In previous parts of this serial review,[1–3] several oncogenes have been described. These oncogenes were loosely grouped into four categories: (i) growth factors; (ii) tyrosine kinases; (iii) intermediate signaling molecules; and (iv) transcription factors. This part focuses on genes that are best classified as regulators of the cell cycle or apoptosis. All of the oncogenes in this review are expressed or overexpressed in cancer tissues, and their presence is often the cause of tumor initiation, growth and progression, or metastasis. Oncogenes are seen as advantageous targets because downregulation of their gene product or inhibition of their action(s) may be sufficient to halt tumor growth or perhaps even eradicate a tumor.

Various strategies are being employed to interrupt the function(s) of oncogenes. Investigators are exploring the use of antisense, interfering RNA, aptamers and antibodies to decrease or completely knock down expression of oncogene mRNA or protein. Other strategies involve specific inhibitors of enzyme activity, competitors for binding, and molecules that disrupt the tertiary or quaternary structure of proteins, or that target other molecules upstream or downstream of the target to limit its actions. Many of these strategies show promise for future clinical use. In addition, directly targeting oncogenes should result in a more specific, more effective treatment. This may lead to a decrease in toxicity for patients receiving cancer therapy. Therefore, oncogenes are an attractive and viable target for cancer therapy.

1. Cell Cycle Regulators

1.1 MDM2

The MDM2 oncogene was first cloned as an amplified gene on the murine double-minute chromosome in 3T3DM cells, and was found to be tumorigenic when NIH3T3 cells harboring amplified MDM2 were injected into nude mice.[4–6] MDM2 overexpression confers tumorigenicity to cultured cells and to intact mice.[6] MDM2 is overexpressed in a number of human malignancies, including cancers of breast, colon, esophagus, bladder, prostate, liver, pancreas, lung and kidney; in lymphomas and leukemias; and in soft tissue sarcoma, melanoma, glioma, and tumors of the peripheral nerve sheath.[5–9] MDM2 overexpression, which is associated with invasive tumors, high grade/late stage tumors, tumor recurrence and metastasis, is a negative prognostic factor for cancer patients.[10,11]

The human MDM2 gene is located on chromosome 12q13-15.[8] Like the mouse gene, the human gene also contains an intronic p53-dependent promoter (P2),[12] which is highly conserved between mice and humans. The P2 structure has two tandem p53 binding elements. In order to activate the promoter, p53 must bind simultaneously to both binding elements. The requirement for dual binding may help prevent the promoter from becoming activated in the presence of low amounts of p53.[12] The MDM2 protein contains a p53-binding domain, a nuclear localization signal (NLS), a nuclear export signal (NES), a central acidic domain, and a C-terminal zinc-finger and RING finger domain.[13]

1.1.1 Molecular Mechanisms in Cancer

Until recently, the major functions of the MDM2 oncoprotein were believed to be associated with its interaction with p53, because p53 and MDM2 operate within a feedback autoregulatory loop. p53 increases the expression of the MDM2 gene via its P2 promoter, and MDM2 in turn negatively regulates p53 by masking its activation site (to keep it inactive) and by increasing its degradation via the proteasome.[13–20] Additionally, some mutants of p53 (gene TP53) stabilize and increase the expression of MDM2 while avoiding the increase in degradation.[21] Many environmental insults and cancer treatments, including γ-irradiation and chemotherapy, increase the levels of p53, leading to G1 arrest or apoptosis.[22,23] Modulation of the p53-mediated induction of cell cycle arrest and apoptosis may lead to the sensitization of tumor cells to DNA damaging treatments such as chemotherapy and radiation.[5] Unfortunately, the activation of p53 by DNA damage may be limited in cancers with MDM2 expression or overexpression. Inactivation of the MDM2 negative feedback loop or interference with the p53-MDM2 interaction may increase the magnitude of p53 activation following DNA-damaging treatment, thus enhancing the therapeutic effectiveness of chemotherapeutic drugs and radiation therapy.

