Targeting Tumor Cell Growth Kinetics
Malignant hematopoietic cells proliferate more and differentiate less than their normal counterparts. The cell cycle consists of a series of stages through which normal and neoplastic cells proceed during the course of cellular replication (shown schematically in Fig. 1). The cell cycle is divided into G 1 (pre-DNA synthetic phase), S phase (in which DNA replication takes place), G 2 (post-DNA synthetic phase), and mitosis (M), during which chromosomal division and segregation occur. In addition, nonproliferating, resting cells reside in G 0, a phase that may theoretically last for an indefinite period. Such cells remain in G 0 until they are induced to cycle (at G 1) by specific triggers (e.g., hematopoietic growth factors). The growth fraction of a tumor represents the percentage of cycling cells relative to the total cell population. The generation time represents the time required for a cell to proceed through a single cell cycle (generally 24 to 36 hours for hematopoietic tissues). Surprisingly, in the case of acute myeloid leukemia (AML), the generation time of leukemic blasts is not shorter than that of normal hematopoietic progenitors and may be longer. The proliferative advantage of malignant hematopoietic cells (and of many nonhematopoietic tumors) stems, at least in part, from the fact that a higher percentage of cells are in cycle at any one point in time (i.e., the growth fraction is higher). The doubling time rep resents the period required for a tumor to double in mass and is, in general, inversely related to the tumor’s growth fraction. Tumor doubling times range from longer than 120 days in the case of some solid tumors (e.g., lung and colon) to less than 2 weeks (in some leukemias and lymphomas). Tumors with high growth fractions and short doubling times tend to be more sensitive to chemotherapy than slowly growing neoplasms with low growth fractions and long doubling times.
Fig1. Progression through the cell cycle is controlled through complex interactions among cyclins, cyclin-dependent kinases, and cyclin-dependent kinase inhibitors. Progression across the G 1 S interface and through S phase involves the E2F transcription factor, which activates numerous enzymes (e.g., thymidylate synthase, dihydrofolate reductase) required for DNA replication. The pRb in its dephosphorylated state binds to and inactivates E2F in con junction with DP proteins, thereby inhibiting S-phase progression. Conversely, phosphorylation of pRb antagonizes binding to E2F allowing S-phase events to proceed. Phosphorylation of pRb results from activation of (1) CDK2:cyclin A/E and (2) CDK4/6:cyclin D complexes. The former complexes are inhib ited by the CDK1s p21, p27 (and p57), and the latter are inhibited by the low-molecular-weight CDK inhibitors (p14–18) but also by p21 and p27. The complex formed by CDK1 (p34 cdc2) and cyclins A and B regulates G 2 M pro gression and is inhibited by the “universal” CDK inhibitor, p21. Moreover, its phosphorylation status, which plays a major role in determining activity, is regu lated by the phosphatase cdc25. Proteins such as pRb, E2F, p21, and p27 can influence the response of cells to chemotherapeutic agents by controlling cell cycle progression and possibly via cell cycle–unrelated actions. CDK, Cyclin dependent kinase; CDK2, cyclin-dependent kinase-2; DHFR, dihydrofolate reductase; pRb, retinoblastoma protein; TS, Thymidine synthase.
Cell cycle progression is governed by a complex network of proteins consisting of cyclins, cyclin-dependent kinases (CDKs), and CDK inhibitors. Progression through S phase is regulated primarily by CDK2 in association with cyclins A and E; progression through G 2 M is regulated by CDK1 (p34cdc2) and cyclins A and B; and progression through G 1 involves CDKs 4 to 6 in conjunction with cyclin D. CDK inhibitors fall into two major categories: the low-molecular weight inhibitors (pINK14, -15, -16, -17, and -18), which primarily inhibit cyclin D (and to some extent, CDK2) complexes, and the higher molecular-weight inhibitors, p21, p27, and p57, which are more universal in their actions and inhibit most or all CDKs. Signals for the progression of cells through G 1 S are essential for maintenance of the neoplastic phenotype. In the commonly accepted model of G1 S progression, inactivation of the retinoblastoma protein (pRb) is required. In quiescent cells, pRb is in an active dephosphorylated state and bound to the transcription factor E2F. Phosphorylation of pRb by CDK4, CDK6, and CDK2 leads to release of E2F, which is then free to activate diverse genes essential for S-phase progression, such as MYC (also known as c-Myc), TS (thymidylate synthetase), and DHFR (dihydrofolate reductase). Conversely, induction of CDK inhibitors (e.g., by transforming growth factor- β [TGF- β ] or differentiation-inducing agents) results in inactivation of CDK4, CDK6, and CDK2, dephosphorylation of pRb, inactivation of E2F, and inhibition of the progression through S phase. Aberrant expression of cyclins and CDK inhibitors is commonly encountered in hematopoietic malignancies.
