SEARCH

SEARCH BY CITATION

Introduction

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

Apoptosis is a unique form of cell death in which the cell activates a self-destruct mechanism, causing its death. The realization that most, if not all, anticancer agents effect tumour killing through the activation of apoptosis has emphasized the key role of the process in tumour growth and cytotoxic resistance [1]. In the last decade, there has been intensive scientific research in the field and as the understanding of the genetic modulation of the apoptotic process evolves, new therapeutic options arise. A critical development in the current understanding of cytotoxic cell killing is that DNA damage induced by an anticancer agent does not necessarily result in cell death. Thus, cells cannot be seen as passive recipients of drug-induced damage. Rather, effective killing involves the cell being able to activate the genes involved in the apoptotic process [2,3]. Therefore, inappropriate expression of oncogenes and tumour suppressor genes (TSGs) is fundamental to the sequential events which lead to the development of cancer and resistance to cytotoxic therapy [4].

In bladder cancer the mechanisms of chemoresistance are not fully understood and this may in part be due to a lack of knowledge of the processes involved in apoptosis. However, many investigators within the bladder cancer research community are now addressing the issue of drug resistance and apoptosis. This review will discuss the current understanding of the concepts of apoptosis and the clinical significance of the processes involved in resistance to cytotoxic therapy in bladder cancer.

Definition of apoptosis

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

The introduction of encompassing terms such as necrobiology (necros death and bios life) [5] illustrates both the confusion over definitions of cell death and the need to view apoptosis within the broader context as a process of growth regulation. Apoptosis describes the events of cell death which occur throughout normal development and as a response to a variety of initiation stimuli. Apoptosis is often incorrectly referred to as being synonymous with programmed cell death (PCD). Not all apoptosis is PCD but, as implied by the term PCD, cell death is predetermined by a genetic clock which is activated at an appropriate time [6,7]. An example of PCD can be seen in the perfect timing of the death of individual cells in invertebrates and in embryonic development [8]. Thus PCD should be reserved for describing death that is a normal part of life or development, but not related to a specific form of cell death.

Another form of cell death, necrosis, is distinct from apoptosis and occurs after injury or toxic damage. It is not seen in normal development and is fundamentally different from apoptosis both in its nature and biological significance [9,10]. In necrosis, the cell is unable to maintain ionic homeostasis and metabolic collapse results. Consequently, the cell swells and the membranes leak, resulting in an inflammatory reaction. In contrast, during apoptosis, the cell membrane maintains its integrity and cytosolic proteins cross-link to produce a net which envelopes the cell and prevents leakage of cellular structures. Therefore, in apoptosis, the surrounding tissue architecture is not disrupted and there is no inflammation [11].

Evaluation of apoptosis in bladder cancer

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

Cell morphology

In 1885, Walther Fleming, who was studying regression of ovarian follicles, observed that cell death was associated with the generation of small cells and small cellular particles in which the nuclei were in the process of disintegrating [12]. Surprisingly, the distinct morphological features were not formally described as apoptosis until 1972 [13]. In 1988, a report on the effect of mitomycin on the cells of patients with TCC clearly described nuclear compaction and clumping of chromatin. Although the authors did not allude to the process by name, this was probably the first description of apoptosis in bladder cancer [14]. There are several methods, based on cell morphology, DNA fragmentation and flow cytometry, which can be used to identify apoptotic cells in bladder cancer. The original description of apoptosis was based on morphological nuclear features as determined by electron microscopy [13]. Electron microscopy, which facilitates the most accurate identification of apoptotic cells, remains the gold standard [15]. However, sampling problems at high magnifications make quantitative cell counts difficult. For this reason, measurements of apoptosis by light microscopy are accepted as the reference standard [5,16,17] and have been used in bladder cancer [18[19]–20].