One concern is that toxicity to normal tissues could result from the activation of p53, possibly narrowing the therapeutic window for p53-activating strategies. Fortunately, recent data from in vitro and in vivo studies suggest that the consequences of p53 activation may not be the same for cancer cells and their normal counterparts. For example, activation of p53 results in cell cycle arrest in normal fibroblasts but causes apoptosis in cancer cells.[24] A study in transgenic mice suggests that p53 regulation in normal tissues may be different from that in cancerous tissues.[25] In transgenic mice, the conditional knockout of MDM2 results in only mild hemato-poietic abnormalities, and does not induce severe toxicity.[26] Our observations support this lack of toxicity. Antisense oligonucleotides targeting human MDM2 inhibit the growth of human cancers in mouse xenograft models, but do not cause any severe adverse effects. To rule out the possibility that the lack of toxicity is due to sequence differences between species, an antisense molecule specific for mouse MDM2 was designed and delivered to normal CD-1 mice using the same treatment protocol; again, no major toxicities were observed.[27]

Recently, p53-independent activities have been discovered for the MDM2 oncoprotein. For example, MDM2 modulates the differentiation of normal muscle cells by interacting with the SP1 transcription factor and RB1.[28] MDM2 also confers a growth advantage to cells lacking p53 and RB1, and overcomes the G1-cell cycle arrest induced by p107.[29] In a transgenic mouse model, overexpression of MDM2 predisposes mice to spontaneous tumor formation in both the presence and absence of functional p53.[30] MDM2 also binds to various proteins, including RB1, E2F1, the ribosomal proteins L5, L11 and L23, promyelocytic leukemia protein, nucleophosmin, and Yin Yang (YY1).[13,28,31–35] Although the effects of these interactions are not fully understood, they may be associated with the transformational properties of MDM2. In addition, MDM2 downregulates CDKN1A (p21WAF1) at both the transcriptional and post-translational levels, and facilitates its degradation, independent of p53.[16,31] MDM2 may act similarly for other proteins, including those mentioned earlier. Moreover, overexpression of MDM2 may be responsible for drug resistance and radio-resistance of human cancers, and may be associated with an increase in MDR-1 expression[36] and transforming growth factor β1 (TGF)-β1 resistance.[37]

In addition to the normal, full-length MDM2 protein, more than 40 alternative and aberrant splicing variants of MDM2 are present in a wide spectrum of human cancers. Most of the variants lack at least part of the p53 binding domain, the NLS, or the acidic domain.[38] Some of these variants transform NIH3T3 cells and are associated with high-grade and late-stage human cancers, suggesting that they play a role in p53-independent tumorigenesis.[39]

1.1.2 MDM2 as a Therapeutic Target

Targeting the p53-dependent activity of MDM2 is necessary because more than 50% of human cancers harbor p53 mutations, and p53 is required for the effects of many chemotherapeutic agents and radiation.[21] Inhibition of the p53-independent activities of MDM2 may be equally important, and it is possible that inhibition of MDM2 can overcome drug resistance in tumors with dysfunctional p53. Several approaches have been used to validate MDM2 as a potential target for cancer therapy: micro-injection of a monoclonal antibody against the p53-binding domain on MDM2;[40] utilization of sequence-specific antisense molecules;[27,32,41–46] expression of a peptide homolog of p53 that binds to MDM2 but does not upregulate expression;[47] and transduction of cells with MDM2-binding peptides coupled with scaffold-stabilized domains.[48] These studies demonstrated that disruption of the p53-MDM2 negative feedback loop leads to accumulation and activation of p53, and thus to p53-dependent cell cycle arrest and apoptosis.

