As noted previously (Figure 7.14), the process of apoptosis involves numerous potential induc- ing agents as well as a number of pathways within cells capable of modulating the apoptotic response to various agents. Primary among these are receptor-mediated events, many of which involve tumor necrosis factor alpha (TNF-α) and related pathways. These are transduced by spe- cific receptors including the Fas antigen, the TNF receptors, and a variety of others (Nagata,1997). Interaction with the appropriate ligand, e.g., TNF-α or the Fas ligand, can result in apop- tosis mediated through signal transduction pathways. Resistance to such apoptotic mechanisms has been demonstrated in malignant lymphoid cells that are also resistant to certain chemothera- peutic drugs, such as doxorubicin or mitoxantrone (Landowski et al., 1997). An example of such an association is seen in Figure 20.12, where increasing concentrations of anti-Fas antibody, a surrogate for the Fas ligand, cause a rapid decrease in survival, which is completely absent in cells resistant to doxorubicin. However, resistance to Fas-mediated apoptosis does not select for drug resistance (Landowski et al., 1999).
Another major component of the apoptotic regulatory pathway is the bcl-2 gene, which occurs in the mitochondrial compartment and regulates the permeability of this organ system and interacts with specific proteins to inhibit apoptosis during potential apoptotic events (Kroemer, 1997; Reed, 1997). A large number of human neoplasms exhibit overexpression of the bcl-2 gene, as noted in Table 20.8. Overexpression of this gene does confer resistance to apoptosis on the cell, as discussed earlier in Chapter 6 in reference to the 14-18 translocation, resulting in an enhanced expression of bcl-2 in large-cell lymphomas (Figure 6.8). More re- cently, Voehringer and Meyn (1998) have suggested that bcl-2 also mediates its effects through alterations in glutathione metabolism. In an experimental system, they have demonstrated that depletion in glutathione in bcl-2–expressing cells restores apoptosis and reverses drug resistance.
Figure 20.12 Dose response of the effects of an anti-Fas antibody on apoptosis (cell survival) in sensi- tive and doxorubicin (S and Dox40 respectively) in cell culture. (Adapted from Landowski et al., 1997, with permission of the authors and publisher.)
This finding further establishes and relates apoptosis resistance as a mechanism of drug resis- tance to the alterations in glutathione metabolism that are similarly related to drug resistance.
Glutathione and Related Enzymes in Multidrug Resistance
As noted in Chapter 8, glutathione and the enzyme glutathione peroxidase play important roles in the inhibition of active oxygen radical–induced alterations in cellular metabolism. These reac- tions appear to play a similar role in the metabolism and resistance to some chemotherapeutic agents (Morrow and Cowan, 1990). But perhaps more important in drug resistance is the family of glutathione S-transferases. These are phase II (Chapter 3) enzymes involved in the conjuga- tion of a variety of substrates, many of which include drugs used in the chemotherapy of cancer or their metabolites. A listing of these is seen in Table 20.9, adapted from the review by O’Brien and Tew (1996). A variety of human neoplasms exhibit altered levels of glutathione-metaboliz- ing enzymes, as can be noted from Table 20.10.
Like phase I enzymes, the glutathione S-transferases may also be regulated in their expres- sion by a variety of different drugs and hormones. However, in the neoplasms listed in Table
20.10, the increased glutathione S-transferase activity is stable and not significantly altered by external factors. Furthermore, a number of the glutathione conjugates formed with chemothera- peutic drugs or their metabolites are removed by membrane pump proteins of the MDR family; thus, in a number of instances of resistance, increases in both the MDR p-glycoprotein or related proteins as well as GSTs may be seen (cf. O’Brien and Tew, 1996). This same review also dis- cusses the suggestion that genetic polymorphisms in the GSTs may alter their response to che- motherapy in specific disease conditions. In any event, it is clear that glutathione and specific aspects of its metabolism, especially conjugation with reactive forms and elimination of active oxygen radicals, play significant roles in drug metabolism and resistance. Furthermore, while this discussion has concentrated on isolating specific drug resistance mechanisms and discussing each separately, both the possibility and reality of multiple pathways of drug resistance in neo- plastic cells presently exist.
The Cell Cycle and Chemotherapy
Although many of the drugs used in chemotherapy eliminate neoplastic cells by effects during cell division, some drugs also have relatively specific effects during individual stages of the cell cycle. A schema of the relationship of the specific action of various cytostatic and cytotoxic drugs to the various periods of the cell cycle is seen in Figure 20.13. Chapter 9 presented a dis- cussion of the doubling times of neoplastic cells as well as some of the characteristics of their growth. A better knowledge of such variables, as well as their determination in the individual patient, is the goal toward which the methods of chemotherapy are directed. The number of cells in each of the various phases of the cell cycle at any one time in a neoplasm is dependent on a variety of factors. During the logarithmic growth, a greater proportion of the cells are in the S phase, whereas later in the natural history of the neoplasm, more neoplastic cells may be seen in the G1 or even G0 stage of the cell cycle. This latter situation is the rule with relatively slowly growing solid neoplasms, whereas rapidly growing embryonal neoplasms and acute leukemias have a much greater proportion of cells entering the S phase during most of their natural history. Not all cells of a neoplasm or a normal tissue have the capacity to replicate indefinitely. In nor- mal tissues there occur cells, termed stem cells, that have extensive cell-renewal capacity extend- ing throughout the whole or most of the life span of the organism (Chapter 14; Trott, 1994). Similarly, it appears that neoplasms have stem cells, which are those cells capable of continued
aGSH Px, glutathione peroxidase; GR, glutathione reductase. Elevated enzyme activity represented by (+) sym- bol. π denotes the specific isozyme, the π form, of glutathione S-transferase (GST).
