Of mice and men … and elephants

The debate about hematopoietic stem cell numbers and productive capacity is relevant to many areas of clinical and experimental hematology. Recently, Abkowitz et al1 reported that the total number of stem cells in mammals is conserved so that mice, cats, humans, and possibly elephants have the same-sized stem cell pools. They emphasized that, since humans live longer than mice and need to produce more blood cells, human stem cells must be capable of producing many more blood cells than murine stem cells. A corollary of this is that the frequency of stem cells in human bone marrow will be lower than the frequency in murine bone marrow. Since we have made similar measurements in mice and calculations in humans,2,3 we were interested in determining whether our data are consistent with those of Abkowitz et al1 and whether the calculated numbers of stem cells in humans can account for the daily requirement to replace mature blood cells lost through senescence.

Table 1 compares earlier data, obtained using ⁵⁹Fe ferrokinetics to estimate marrow cellularity and the spleen colony (CFU-S) assay4 to measure hematopoietic stem cells in 3 strains of mice,2 with the values reported by Abkowitz et al.1 Our data indicate slightly greater values than Abkowitz et al,1 but they are of the same order of magnitude. Table 1 also shows data for the derived frequency of human hematopoietic stem cells, assuming that the human hematopoietic stem cell population is the same size as that of the mouse, which again is consistent with that reported by Abkowitz et al.1 

Adult humans need to produce 1 × 10¹¹ neutrophils and 2 × 10¹¹ red blood cells per day throughout life. Assuming a lifespan of 80 years, this amounts to a total of 80 × 365 × 3 × 10¹¹ (8.8 × 10¹⁵) for these 2 lineages alone, which must be generated by 16 800 (Abkowitz et al1) or 81 000 (our data) stem cells. Assuming steady-state stem cell kinetics5 and 100 stem cell replications per lifetime, stem cell division and self-renewal with a probability of 0.5 will result in 1.68 × 10⁶ to 8.1 × 10⁶ new stem cells available for differentiation and, ultimately, mature blood cell formation. Consequently, each of these stem cells must be responsible for producing 1.1 × 10⁹ to 5.2 × 10⁹mature cells. It is noteworthy therefore that Brummendorf et al6 reported that single stem cells generate, on average, nearly 10⁹ CD34⁺CD38⁻ cells in vitro and the greatest number observed was more than 10¹². The corresponding stem cell frequency would be 1 in 5 × 10⁸ to 10 × 10⁸. As noted by Abkowitz et al,1 these frequencies are much lower than the frequencies of candidate human stem cells measured in the NOD/SCID transplantation model.7 Clearly, if stem cells did not self-renew, they would need to be capable of producing many more cells to sustain hematopoietic cell production throughout life. If this were the case, each of the 16 800 to 81 000 stem cells would be required to produce on average 1.1 × 10¹¹ to 5.2 × 10¹¹ mature cells and their frequencies would be commensurately 2 orders of magnitude lower. Moreover, at such low frequencies a total marrow cellularity of 1.5 × 10¹²could accommodate only a very few stem cells.

Insufficient data are available to estimate the size of the stem cell pool and kinetic parameters in elephants accurately (Table 2). But simple calculations of the possible frequency and productive capacity of elephantine stem cells can be made. The data in Table 1 and Abkowitz et al1 indicate that mice, cats, and humans have on the order of 1 × 10¹⁰ to 2 × 10¹⁰bone marrow cells per kilogram of body weight. If the same is true of elephants, a large African elephant of 7 500 kg (Table 2) could have as many as 1.5 × 10¹⁴ bone marrow cells. The corresponding frequencies of stem cells in such animals would be 1 to 8 per 10¹⁰ if all species have the same number.1Furthermore, elephants have many more blood cells than mice or humans (Table 1), up to 10 × 10¹². If red and white cell kinetics are similar in humans and elephants, who live about the same length of time, a large African elephant will need to produce 17.5 × 10²⁰ cells in a lifetime. This number corresponds to 2 × 10¹⁴ to 10 × 10¹⁴ cells per stem cell.

In summary, the experimental and theoretical estimates of stem cell pool size in humans are not inconsistent with lifelong requirements for blood cell production sustained by a self-renewing stem cell population. Nor do the values we have derived appear to be inconsistent with the idea that mice, cats, humans, and elephants could have similar numbers of stem cells.1