Blocking expression of specific proteins by the use of antisense oligonucleotides, which apparently act by hybridization with and degradation of targeted mRNA molecules, is effective against various diseases, including cancer. We have recently designed anti-MDM2 oligonucleotides that inhibit MDM2 expression in tumor cells in vitro and in vivo. In a variety of tumor models, regardless of their p53 status, these antisense inhibitors have antitumor activity and sensitize tumors to chemotherapy and radiation.[34]

Efforts are being made to develop other MDM2 inhibitors, including small molecules, antibodies, and peptides. Some small molecule inhibitors of MDM2 show in vitro and in vivo effects in the activation of p53 and inhibition of tumor growth.[49] Another strategy is to elicit a host immune response against the MDM2 protein, which has two highly conserved major histocompatibility complex (MHC) class I binding motifs (MDM2100 and MDM2441). Dendritic cells presenting MDM2100 stimulate an effective cytotoxic T-lymphocyte reaction against melanoma cells.[50] While these strategies may prove to be effective for reducing the effects of the MDM2 oncoprotein, the studies are preclinical, and the utility of MDM2 as a drug target in cancer patients remains to be confirmed.

2. Apoptosis Regulators

2.1 B-Cell Lymphoma-2 (BCL2)

B-Cell lymphoma-2 (BCL2) was identified in studies of the chromosomal translocation (t14;18)(q32;q21), which places the BCL2 gene under the control of the immunoglobulin heavy chain and results in its deregulated expression.[51] BCL2 is present in 60–85% of follicular B-cell lymphomas.[51,52] The human BCL2 gene, localized on chromosome 18,[53] encodes a protein that is a member of an extended family of proteins involved in the regulation of cellular apoptosis.[54] BCL2 is expressed in cells that are rapidly dividing and differentiating. It is generally absent in adult tissues but is expressed in a variety of fetal tissues. BCL2 is thought to maintain survival of antigen-receptor selected B cells in lymph nodes. It also parallels the differentiation of T cells in the thymus, with those most sensitive to apoptotic stimuli having less BCL2.[55] These physiological functions are illustrated by BCL2 knockout mice, which are smaller in size but still viable, although viability drops to 50% by 6 weeks after birth.[56] All phenotypes show growth retardation, loss of B cells and T cells, reduced bone deposition, and limited clonal expansion of muscle cells.[57]

BCL2 promotes cellular survival rather than proliferation[58] and protects cells from a variety of cytotoxic insults such as cytokine deprivation, irradiation and chemotherapeutic drugs.[54] This immortalization effect on cells apparently comes from inhibition of apoptosis in cells destined for death.[59] There have been two mechanisms suggested for the functions of BCL2.[60] One is that BCL2, possibly with other family members such as BCL-XL (BCL2L1), sequesters ’BH3 domain-only’ molecules in stable mitochondrial complexes, preventing the activation of the proapoptotic proteins BAX and BAK.[61] The BH3 domain-only molecules are BCL2-related proteins characterized by a sequence that is required for their killing function. The other proposed mechanism is that BCL2 promotes caspase activation independent of the cytochrome c/APAF1/caspase-9 ‘apoptosome’ .[62]

2.1.1 Molecular Mechanisms in Cancer

The presence of deregulated BCL2, which promotes cell survival and inhibits apoptosis, apparently contributes to tumorigenesis.[63] In vitro data demonstrate that high levels of the BCL2 protein allow interleukin-3-dependent lymphoid and myeloid cell lines to survive in the absence of cytokines.[63] In agreement with this observation, BCL2 transgenic mice develop neoplasms.[64,65] In addition, BCL2 in mice cooperates with the co-transfected MYC gene, which is also targeted to lymphoid cells, to promote formation of tumors, including early B-cell precursor, and mature B-and T-cell lymphomas.[65,66] This synergistic tumorigenic activity is also observed in other transgenic mouse models that develop breast and prostate cancer.[55] Further, BCL2 is essential for the maintenance of leukemia in transgenic mice: elimination of BCL2 results in the loss of leukemic cells and in prolonged survival.[65]

There is overexpression of BCL2 in human malignancies of hematological origin, such as acute myeloid leukemia (AML), B-cell cancers, and multiple myeloma, and those of non-hematological origin, including melanoma, breast cancer, prostate cancer, small cell lung cancer (SCLC), cervical cancer, and colorectal carcinoma; overexpression is associated with poor prognosis.[67–69] Moreover, BCL2 upregulation is associated with angiogenesis, metastasis, and resistance to chemotherapy and radiation.[60,70] In particular, the presence of BCL2 limits the success of strategies targeting either survival signaling or the activation of a tumor suppressor.