Adapted from O’Brien and Tew, 1996, with permission of authors and publisher.
replication as long as the neoplasm grows within the host. Under ideal circumstances, the delin- eation of stem cells in the neoplastic population and of cells in the nonproliferating compartment is critical for successful drug therapy aimed at killing such populations.
A method that has been utilized to isolate in part and determine the effects of drugs on the tumor stem cell population is the human tumor stem cell assay originally described by Salmon et al. (1978). The assay, depicted in Figure 20.14, is carried out by obtaining individual neoplastic cells from a surgical specimen of the neoplasm under study through mechanical dissociation, followed by the incubation of such cell preparations in the presence or absence of a specific test drug and subsequent plating in soft agar. Malignant “stem” cells from the neoplasm will form individual colonies in the soft agar, such as has been described for cells transformed in culture (Chapter 14). One may then monitor the effect of the drug on colony formation and growth. This assay has proved valuable in the in vitro phase II (Figure 20.3) studies of new agents and in the initial screening of new analogs. One of its most notable accomplishments has been in establish- ing patterns of cross-resistance and sensitivity in relapsing patients as well as monitoring the development of clinical drug resistance through the use of serial tumor biopsies (cf. Bellamy,1992). As might be expected, however, a variety of technical difficulties prevent the use of this assay in routine clinical diagnosis and management of neoplastic therapy. Such difficulties in- clude (1) a low plating efficiency of freshly isolated human neoplasms, limiting the number of neoplasms that will form colonies for the assessment of drug effects; (2) potential selection of only a portion of the stem cells from the original sample; (3) difficulties in obtaining pure sus- pensions of single cells that have not been damaged by the technology; (4) potential for signifi- cant errors in assessment of “response” rates in both the assay and in patients; and (5) not all neoplasms have stem cells capable of growth in soft agar, as predicted from our discussions in
Figure 20.14 Steps involved in the establishment of the clonogenic (human tumor stem cell) assay of human neoplasms for the purpose of testing drug sensitivity and resistance of neoplastic stem cells. (Adapted from Riou and Bernard, 1982, with permission of the authors and publisher.)
Chapter 14. Despite these problems, there is some evidence for a correlation between clinical response and the in vitro response of neoplastic cells in the human tumor stem cell assay. These findings include the observation that cells from human neoplasms of a given histologic type have a response in vitro to therapeutic drug administration that is similar to the known clinical re- sponses of such neoplasms and in particular the patient’s clinical response (cf. Bellamy, 1992). Perhaps more important is that the assay has an extremely high capability (85%) to predict clin- ical resistance to the drug being studied (Von Hoff et al., 1983).
Unfortunately, the human tumor stem cell assay tells nothing about cells in the G1 or G0 phase of the cell cycle. However, a number of drugs (Figure 20.13) are effective during these phases, and a number of drugs appear to be non–cell-cycle-specific in their action. Thus, it is clear that the use of a single drug in the chemotherapy of neoplasia is doomed to failure both because of specific actions during the cell cycle and the development of drug resistance by the neoplastic cell through the variety of mechanisms discussed above (Figure 20.12). The apparent solution to this problem has been to use multiple drug combinations composed of chemicals active during different phases of the cell cycle. In some instances, attempts have been made ei- ther to synchronize neoplastic cells in vivo with subsequent administration of the drugs at the most sensitive times in relation to the synchronization of certain nonneoplastic tissues. This type of therapy has been termed chronotherapy, in which the peak of cell proliferation of neoplastic and normal cells during the daily cycle is determined by a variety of methods, with subsequent therapy administered at time periods giving appropriate and maximal differences between sensi- tivity of the neoplastic cells and resistance of the normal cells (Focan, 1995). An example of the different 24-hour patterns of DNA synthesis (S phase) in cells of bone marrow and a lymphoid neoplasm in the human is seen in Figure 20.15 (cf. Hrushesky and Bjarnason, 1993). Therapy
Figure 20.15 Patterns of DNA synthesis (S phase) activity in bone marrow (solid line) or malignant lymphoma (dashed line) sampled through two consecutive 24-hour periods. The curves represent best-fitting cosigned functions to the raw S phase data expressed as a predictable variation around the 24-hour mean (100%). (Adapted from Hrushesky and Bjarnason, 1993, with permission of the authors and publisher.)
with a drug active during the DNA synthesis phase (Figure 20.13) would expose a much higher proportion of neoplastic cells when given near 12 A.M. than bone marrow cells, which have a significantly lower number of cells in S phase at that time. Halberg (1977) was one of the first individuals to suggest taking advantage of differences in circadian rhythmicity for more effec- tive therapy of not only cancer but other diseases as well (Belanger, 1993). However, unless careful analyses of parameters within the neoplasm compared with the host are undertaken, dif- ficulties in results may occur (e.g., Adler et al., 1994). By combining knowledge of cell kinetics as applied to the human disease, differences in endogenous parameters such as seen in circadian rhythms, and the use of multiple drugs affecting different phases of the cell cycle, a rationale for therapy of both leukemias and solid neoplasms has been developed.