. Res. 1 :241, 1954.

14. Goodman, J. E., Hodgson, G. S. Evidence for stem cells in the peripheral blood of mice. Blood 19:702, 1962.

15. Trobaugh, F. E., Lewis, J. P., Jr. Repopulating potential of blood and marrow. J. Clin. Invest. 43:1306,1964.

16. Hellman, S. and Grate, H. E. Kinetics of circulating haemopoietic colonyforming units in the mouse. In Doyle, E. (ed.) Effects of Radiation on Cellular Proliferation and Differentiation. Vienna, International Atomic Energy Agency, 1968(a), p. 187.

17. Hodgson, G., Guzman, E. and Herrera, C. Characterization of the stem cell population of phenylhydrazine-treated rodents In: Doyle, E. (ed.) Effects of Radiation on Cellular Proliferation and Differentiation. Vienna, International Atomic Energy Agency, 1968, p. 163.

18. Senn, J. S. and McCulloch, E. A. Radiation sensitivity of human bone marrow cells measured by a cell culture method. Blood 35 :56, 1970.

19. Moore, M. A. S. and Williams, N. Functional, morphologic, and kinetic analysis of the granulocyte-macrophage progenitor cell. In Robinson, W. A. (ed) Hemopoiesis in Culture, DHEW Publication No. (NIH) 740205, Bethesda, 1973, p 17.

20. Cronkite, E. P. and Vincent, P. C. Granulocytopoiesis. Ser. Haemat. VoI. II, 4, 1969, 3-43.

21. Cartwright, G. E., Athens, J. W. and Wintrobe, M. M. The kinetics of granu lopoiesis in normal man. Blood 24 :780, 1964.

22. Donohue, D. M., Reiff, R. H., Hanson, M. L., Betson, Y and Finch, C. A. Quantitative measurements of the erythrocytic and granulocytic cells of the marrow and blood. J. Clin. Invest. 37:1571,1958.

23. Donuhue, D. M., Gabrio, B. W. and Finch, C. A. Quantitative measurement of hematopoietic cells of the marrow. J. Clin. Invest. 37:1564,1958.

24. Stryckmans, P., Cronkite, E. P., Fliedner, T. M. and Tamos, J. DNA synthesis time of erythropoietic and granulopoietic cells in human beings. Nature 211:717, 1966.

25. Odartchenko, N., Cottier, H., Feinendegen, L. E. and Bond, V. P. Evaluation of mitotic time in viva using tritiated thymidine as a cell marker: Successive labeling with time of separate mitotic phases. Exp. Cell. Res. 35 :402, 1964.

26. Killmann, S. A., Cronkite, E P., Fliedner, T. M. and Bond, V. P. Mitotic index of human bone marrow cells I. Number and cytologic distribution of mitoses. Blood 19:743,1962.