BCL2 is associated with the nuclear envelope, the outer mitochondrial membrane and the endoplasmic reticulum (ER). Expression of BCL2 in the ER reduces the amount of calcium that can be released from storage in the ER,[71] a factor that may be involved in its anti-apoptotic activity.

2.1.2 BCL2 as a Therapeutic Target

In colorectal carcinoma cells, knockdown of BCL2by use of small interfering RNA (siRNA) results in p53-dependent apoptosis under normal growth conditions, without prior exposure to genotoxic agents.[72] Since BCL2 is upregulated in a variety of human cancers, and apoptosis induced by BCL2 knockdown does not require DNA damage, the protein appears to be a promising target for cancer therapy. This approach could allow development of therapeutics with high specificity, efficiency and sensitivity, with less toxicity.

For lymphoma cells containing the (t14;18) translocation, an antisense molecule, oblimersen (G3139, augmerosen, Genasense®) downregulates BCL2 expression and shows antitumor activity in vitroand in vivo.[73] Oblimersen, designed to be complementary to the first 18 nucleotides of the BCL2coding region,[60] has no effects on viability of normal fibroblasts, which do not express BCL2, or on lymphoblastoid cells that do not have the translocation. On the basis of promising preclinical studies, a phase I clinical trial was undertaken with relapsed non-Hodgkin lymphoma patients. In this study, oblimersen downregulated BCL2in 44% of the cases and produced responses in 14%.[74] Moreover, oblimersen enhanced the efficacy of dacarbazine in preclinical studies and in phase I/II clinical trials, and there were positive results of this antisense strategy in non-hematological solid tumors, including breast cancer,[67,75] melanoma,[76] SCLC,[77] prostate cancer[78] and glioma.[79] This oligonucleotide is also apparently well tolerated.[73] Unfortunately, although oblimersen has shown promise in several studies, the largest and most recent clinical trial using the drug did not show significant improvement in the survival of patients with melanoma. These disappointing results led to a negative review of a recent New Drug Application (NDA) of this drug. Future studies should be concerned with improving the stability of the compound, with the efficiency of its delivery, determining whether it specifically targets cancer cells, and concerned with identifying cancers (if any) for which it can be effective as a single agent.

An alternative strategy is to design peptide inhibitors to reduce the protective effects of BCL2 on cell death. The hydrophobic pocket on the BCL2 protein, revealed by analysis of its three-dimensional structure,[80] serves as the recipient of the BH3 domain of pro-apoptotic BH3-only molecules and provides the basis for development of antagonists to BCL2.[81] Peptides that mimic the structure of the BH3 domain have demonstrated promising activity in experimental cancer therapy.[82] A single change from the normal sequence, however, does not result in apoptosis.[83] The toxicities of these compounds may limit their future use.

Recently, there have been searches for small molecule inhibitors with advantages of increased solubility, cell permeability, stability, and a higher affinity for BCL2.[59] The inhibitor tetrocarcin A, identified by screening large numbers of compounds in existing chemical libraries, enhances the apoptotic effects of tumor necrosis factor (TNF)-α and FAS in BCL2-overexpressing cells, and, as a single agent, displays anti-tumor activity.[84] There is concern that knockdown of BCL2 may not be sufficient to induce apoptosis in all kinds of cancer cells, and it is not clear whether Bcl-2 is involved in the homeostatic functions of cells. A potential problem for this kind of molecule is its lack of specificity for the target protein. Further mechanistic studies are needed, and, as is customary, it will be necessary to ensure that only specific function(s) are affected by the targeting.