The morphological features of apoptosis include compaction of the nuclear chromatin (pyknosis), loss of cellular volume and chromatin condensation on the nuclear envelope (nuclear crescents) [21]. These features, which are common to all apoptotic cells, are illustrated in bladder cancer (Fig. 1). After chromatin condensation, convolution of the nuclear and cellular outlines, referred to as ‘zeiosis’ [22], precedes the collapse of the nucleus and its disintegration into sphere-like fragments. These fragments become sealed by plasma membrane to form several apoptotic bodies in which the closely packed organelles appear intact [23].

image

Figure 1. Apoptotic cells in bladder cancer: originally ×600. The apoptotic cell loses its cell-to-cell contacts and the cytoplasm becomes eosinophilic. DNA clumps and abuts the nuclear membrane before disintegrating into apoptotic bodies. These membrane-bound bodies are dispersed and taken up by surrounding cells and macrophages.

Download figure to PowerPoint

The morphological features of apoptosis occur as a result of biochemical events within the cell. Several calcium- and magnesium-dependent enzymes become activated and cleave DNA at internucleosomal sites into uniform lengths of 185 bases [10,24,25]. This cleavage of DNA by endogenous endonuclease activity reduces nuclear bulk and is morphologically seen as chromatin condensation. Interestingly, the fragmentation of DNA into low-molecular weight material has a protective function, limiting the probability of transfer of damaged genes, in a potentially active state, from dying cells to the nuclei of neighbouring viable cells [10]. It is feasible to monitor cytotoxic therapy and disease progression by expressing quantitative apoptotic measurements in vivo. Apoptosis can be expressed as an apoptotic index (AI), i.e. the number of apoptotic cells divided by the total tumour cell population. Several reports have related quantitative estimates of apoptosis to outcomes in bladder cancer. A study examining patients with invasive TCC, treated by radiotherapy and cystectomy, reported that the AI correlated with potential prognostic factors such as stage, grade and mitotic index. Interestingly, the AI was the only significant predictor of radiation response for patients with stage T3b TCC [20]. A second study, which analysed 400 patients with TCC, reported that the AI was increased in high-grade nonpapillary TCC. The AI predicted recurrence-free survival for patients with Ta-T1 tumours and was independently related to nuclear size and mitotic index [19].

One of the difficulties encountered when measuring apoptosis or cell proliferation indices is determining how to measure the total cell number used as the denominator in the calculation of an AI [17,26]. In solid tissues, the denominator must accurately represent the cell density of the tissue and be easily reproducible. Various methods to estimate cell populations have been described in solid tissues and bladder cancer [19,27]. These attempt to measure the area or volume of tissue, which is then used as an indicator of cell number. The papillary nature of superficial TCC and the consequent variations in cellularity mean that accurate apoptotic measurements will only be determined if the number of apoptotic cells is expressed as a fraction of the total tumour cell population. To avoid tedious counting procedures, we described a rapid method for estimating the cellularity in bladder tumours which is fast, reproducible and will facilitate studies examining the relevance of apoptosis measurements in TCC [18].

DNA fragmentation

Various methods can be used to detect the biochemical events associated with the apoptotic process, such as the formation of DNA fragments; these fragments can be detected by DNA electrophoresis, which requires about 5×105 homogenized cells, thus precluding the examination of apoptosis on a cell-to-cell basis [28,29]. Alternatively, in situ end-labelling (ISEL) techniques, which identify DNA breaks by the incorporation of labelled nucleotides into the breaks, can facilitate both the identification and quantification of individual apoptotic cells [30]. In situ labelling of fragmented DNA using the terminal deoxynucleotidyl transferase reaction (TdT or TUNEL) has been used to identify apoptotic cells in bladder cancer [31] but there are several disadvantages which limit its application. In situ techniques require elaborate handling of cells and the binding of TdT is dependent on the cleavage activity of cellular endonucleases. The activation of cellular endonucleases varies between tissues and the reliance on a single assay which is dependent on a specific endonuclease may underestimate the number of apoptotic cells present in a population [32]. In addition, the effects of tissue fixation, processing and the generation of non-apoptotic DNA fragments by cytotoxic agents [29,33,34] reduce the sensitivity and specificity of in situ methods [35[36]–37].

