1. Thrombopoietin responsiveness reflects the number of doublings undergone by megakaryocyte progenitors
by Jean-Michel Paulus, Najet Debili, Frédéric Larbret, Jack Levin, and William Vainchenker
Blood
Volume 104(8):
October 15, 2004
©2004 by American Society of Hematology

2. Cumulative and derivative dose-response curves for the overall sample of clones derived from CD34+CD41+ progenitors.
Cumulative and derivative dose-response curves for the overall sample of clones derived from CD34+CD41+ progenitors. The cumulative dose-response curve is shown in the top panel, and derivative dose-response curve is shown in the bottom panel. Sorted progenitors were individually cultured for 5 days in serum-deprived medium with the indicated concentrations of TPO. Clones contained 1 to 16 megakaryocytes, corresponding to 0 to 4 progenitor doublings. Data points are the mean ± 1 SEM of quadruplicate determinations of the number of clones at day 5 per 96-well Terasaki plate. A gaussian cumulative equation gave the best fit to data among 29 functions based on transitional models (r2 = 0.99; P < .001). The bell-shaped bottom curve, the gaussian analytical derivative of the top curve, is the frequency distribution of TPO thresholds of individual progenitors. Counts obtained for wells with no TPO were assigned a concentration of 10–2 pg/mL to preclude entering the log of zero values in regressions. In this experiment, the mean absolute number of clones per 96-well plate which survived in absence of TPO was 4.44 ± 0.82 (ie, 6.05% of the total clones). The dotted vertical drop lines indicate TPO50 (ie, the TPO dose at which 50% of the plateau number of clones were counted). Results are from experiment number 1 and are representative of 3 experiments.
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

3. Dose-response curves for CD34+CD41+ progenitors of megakaryocyte clones of various sizes.
Dose-response curves for CD34+CD41+ progenitors of megakaryocyte clones of various sizes. In the 5 top panels, dose-response curves have been fitted to the absolute frequencies of clones with the indicated numbers of megakaryocytes. In the bottom panel, the curves have been normalized to a plateau frequency of 100% and superimposed to facilitate comparison. Data points are the mean ± 1 SEM of quadruplicate determinations of the number of clones at day 5 per 96-well Terasaki plate. R2 values for the gaussian cumulative model were 0.99 or more (P < .001) for all curves. Counts obtained for wells with no TPO were assigned a concentration of 10–2 pg/mL to preclude entering the log of zero values in regressions. For 1 MK clones, the mean absolute number of clones per 96-well plate that survived in absence of TPO was 4.28 ± 0.72 (ie, 5.92% of the total clones). The dotted drop lines indicate TPO50 (ie, the TPO dose at which 50% of the plateau number of clones were counted). Results are representative of 3 experiments.
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

4. Distributions of TPO thresholds for progenitors of 1-, 2-, 3-, and 4-cell MK clones and the total population.
Distributions of TPO thresholds for progenitors of 1-, 2-, 3-, and 4-cell MK clones and the total population. The bell-shaped curves are the gaussian analytical derivatives of the cumulative dose-response curves for 4 clone sizes and the total population. The peak of each curve indicates the median TPO threshold for the indicated class and corresponds to the respective TPO50 shown in Figure 2. The solid line represents the TPO threshold distribution for the total MK progenitor population. The 95% confidence intervals (95% CIs) of TPO50 were 2.68 to 9.04 pg/mL (total MK), 0.62 to 2.55 pg/mL (1 MK), 3.36 to 9.32 pg/mL (2 MKs), 13.9 to 21.7 pg/mL (3 MKs), and 32.3 to 35.7 pg/mL (4 MKs). Because the CIs did not overlap, TPO50 values differed significantly from each other at the 5% level. A representative experiment is shown.
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

5. Interrelationships between TPO responsiveness and NbD undergone by megakaryocyte progenitors in vivo and in vitro.
Interrelationships between TPO responsiveness and NbD undergone by megakaryocyte progenitors in vivo and in vitro. Relative responsiveness was defined as the absolute responsiveness (1/TPO50) at each data point divided by the responsiveness of the least responsive MK progenitor. NbD have been calculated as detailed in “Materials and methods.” Inverse correlations were obtained between responsiveness and numbers of residual doublings undergone in vitro (NbDr; left), and positive correlations were obtained between responsiveness and number of previous doublings undergone in vivo (NbDp; middle). As a result, a reciprocal relationship was found between NbDr and NbDp, indicating that a high number of previous doublings were associated with a low number of residual doublings. The converse was also true (right). The P values for the correlation coefficients were less than .005.
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

6. Analysis of correlation between TPO threshold of individual CD34+CD41+ progenitors and clone size achieved, expressed as NbDr.
Analysis of correlation between TPO threshold of individual CD34+CD41+ progenitors and clone size achieved, expressed as NbDr. Samples of 500 TPO thresholds (□) have been simulated so that the corresponding TPO50 and coefficients of variation were those in Table 1. The 4 horizontal groupings represent the sets of simulated values obtained for the 4 NbD size classes, from top to bottom: 8 to 15 MKs; 4 to 7 MKs; 2 to 3 MKs; and 1 MK. Solid and dashed lines indicate the regression line through the total set of points and through percentiles 2.5% and 97.5% of each of the 4 sets of simulated TPO threshold values, respectively. R2 is the squared Spearman rank order correlation (P < .001).
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

7. Size of megakaryocyte clonal progeny, TPO responsiveness, and Mpl expression recorded for CD42a+ and CD42– fractions of CD34+CD41+ megakaryocyte progenitors.
Size of megakaryocyte clonal progeny, TPO responsiveness, and Mpl expression recorded for CD42a+ and CD42– fractions of CD34+CD41+ megakaryocyte progenitors. (Left) Histogram columns indicate the percentages of progenitors of each fraction that generated clones of the sizes indicated. Data are the means of 3 duplicate experiments ± 1 SD. Distributions of percentages in the 2 fractions differed by chi-square analysis (P < .001). (Middle) The bell-shaped curves are TPO threshold distributions for the CD42a+ (left) and CD42a– (right) fractions. The means and 95% confidence intervals (95% CIs) of the corresponding TPO50 were 1.13 pg/mL (95% CI, pg/mL) and 2.75 pg/mL (95% CI, pg/mL), respectively. Because the CI did not overlap, TPO50 values differed significantly from each other at the 5% level. (Right) Flow cytometric analysis of Mpl expression on CD42a+ (right) and CD42a– (left) megakaryocyte progenitors. Mononuclear bone marrow cells were labeled with anti-Mpl followed by antimouse IgG-PE, washed with mouse serum, and further labeled with anti-CD34–PC5, anti-CD41–APC, and anti-CD42a–FITC. Median Mpl expression was 2.8-fold higher on CD42a+ than on CD42a– progenitors (P < .001).
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology

8. Relationship between dose-response curves and clone size distributions expressed as NbD.
Relationship between dose-response curves and clone size distributions expressed as NbD. (Left) Dose-response curves for clones with 0 or more, 1 or more, 2 or more, or 3 or more doublings. (Right) Normalized plateau values versus the corresponding NbDr.
Jean-Michel Paulus et al. Blood 2004;104:
©2004 by American Society of Hematology