2.2 XIAP (BIRC4)

Apoptosis is a tightly regulated cell-death program that is of fundamental importance in various biological processes such as cell turnover, development, metamorphosis and maintenance of homeostasis.[83] Abnormalities in apoptosis are involved in carcinogenesis and in the resistance of cancer cells to therapy.[85] Upon being activated, apoptosis-inducing pathways eventually converge to activate caspases,[86] leading to the cleavage of important cellular substrates, including poly(adenosine diphosphate [ADP]-ri-bose) polymerase and the laminins.[87] The activities of caspases are negatively regulated by the inhibitor-of-apoptosis family of proteins (IAP), the most potent of which is X-chromosome-linked IAP (XIAP).[87]

The identification of IAPs in baculoviruses was followed by the discovery of their presence in other species, including Drosophila, Caenorhabditis elegans, yeast and vertebrates.[88] The human XIAP gene (BIRC4) is located on chromosome Xq25[89] and encodes a 54 kDa protein that has three BIR (baculovirus inhibition of apoptosis repeat) motifs, a structure characteristic of IAP proteins that includes a conserved cysteine and histidine core sequence, Cx2Cx6Wx3Dx5Hx6C.[90]

The XIAP protein binds to and inhibits caspase-9 through its BIR3 domain, and caspases-3 and -7 through its linker-BIR2 domain.[91] It also interferes with the BAX/cytochrome c cell-death pathway.[92] XIAP promotes caspase-3 degradation through the proteasome by facilitating its ubiquitination via the ubiquitin E3 ligase activity in its RING finger domain near the C-terminus.[93] In addition to regulation of apoptosis, XIAP is involved in several other cellular processes. By direct binding, XIAP activates TAB1, a signaling molecule of the bone morphogenic protein (BMP) pathway,[94] and there is binding between XIAP and the BMP receptor (type 1).[95] Additionally, XIAP has roles in the signaling pathways of SMAD, NF-κB and JNK,[96] in copper homeostasis[97] and in regulation of cell cycle progression during mitosis.[98]

2.2.1 Molecular Mechanisms in Cancer

Overexpression of XIAP inhibits cellular apoptosis induced by a variety of stimuli, including TNFα, Fas, serum or growth factor withdrawal, ischemia, and chemotherapy and irradiation.[99] XIAP is present at basal levels in normal adult tissues, but is upregulated in many types of human tumors.[100] In AML patients, higher XIAP protein levels may indicate a poor prognosis.[101] In renal cell carcinomas, XIAP expression correlates with tumor stages, with the highest levels being present in poorly differentiated cancers.[102,103] Moreover, its overexpression correlates with the resistance of cancer cells to chemotherapy and irradiation.[100] Although XIAP appears to be involved in tumorigenesis, its potential as a cancer target is compromised by the observation that an apoptotic phenotype is not evident in XIAP-deficient mice.[104] However, cancer cells with deleted XIAP show remarkable sensitivity to TNF-related apoptosis-inducing ligand (TRAIL), suggesting that XIAP is a non-redundant modulator of TRAIL-induced apoptosis.[105]

In tumor cell lines and breast carcinoma tissues, there is upregulation of both pro-apoptotic and anti-apoptotic signals.[100] It is possible that, during carcinogenesis, the abnormalities in transformed cells trigger the activation of apoptosis. A small fraction of transformed cells overexpressing anti-apoptotic molecules, including XIAP, can survive, leading to tumor development.[100] This hypothesis is strengthened by the observation that, in human cancer cells, small-molecule antagonists of XIAP release the inhibition of caspase-3 and induce cell death, while showing little toxicity to normal cells.[106] These data suggest that, in cancer cells, caspases are in an ‘activated’ state compared with normal cells. These ‘activated’ caspases are restrained by multiple mechanisms, including IAPs, to maintain the survival of cancer cells.