Flow cytometry

Flow cytometry can be used to assess rapidly the changes associated with apoptotic cells in a large number of single cells in suspension. This is achieved by analysing the features which are characteristic hallmarks of apoptotic cells, e.g. cell size, chromatin distribution, DNA content, membrane characteristics and gene protein expression [5]. Furthermore, standardized inter-cell population comparisons between levels of uptake of dyes, levels of antibody binding and incorporation of nucleotides are facilitated by flow cytometric techniques. It is possible to obtain urothelial cells by bladder irrigation and labelling with cytokeratin can be used to ensure that apoptosis is only examined within epithelial cell populations [38].

In flow cytometry, the cell’s ability to scatter light is proportional to its size, as assessed by forward scatter, and its granularity, as assessed by side scatter. In the late stages of apoptosis, both forward and side scatter properties decrease and apoptotic cells can be discriminated from live cells. Unfortunately, misclassification of apoptotic bodies and nuclear fragments as single apoptotic cells can cause confusion. This underlines the importance of conducting flow cytometry assays in tandem with other more specific assays, such as morphological assessment, to ensure that apoptotic cells are identified correctly [5].

During the apoptotic process, the permeability of the plasma membrane alters, which permits the leakage of DNA. This characteristic can be exploited to identify the percentage of cells with a subnormal DNA content. However, the use of sub-G1 peaks as direct measures of the percentage of apoptotic cells is inaccurate because of the errors that will be introduced by overlapping cell-cycle phase populations and debris. Other novel methods to detect early changes in the cell membrane of apoptotic cells have been described [39]. The transference of a membrane phospholipid, phosphatidylserine, from the inner to the outer leaflet of the cell plasma membrane occurs early in the apoptotic process. Phosphatidylserine can be detected by flow cytometry with fluorochrome-conjugated annexin V, which binds naturally to the exposed phospholipid [40]. A decrease in the mitochondrial transmembrane potential is another early event in apoptosis. This can be recognized by a decreased ability to accumulate rhodamine in the mitochondria of apoptotic cells compared with normal cells [5].

Molecular biology of apoptosis

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

In the bladder, as in other tissues, failure of the regulatory genes involved at any point in the apoptotic pathway may result in prolonged survival of an abnormal genome, tumorigenesis and resistance to anticancer agents (Fig. 2). Understanding of the genetic control of the apoptotic process is as yet incomplete. In humans the interactions are complex and involve multiple regulatory processes. Thus much of the understanding of the apoptotic process has come from studying the nematode Caenorhabditis elegans [41,42]. The apoptotic process of this nematode appears to be similar to that in humans and many of the protein products of the nematode genes are homologous with those in humans [43[44]–45]. Indeed, in humans, as in the nematode, genes responsible for the regulation of apoptosis interact in a cascade, inducing, checking and inhibiting cell death. This cascade has been described as a pathway which can be divided into initiation, regulation and degradation phases (Fig. 3) [46,47].

image

Figure 2. A normal cell with damaged DNA can repair itself or be removed by apoptosis. A cell in which apoptotic genes are defective can survive to develop cancer. The subsequent failure to induce apoptosis affects tumour growth regulation and response to cytotoxic therapy.

Download figure to PowerPoint

image

Figure 3. The Bcl-2 checkpoint is can be considered as a rheostat and is regulated by initiating genes which signal the cell to enter the pathway. If a cell passes through the checkpoint the caspase effector genes dismantle its components and apoptosis is inevitable.

Download figure to PowerPoint

Initiation

The p53 gene.

