Hematopoiesis and its disorders: a systems biology approach

Abstract

Scientists have traditionally studied complex biologic systems by reducing them to simple building blocks. Genome sequencing, high-throughput screening, and proteomics have, however, generated large datasets, revealing a high level of complexity in components and interactions. Systems biology embraces this complexity with a combination of mathematical, engineering, and computational tools for constructing and validating models of biologic phenomena. The validity of mathematical modeling in hematopoiesis was established early by the pioneering work of Till and McCulloch. In reviewing more recent papers, we highlight deterministic, stochastic, statistical, and network-based models that have been used to better understand a range of topics in hematopoiesis, including blood cell production, the periodicity of cyclical neutropenia, stem cell production in response to cytokine administration, and the emergence of imatinib resistance in chronic myeloid leukemia. Future advances require technologic improvements in computing power, imaging, and proteomics as well as greater collaboration between experimentalists and modelers. Altogether, systems biology will improve our understanding of normal and abnormal hematopoiesis, better define stem cells and their daughter cells, and potentially lead to more effective therapies.

Figure 1

Graph of protein-protein interaction. (A) The network of the 1548 proteins of the yeast proteome and their interactions. Proteins with different functions are represented by different colors and include lipid metabolism (yellow), cell structure (green), membrane fusion(blue), chromatin structure(gray), and cytokinesis (red). (B) Because of node and edge density, a subset of the network is magnified to reveal the complexity of components and interactions. (Figure from Uetz et al reprinted with permission from Nature.)

Figure 2

Feedback and feedforward loops in biologic systems. (A) Positive feedback loops result when (x) leads to the production of (y), which up-regulates the level of (x). (B) Negative feedback loops result when (x) leads to the production of (y) which down-regulates the level of (x). (C) Feed-forward loops are often found in biochemical or genetic regulatory networks. An example involves a general (x) and a specific (y) gene activator. The general activator sends an activation signal to the target gene (z) and the specific activator (y). If the signal is sustained, (y) becomes activated, permitting it to reinforce an activation signal for (z), completing the gene activation process. Once the signal from (x) stops, (z) ceases to be activated. The 2-step activation process is ideal for noisy systems where random fluctuations in the signal from (x) are less likely to cause activation of the gene (z) because of the signaling delay via (y).

Figure 3

Deterministic and stochastic processes in hematopoiesis. (A) In the deterministic model, hematopoietic growth factors such as Epo, G-CSF, macrophage colony-stimulating factor (M-CSF) instruct differentiation of blood stem cells. (B) In the stochastic model, these growth factors as well as others such as interleukin-3 (IL-3) or stem cell factor (SCF) promote survival, allowing stem cells to differentiate. (C) Results of 6-generation stochastic cell-fate simulations. A pluripotent stem cell can either divide into 2 stem cells or differentiate, losing its proliferative capacity. Simulations were run using a random number generator. The probability of birth was 60%, death 40%, the decision for each cell was performed by picking a random number 0 to 9 with 7 of these numbers (0-5) resulting in symmetric division and 3 (6-9) resulting in stem cell loss due to differentiation. Starting with 1 cell and drawing random numbers for each stem cell present in each generation leads to very different final populations, illustrating the randomness a stochastic model can incorporate. (Figure modified from Till et al with permission from Dr Till.)

Figure 4

Cyclical hematopoiesis. A deterministic model for cyclic erythropoiesis was constructed from experimental data derived from a rabbit with auto-immune hemolytic anemia. The model describes periodic changes in Epo levels as a result of feedback loop with erythrocyte mass. The solid lines represent model predictions and the dashed line represents experimentally obtained erythrocyte counts. (Reprinted from Mahaffy et al with permission from Elsevier.)

Figure 5

Population dynamics model for G-CSF mobilized peripheral stem cells. (A) A compartment model for stem cell mobilization. This model describes the G-CSF concentration (G) in the microenvironment and 4 different populations of cells in the bone marrow and peripheral blood: stem cells (S), peripheral blood cells (B), white blood cells (W), and platelets (P). Other parameters that the model incorporates are as follows: a_G, a_P: the production rates of G-CSF and platelets. a_S(G), a_w(G): The production rates of stem cells and white blood cells as functions of G-CSF concentration. f_1, f_2, f₃: Rates at which stem cells, white blood cells, and platelet concentrations down-regulate the stem cell production rate. f₄: Rate at which white blood cell concentration down-regulates G-CSF production. d_G, d_B, d_W, d_P: The rates of destruction of G-CSF, peripheral blood progenitor cells, white blood cells, and platelets. T_L: The population of progenitor cells in the bone marrow. T_SB: The population of stem cells passing from the blood to the bone marrow. T_BS: The population of stem cells passing from the bone marrow to the blood r(G). (B) Stimulation of G-CSF treatment every day over a 9-day period shows a rapid increase in white blood cell count (solid line) with each treatment (dotted line). After 5 days, further treatments have no effect on total white blood cell count. Once treatment has stopped (day 10), white blood cell counts slowly decline, returning to pretreatment levels. (Figure from Obeyesekere et al reprinted with permission from Wiley-Blackwell.)