Genesis of clone size heterogeneity in megakaryocytic and other hemopoietic colonies  Jean-Michel Paulus, Jack Levin, Najet Debili, Adelin Albert, William Vainchenker  Experimental Hematology  Volume 29, Issue 11, Pages 1256-1269 (November 2001) DOI: 10.1016/S0301-472X(01)00728-7

Figure 1 Graphic computer map of pure megakaryocyte clones grown in 35-mm Petri dish plated with 30,000 bone marrow cells in 1 mL of plasma clot medium supplemented with WEHI-CM + EPO (3 U/mL). Cultured cells were stained for DNA with the Feulgen reaction [18] following AChE staining to exclude AChE-positive cytoplasmic debris. Megakaryocyte identification was made at 250× magnification to ensure that even small, single megakaryocytes were detected. Each point represents one megakaryocyte; its location was recorded by an operator using x-y sensors attached to the microscope stage and connected to a graphic computer. The latter delineated clones based on the criterion [28] that a megakaryocyte belonged to a clone if it had not wandered >1,000 μm from the nearest colony member. Colonies, i.e., clones with ≥3 megakaryocytes, were delineated by convex polygons. This formulation was introduced to display all cells of a colony within a simple geometric figure. For each clone, the NbD undergone by its progenitor was obtained as the log base 2 of the number of AchE-positive, Feulgen-positive cells. Four megakaryocyte colonies and two single megakaryocytes are shown Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)

Figure 2 Steps involved in drawing distributions of the number of doublings (NbD) undergone by megakaryocyte (MKC) clones. Delineation was made as shown in Figure 1. (A) Frequency histogram of clone sizes. Upper abscissa indicates the number of AChE-positive cells per clone from which the NbD indicated on the lower abscissa was derived. (B) Corresponding cumulative distribution recording the number of clones that have undergone at least 0, 1,…8 doublings. (C) Graph obtained by plotting the cumulative distributions of panel B on semilog coordinates and drawing the regression line as detailed in Figure 3. The fraction of clones that arrest proliferation and commit to polyploidization during one doubling, termed probability of proliferation arrest, is indicated on the exponential slope Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)

Figure 3 Semilogarithmic plots of cumulative distributions of population doublings (number of clones with at least the indicated number of doublings) undergone by mouse CFU-MK (first three rows) or BFU-E/MK (fourth row) clones in cultures incubated with various stimulators or combinations of stimulators. The plots were generated as shown in Figures 1 and 2. The upper abscissa indicates the number of AChE-positive cells per colony from which the number of doublings indicated on the lower abscissa was derived. The logarithms of cumulative frequencies were entered into a multiple regression program that fitted one straight line or two straight lines with unknown slopes and point of intersection to the data [192]. As judged by the r2 value, generation of a continuous curvature by a second- or third-degree polynomial did not give a better fit than the one- or two-line model. The fraction of CFU-MK clones that arrest proliferation and commit to polyploidization during one doubling, termed probability of proliferation arrest, is indicated on the first exponential slope. Note that on each graph, the point that represents the cumulated frequency of entities containing at least one megakaryocyte lies approximately on the regression line. The figure describes the results of three identical experiments in which the data from two or three dishes were pooled for each condition. The r2 values were higher than 0.986 in all cases. EPO = erythropoietin; PW = pokeweed-spleen conditioned medium; WEHI = WEHI-conditioned medium Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)

Figure 4 Doubling distributions of single human CD34+CD41+ progenitors sorted in individual wells and cultured for 6 days in serum free-medium with thrombopoietin (TPO) added as described [39]. Under the conditions used, the progeny of CD34+CD41+ progenitors consist solely of CD41+ cells. Data were obtained from cultures of adult bone marrow (top panels) or cord blood (bottom panels) Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)

Figure 5 Redrawing of the NbD distribution shown in Figure 2 to illustrate that single megakaryocyte (SM) clones, two-megakaryocyte clones, and colonies with ≥3 megakaryocytes may belong to the same exponential continuum despite the evidence for intrinsic differences between these progenitor types. CFU-MK is defined as pure megakaryocyte colonies containing 3 to 41 cells and BFU-MK as pure megakaryocyte colonies containing ≥42 cells [23], corresponding to 5.4 doublings Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)

Figure 6 Model illustrating the genesis of megakaryocyte clone size heterogeneity. A given stem cell can generate any paired combination of progenitor cells with different lineage potential, including various monopotent (single megakaryocyte [SM], CFU-MK, BFU-MK), bipotent (BFU-E/MK), and multipotent (CFU-GEMMK) megakaryocyte progenitors. Abcissae refer to the number of doublings (NbD) undergone by individual clones prior to polyploidization. Monopotent and bipotent progenitors (for instance, CFU-MK and BFU-E/MK) and presumably multipotent progenitors generate NbD distributions with different slopes, reflecting the fact that progressive restriction in lineage and proliferative potential is associated with increasing polyploidization probability. Differential shading as a function of responsiveness to polyploidization inducers is meant to reflect the deterministic concept that the most responsive progenitors (black shading) commit to polyploidization after fewer doublings than the least responsive ones (white shading). In a stochastic view, all progenitors would have the same shade because clone size heterogeneity would be generated by a random commitment event independent of progenitor properties, arresting proliferation asynchronously (i.e., after a different number of doublings) in different clones Experimental Hematology 2001 29, 1256-1269DOI: (10.1016/S0301-472X(01)00728-7)