In addition to growth control, cell cycle proteins are intimately involved in the regulation of programmed cell death (apoptosis) and checkpoint control mechanisms. Consequently, cell cycle regulatory proteins can exert a major influence on the response of neoplastic cells to cytotoxic agents. For example, when cells undergo DNA damage, they may arrest in G 2 M or G 1 , during which repair occurs, or if the damage is too severe, the cells undergo apoptosis. In particular, the tumor suppressor gene TP53 and its downstream inducible target p21 have been implicated in the G 1 arrest process after genotoxic insult. Dysregulation of various cell cycle regulatory proteins can have a major impact on the sensitivity of neoplastic cells to chemotherapeutic agents. Loss of the TP53 gene renders cells resistant to diverse chemotherapeutic agents, presumably by preventing cells from undergoing repair in G 1 and thereby inhibiting the cell death processes and allowing DNA dam age to accumulate, culminating in cellular transformation. Conversely, transfection of P53-negative cells with wild-type P53 restores responsiveness to most drugs. Dysregulation of the CDK inhibitors p21 (a downstream target of P53) and p27 increases the sensitivity of neo plastic cells to various cytotoxic agents, possibly by uncoupling S-phase progression and mitosis. After DNA damage, checkpoints block the cell cycle, but loss of the CDK inhibitors p21 or p27 prevents cells from arresting in G 1 and cells die during G 2 M. Mutations in the E2F protein have been shown to lengthen S phase and increase the sensitivity of malignant cells to S-phase–specific agents. Furthermore, cells lacking functional pRb have been shown to be significantly less sensitive to the actions of antimetabolites, including methotrexate.
In vivo the growth of tumors is limited by various factors such as vascular supply, nutritional requirements, and possibly physical restraints. Consequently, the rate of tumor growth declines as the number of cells increases. To the extent that tumor-doubling times are inversely correlated with drug responsiveness, large, late-stage tumors are less susceptible to cytotoxic drugs than early-stage tumors, with higher growth fractions. Most chemotherapeutic drugs kill by first-order kinetics. The implication of this phenomenon is that it requires the same drug dose to reduce the number of tumor cells from 104 to 101 cells as it does to reduce the tumor burden from 1010 to 107 cells.
Hematopoietic malignancy-initiating or stem cells, such as leukemia stem cells, appear to explain resistance and treatment failure. These cells express high levels of hematopoietic stem cell proteins and markers and are resistant to cell cycle–specific agents because of an increase in quiescent cell populations. They overlap with nor mal hematopoietic stem cells (HSCs) and there appears to be a set of HSC s that contain preleukemic-promoting mutations that both predispose and may be sufficient for conversion to leukemic stem cells. The proteins expressed by these cells have become targets of therapy, including alterations in DNA damage, quiescent cell cycle factors such as the thrombopoietin receptor, MPL , stem cell proliferation signals such as NOTCH and WNT protein families, and niche occupancy proteins such as KIT .
Cytotoxic agents may be divided into several categories with respect to their effects on the cell cycle or the cell cycle specificity of their actions, or both.
1. Noncycle-active drugs kill both cycling and noncycling cells in all phases of the cell cycle. Examples include steroids and antitumor antibiotics (except bleomycin).
2. Cycle-active, nonphase-specific drugs are more active against cycling cells and can kill cells in each phase of the cell cycle. However, such drugs may preferentially kill cells in a particular phase of the cell cycle. Examples include alkylating agents, cisplatin, and 5-fluorouracil (5-FU).
3. Cycle-active, phase-specific drugs primarily kill cells in a specific phase of the cell cycle. Examples include most antimetabolites, which are active against cells engaged in DNA synthesis (S-phase cells), and microtubule-active drugs (e.g., vinca alkaloids, taxanes), which kill cells in G 2 M.
An example of a cytokinetically rational approach to chemo therapy involves the combination of a noncycle-active agent (e.g., daunorubicin) with a cycle- and a phase-specific agent (e.g., ara-C, fludarabine, decitabine, gemcitabine, clofarabine, and nelarabine). From a theoretical standpoint, administration of a noncycle-active agent may reduce tumor mass, leading in turn to an increase in the growth fraction caused by recruitment of cells into cycle. Such cells would then be more susceptible to a cycle- and phase-specific agent, particularly one administered over a prolonged interval. In the case of hematopoietic malignancies, attempts have been made to recruit neo plastic cells into the more susceptible S phase of the cell cycle through the use of hematopoietic growth factors. The success of such a strategy has been limited because of several factors, including the inability of growth factors to increase the S-phase fraction significantly, the lack of selectivity of this strategy, and the theoretical possibility that growth factors may protect neoplastic cells from apoptosis.
Unfortunately, cytokinetic differences between normal and neo plastic tissues have been difficult to exploit. Both normal hematopoietic stem cells and hematologic malignant stem cells have a low proportion of cells in G 1 . However, prolonged dosage schedules can provoke these malignant cells into cell cycle and may explain their efficacy. Consequently, rapidly dividing normal tissues such as gastro intestinal epithelium and normal hematopoietic progenitors tend to be very sensitive to most chemotherapeutic agents. As a result, mucositis and myelosuppression represent frequent dose-limiting toxicities for many cytotoxic drugs.
Newer agents that target cell cycle proteins are often first utilized in patients with hematologic malignancies. As noted in the section on CDK and CHK1 inhibitors, these agents can have potent cytotoxic effects on dividing cells.