2.2.2 XIAP as a Therapeutic Target

By using inhibitors designed to target IAPs, including XIAP, activated caspases could be modulated to induce apoptosis in cancer cells. Other lines of evidence can also be used to rationalize the targeting of XIAP for cancer therapy. For example, over-expression of XIAP-associated factor 1 (XAF1), a protein counteracting XIAP, induces apoptosis in human tumor cell lines but not in normal cells.[100] Therefore, downregulation of XIAP in human cancers by small molecules, small peptides, or RNA interference may support pro-apoptotic pathways in tumor cells.

Preliminary studies show anti-tumor activity and chemosensitization effects of XIAP-specific antisense nucleotides,[99,107] siRNA molecules[99,108–110] and the SMAC (second mitochondria-derived activator of caspase) peptide.[87] Now that the structures of the XIAP functional domains essential for the inhibition of caspases have been resolved, there is a search for small molecular inhibitors of XIAP that can reverse the inhibition of caspases.[111,112] These cell-permeable molecules may have advantages over other methods of inhibition, including higher specificity, bioavailability and stability. Given the possible roles of XIAP in normal cellular processes, however, it is possible that molecules targeting XIAP may affect normal cellular homeostasis. Additional studies of the mechanisms of action of XIAP are needed.

2.3 Survivin (BIRC4)

2.3.1 Molecular Mechanisms in Cancer

Survivin, which plays important roles in the inhibition of apoptosis and regulation of cell division,[113] was initially identified through hybridization screening of a human genomic library with the cDNA of the effector cell protease receptor-1.[114] The human survivin gene (BIRC5)is located on chromosome 17q25 and encodes a 16.5 kDa protein, making it the smallest member in the IAP family.[115] The survivin protein has only one BIR (baculovirus IAP) domain. In solution, it is a stable homodimer with the C-terminal α-helices protruding from the core.[115] Survivin expression, believed to be regulated largely at the transcriptional level, is dependent on the cell cycle; there are large amounts present in the G2/M phase.[116] The canonical cell-cycle dependent element (CDE)/cell cycle genes homology region (CHR) boxes on the BIRC5promoter may act as G1-repressor elements.[117] At the post-translational level, the survivin protein is regulated either by ubiquitination and proteasome degradation during the interphase[118] or by the stabilizing effects of phosphorylation on Thr34 by cdc2-cyclin B1 during metaphase.[119] In addition, survivin is upregulated by thrombopoietin, stem cell factor, and FLT3 ligand (FL)[120] through PI3K/AKT activation and through the STAT3 pathway.[121,122]

During cell cycle progression, survivin is found on centromeres at the prophase/metaphase, in the spindle mid-zone during anaphase,[123] and in the cleavage plane during telophase and cytokinesis.[124] Its pattern of localization suggests that survivin is a chromosome passenger protein involved in cytokinesis and chromosome movement during cell division.[125] This hypothesis has been substantiated in animal studies in which disruption of survivin caused cell division defects.[125]

Overexpression of survivin prevents cellular apoptosis induced by a variety of stimuli;[125] transgenic mice expressing survivin in the skin show impaired UVB-induced apoptosis.[125] In a variety of tumor models, downregulation of survivin through techniques involving antisense oligonucleotides, ribozymes or siRNA leads to suppression of tumor growth.[125] Although the mechanisms mediating its inhibitory effects remain to be elucidated, there is evidence that survivin blocks apoptosis through CDKN1A (p21WAF1). [126] It may also bindto and inhibit the activities of caspases 3,7, and 9.[127] Another mechanism could involve regulation of the subcellular location of apoptosis-inducing factor (AIF).[128]

