The replication rate of human hematopoietic stem cells in vivo

Abstract

Hematopoietic stem cells (HSCs) replicate (self-renew) to create 2 daughter cells with capabilities equivalent to their parent, as well as differentiate, and thus can both maintain and restore blood cell production. Cell labeling with division-sensitive markers and competitive transplantation studies have been used to estimate the replication rate of murine HSCs in vivo. However, these methods are not feasible in humans and surrogate assays are required. In this report, we analyze the changing ratio with age of maternal/paternal X-chromosome phenotypes in blood cells from females and infer that human HSCs replicate on average once every 40 weeks (range, 25-50 weeks). We then confirm this estimate with 2 independent approaches, use the estimate to simulate human hematopoiesis, and show that the simulations accurately reproduce marrow transplantation data. Our simulations also provide evidence that the number of human HSCs increases from birth until adolescence and then plateaus, and that the ratio of contributing to quiescent HSCs in humans significantly differs from mouse. In addition, they suggest that human marrow failure, such as the marrow failure that occurs after umbilical cord blood transplantation and with aplastic anemia, results from insufficient numbers of early progenitor cells, and not the absence of HSCs.

Figure 1

A stochastic model of hematopoiesis. The hematopoietic stem cell reserve (compartment 1) contains HSCs. Each HSC can replicate, differentiate, or die. Mean rates of HSC replication, differentiation, and HSC death (apoptosis) are denoted λ, ν, and α, respectively. Once an HSC commits to differentiation, it heads a clone that contributes mature blood cells for a finite period of time and then exhausts (mean rate μ). As HSCs act based on their unique intrinsic and microenvironmental signals, we assume that these fates are independent (the Markovian assumption). R₀ and C₀ are the numbers of HSCs and contributing (ie, short-term repopulating cell) clones at birth, respectively. The steady-state number of HSCs is termed K.

Figure 2

Choosing human HSC replication rates for simulations. For each simulated person, we selected the replication rates for HSCs expressing the maternal X-chromosome and for HSCs expressing the paternal X-chromosome at random from a distribution with common mean λ. The SD of this distribution was defined such that the ratio between values 2 SDs above and below this mean would be equal to the ratio between λ_d and λ_G in cats (the horizontal line). If the variability in humans is smaller than assumed here, there should be a narrower range of possible values for λ. We suspect that this more stringent limit (which would result in a narrower range of ∼ 1 per 40 weeks) is indeed probable.

Figure 3

Estimating human replication rate through the analysis of changing maternal/paternal X-chromosome ratios with age in blood cells of females. (A) The percentage dominant allele for 1219 females 18 to 100 years of age. A value of 50% indicates no skewing, and 100% indicates complete skewing. (B-C) Means and percentages of persons with a skewed X-chromosome inactivation ratio (> 75% predominant allele), calculated using bin widths of 10 years. Means instead of the raw data were used to avoid overweighting more frequently occurring ages. (D-F) Percentage of persons with skewing in 3 sets of 100 simulations. For each simulation, we randomly drew replication rates for HSCs expressing the maternal X-chromosome and for HSCs expressing the paternal X-chromosome from a distribution (mean λ), with variance chosen similar to the range observed in Safari cats (Figure 2). The first simulations (D) use λ = 1 per 40 weeks and R₀ (the number of HSCs at birth) = 300. Because R₀ is uncertain, we considered this as a variable. The second set (E) uses λ = 1 per 20 weeks and R₀ = 700, and the third (F) uses λ = 1 per 55 weeks and R₀ = 300. To evaluate the appropriateness of R₀ and λ, we compared the intercepts and slopes (representing value at birth and increase per year of age, respectively) for each of the regressions to the intercepts and slopes of the regressions for 100 sets of 1219 simulations (“Determining acceptable values of λ and R₀”). Fitted regression lines are included. Parameter values in panel D yielded a similar y-intercept and slope to the observed data and were accepted; those in panels E and F did not and were rejected.