In response to DNA damage the TSG p53 can induce either apoptosis or DNA repair, and has thus been described as the ‘guardian of the genome’ [48]. The decision to signal a cell to undergo apoptosis or repair depends on the extent of DNA damage, as well as the extent of induction of downstream target genes [49,50]. Following DNA damage, increased levels of p53 can result in cell-cycle arrest through transcriptional activation of p21WAF‘oo1‘ox [51]. The p21WAF‘oo1‘ox gene is an inhibitor of cyclin-dependent kinases that are required for G1- to S-phase progression [52]. After p21WAF‘oo1‘ox induction, cells fail to exit G1 and growth is checked until the cells are repaired [53]. Failure of repair processes may result in the cell re-entering the apoptotic pathway, although this mechanism is still unclear. Alternatively, p53 can induce and repress effector genes which are downstream in the apoptotic pathway, such as bcl-2 and bax [54]. So far, 34 genes have been shown to be regulated by p53, although the complex interactions involved and the identity of several of these genes remains to be clarified [55]. However, it is currently thought that p53 and its transcripts induce apoptosis through the generation of reactive oxygen species [55]. It has been postulated that these reactive oxygen species act on the mitochondrial membrane, causing loss of selective ion permeability and collapse of the membrane potential, resulting in activation of effector caspase genes [56].

The induction of apoptosis by many anticancer therapies correlates with functional or wild type (wt) p53 status [57]. The importance of the gene as a major trigger of apoptosis in bladder cancer has been suggested by the presence of wt p53 in sensitive bladder cell lines and conversely the absence of wt p53 in resistant cell lines [58,59]. However, the induction of apoptosis depends not only on wt p53. Both mitomycin and cisplatinum have been shown to induce apoptosis in a bladder cell line expressing active p21WAF1 but with nonfunctional p53, and both agents failed to induce apoptosis in a cell line expressing nonfunctional p53 and inactive p21WAF1 [59].

In bladder cancer, mutations of p53 are implicated in carcinogenesis [60] and associated with pathological stage and disease progression [61]. To date there are few reports relating p53 alterations to clinical outcome after chemotherapy for TCC. The wt p53 protein has a short half-life, but mutations stabilize the protein, extending its half-life and enabling detection by immunohistochemistry. Although not always the case, the immunohistochemical detection of p53 can be associated with alterations or mutations of the gene [62,63]. Several reports have suggested a relationship between p53 overexpression, as detected by immunohistochemistry, and chemoresistance. Sarkis et al. [64] reported that p53 overexpression had independent prognostic value which was inversely related to survival in patients after treatment with neoadjuvant chemotherapy for invasive bladder cancer. This report corroborates two studies which examined p53 overexpression in patients with CIS treated with BCG [65,66]. Although there were few patients, Kurth et al. [65] reported that 40% of patients with CIS and p53 overexpression did not respond to BCG, and subsequently developed invasive disease. These results do not concur with a study by Cote et al. [67] who reported a threefold reduction in risk of recurrence after adjuvant chemotherapy when p53 was overexpressed in TCC. An interesting paper described a novel mechanism of action for p53 to explain these results; Waldman et al. [68] postulated that the sensitivity of cells lacking functional p53 was a result of an inability of mutant p53 to upregulate p21WAF1 . Consequently, cells fail to exit the S-phase, repeatedly synthesize DNA, but do not undergo mitosis, become aneuploid and develop gross nuclear abnormalities which are susceptible to an increased ‘hit rate’ by cytotoxic agents. These conflicting clinical results of p53 analysis in bladder cancer highlight the complexity of the interpretation of p53 expression, as assessed by immunohistochemistry [69], and the limitations imposed by technical aspects of the procedure [70].

Fas TNF.

The cytokine family of proteins includes tumour necrosis factor (TNF) and Fas Ligand (FasL). Both TNF and FasL bind to their receptors and regulate cellular proliferation and differentiation. The Fas receptor is ubiquitously expressed in various tissues, while FasL is predominantly expressed in activated T lymphocytes and natural killer (NK) cells. Activation of T cells induces expression of the FasL which binds to and causes trimerization of Fas on the target cells [71,72]. This interaction plays an important role in the cytotoxic T cell-mediated apoptosis against cancer cells [73]. It is not clear how the FasL TNF system is regulated, and while Bcl-2 can inhibit Fas-mediated apoptosis [74], pathways independent of Bcl-2 also exist [75]. Similarly, the understanding of events which signal apoptosis after receptor binding is incomplete. Cytosolic proteins are required to signal apoptosis by binding to the Fas receptor [76]. This signal can be inhibited by antioxidant proteins which regulate the cell redox environment, such as glutathione [77,78]. Mizutani et al. [79] reported that doxorubicin enhanced the expression of Fas in T24 bladder cells, and synergy was achieved by treatment of T24 cells with anti-Fas antibody and doxorubicin. In that study, anti-Fas antibody did not alter the levels of the antioxidant enzyme glutathione-π, but glutathione, which is present in TCC [80], has been reported to be increased in doxorubicin-resistant bladder cell lines [81]. In leukaemia cells, Fas resistance has been reversed by lowering intracellular glutathione [78] but this effect has not been examined in bladder cancer.

The FasL/Fas pathway may be important for BCG-induced killing by T cell activation. The host immune response is a significant factor in the cytotoxicity of BCG [82], which may increase tumour cell killing by activation of cytotoxic T cells and macrophages [83,84]. The observation that apoptosis was induced in the T24 bladder cell line when incubated with lymphokine activated killer cells [85] suggests that the ultimate effect of BCG in bladder cancer is to induce apoptosis [85]. Further investigation of the FasL pathway may help to determine the mechanisms involved in the mode of action of both BCG and chemotherapy.

Regulation

The bcl-2 checkpoint.

After initiation, the apoptotic process converges towards a common pathway in which two important and distinct checkpoints have been identified. These are controlled by the bcl-2 family of genes and the caspases. The bcl-2 checkpoint controls and regulates the effector phase, either inhibiting or promoting the passage of cells towards the degradation stage of apoptosis. Bcl-2 is located on the outer mitochondrial membrane, where it regulates the formation of channels or ‘megapores’ which control the flux of ions to maintain the potential across the mitochondrial membrane [86,87]. Opening of mitochondrial pores alters the membrane potential, with the resultant generation of oxygen free radicals, release of stored calcium and activation of the terminal effector caspases [88].

Bcl-2 can function as an inhibitor of apoptosis induced by chemotherapeutic agents and ionizing radiation [89,90]. Bcl-2 does not function by protecting cells against drug-induced DNA damage [91,92]. Instead, it acts downstream of this event by possibly preventing damaged DNA from being translated into a signal for activating genes involved with apoptosis [93]. The bcl-2 gene is the prototype member of a family of genes which can act either to promote or inhibit apoptosis. The various bcl-2 family members can dimerize with one another, with one monomer antagonizing or enhancing the function of the other. bcl-2, bcl-xL , bcl-w and mcl-1 inhibit apoptosis. Conversely, bax, bik, bak, bad and bcl-xs activate apoptosis [94]. It is believed that the ratio of inhibitors to activators in a cell determines the propensity of the cell to undergo apoptosis and its susceptibility to a given apoptotic stimulus [76,95[96]–97].

Overexpression of bcl-2 has been reported to be associated with hormone resistance in prostate cancer [98], and reduced expression of bax has been found to be associated with failure to respond to therapy in patients with breast cancer [99]. In bladder cancer, bcl-2 has been investigated mainly by using immunohistochemical analysis of archival material. Caution must be exercised in the interpretation of these reports, as results are conflicting, and observations that the Bcl-2 protein is labile may in part explain this [100]. It was initially suggested that Bcl-2 protein and mRNA were not present in normal urothelium [101,102], but were present in 63% of cases of low-grade TCC of the bladder [102]. More recently it has been reported that Bcl-2 protein is expressed in the basal layer of normal transitional urothelium [103] and in ≈24% of invasive TCC (Fig. 4) [104]. The pattern of Bcl-2 staining in normal urothelium would explain the reduced susceptibility of the basal urothelial cells to apoptosis.

image

Figure 4. Immunohistochemical detection of Bcl-2 in superficial bladder tumours. ×200. a, The expression of the protein in the basal urothelial cells, which is the distribution in normal urothelium. Variation in expression of Bcl-2 is illustrated in b, in which only sporadic urothelial cells are positive, and in c, in which staining extends beyond the basal cells.

Download figure to PowerPoint

Our group has determined the change in expression of the Bcl-2 and Bax proteins by flow cytometry in patients’ TCC cultures after exposure to mitomycin, and related these changes to the induction of apoptosis (unpublished data). In tumours with a high AI after exposure to mitomycin, there was a trend for increased expression of Bax and reduced expression in Bcl-2. In tumours which did not undergo apoptosis there was no clear association between the changes in Bcl-2 and Bax. These preliminary results would suggest that the Bcl-2 checkpoint is unregulated, possibly through a defect in upstream genes which initiate the process.

Degradation

The caspase checkpoint.

A family of 10 or more proteolytic enzymes termed caspases are an essential component of the degradation stage of apoptosis [105]. Caspases are expressed as precursors which are activated when cleaved, resulting in the assembly of active proteases [106]. The caspases are activated in a proteolytic cascade and function to cleave target proteins which are integral to the cell structure and repair processes [107]. As such, the caspases are viewed as the executioners of the apoptotic process. The destruction of structural proteins such as laminins and repair proteins such as poly(ADP)ribose polymerase (PARP) and DNA-dependent protein kinase (DNA-PKcs) ensures the crippling, rapid and irreversible death seen in apoptosis [107]. The regulation of the caspases and the role of these enzymes remains to be elucidated. The bcl-2 family can act to regulate the caspase genes. However, recently it has been shown in vitro that the caspases can also act on the Bcl-2 family of proteins to reverse their anti-apoptotic effects [108,109]. This may thus provide an active area for future research.

The future: a rational approach to cancer therapy

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

Research into the mechanisms involved in the process of cancer cell apoptosis has entered an exciting phase. The possibility exists to target therapies at specific sites of the apoptotic pathway, enabling defects to be bypassed and checkpoints to be manipulated. It may be possible to block a gene which provides a selective survival advantage using antisense oligonucleotides and thus sensitize cells to a cytotoxic agent. For example, Bcl-2 antisense oligonucleotides, when incorporated into cells, bind to bcl-2 mRNA and have been shown to down-regulate bcl-2 protein translation, enhancing the in vivo and in vitro response to chemotherapy [110,111].

A second approach is to bypass a defect in the regulation stage. For example, it is possible to activate Fas by cross-linking of the receptor with agonistic anti-Fas antibodies and thus initiate apoptosis in targeted cells [112]. A third approach is to target tumour cells which have a specific gene defect, e.g. using an engineered adenovirus that selectively replicates in and kills cells lacking functional p53 [113].

The major challenge in bladder cancer will be to identify the defects in the apoptotic mechanism responsible for chemoresistance. A dilemma is the possibility that the different agents used in treating bladder cancer initiate differing pathways to cell death. However, a greater understanding of the mechanisms of action of agents and of tumour cell response will enable existing and novel therapies to be applied rationally and effectively. In future, it will be possible to ‘trick’ a tumour into killing itself by activating its own ‘death machine’.

Conclusion

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References

The accessibility of bladder tumours and the relatively impermeable urothelial–blood barrier make it possible to administer cytotoxic agents topically, at high concentrations, without significant systemic absorption. Despite these advantages, the efficacy of chemotherapy and immunotherapy for TCC is limited, with objective response rates of ≈50% [114,115]. Understanding apoptosis and its molecular controls is redefining the traditional concepts of tumour growth, response to therapy and cytotoxic resistance. The notion that cytotoxic therapies operate through the induction of apoptosis holds enormous promise for successful modifications to cancer chemotherapy.

References

  1. Top of page
  2. Introduction
  3. Definition of apoptosis
  4. Evaluation of apoptosis in bladder cancer
  5. Molecular biology of apoptosis
  6. The future: a rational approach to cancer therapy
  7. Conclusion
  8. Acknowledgements
  9. References