2.3.2 Survivin as a Therapeutic Target

The potent inhibitory effects of survivin on apoptosis and its basal expression levels in normal adult tissues make it a reasonable target for cancer therapy. Moreover, survivin is present at high levels in a variety of human malignancies, including carcinomas of lung, breast, colon, stomach, esophagus, liver, pancreas, uterus, ovaries, Hodgkin- and non-Hodgkin lymphoma, neuroblastoma, various leukemias, sarcoma and melanoma.[115] Considerable retrospective evidence indicates that survivin levels correlate negatively with clinical prognosis and survival, and positively with aggressiveness, recurrence and resistance to the therapy.[115] Survivin, transfected into rat tumor cells, inhibits apoptosis; it also promotes tumorigenesis of these cells when they are transplanted into severe combined immunodeficient (SCID) mice.[129]

The potential of survivin as a target for cancer therapy is also indicated by evidence derived from use of an antisense molecule to downregulate survivin expression in leukemic HL60 cells. Accompanying survivin inhibition, there is cell cycle arrest at the G2/ M phase, followed by apoptosis.[130] In xenograft models of human gastric carcinoma, non-Hodgkin lymphoma, thymic lymphoma and lung cancer, antisense oligonucleotides have antitumor and radiosensitization effects.[107,131–133] Based on preclinical results, antisense molecules targeting survivin are under development for clinical use.[127]

Several groups working on the development of siRNA molecules to downregulate survivin have generated promising results.[110,134,135] There is concern, however, that the low in vivo efficacy of the siRNA must be improved before these molecules enter clinical studies. Use of ribozymes is an alternative strategy that does not require repeated doses;[136] ribozymes demonstrate in vitroand in vivoeffects in human prostate cancer and melanoma.[137,138]

Stimulation of the host immune system to generate a response against survivin-bearing tumor cells is also effective.[139] Human cytotoxic T lymphocytes (CTL) specific for survivin-derived determinants[140] kill primary tumor cells isolated from acute leukemia patients.[141] In a phase I clinical trial involving a peptide derived from survivin as a vaccine, there was a mild response in 1 of 11 colorectal patients; no severe adverse effects were evident.[142] Another strategy relies on the dominant-negative effects of the T34A mutant of survivin, which demonstrates pro-apoptotic and anti-tumor activities in vitroand in vivo.[143] Evaluation of additional experimental therapeutics that interfere with survivin is anticipated in future clinical trials. Because of the important role of survivin in the regulation of proliferation and survival of normal adult hematopoietic cells, however, therapy directed toward survivin might affect normal hematopoiesis.

3. Future Prospects and Conclusions

Most of the chemotherapeutic drugs presently available — even those recently brought into clinical use, such as irinotecan for treatment of colorectal cancer; taxanes for breast, ovarian and lung cancer; and carboplatin for ovarian cancer — are cytotoxic agents characterized by their cell-killing capacity (to all dividing cells) rather than by their tumor-specific mechanisms. For these agents, there are notable problems related to host toxicity and drug resistance. With progress being made in understanding the cellular and molecular mechanisms of carcinogenesis and tumor progression, a more rational approach to drug design is within reach. For improvement of therapy, oncogenes present new targets that can be valuable for the development of highly specific and effective antitumor drugs.[144]

A variety of studies have validated the targeting of oncogenes for cancer therapy. In addition to the preclinical and clinical studies mentioned in this review, numerous others have shown that interference with the function of oncogenes or prevention of their expression can decrease the rate of tumor growth and often cause complete tumor regression.[145–149] Oncogenes are involved in all major aspects of cancer, from initiation to progression, invasion, and response to therapy. Inhibiting the function of oncogenes or eliminating the expression of oncoproteins can thus act at each stage of tumor development. Although cancer is a consequence of many steps involving a variety of genetic events, removal or interference with the function of a single oncogene can result in cancer prevention, regression or improved response to treatment.[145–149]

Recently, many studies have indicated in transgenic mice that conditional inactivation of oncogenes results in the regression of tumors and leads to a reversal of tumorigenesis.[145–149] Moreover,even tumors demonstrating complex genetic alterations regress upon oncogene inactivation. The regression has also been demonstrated in vitro, further validating the approaches used to inhibit the action or decrease the expression of oncogenes.

It is noteworthy that regardless of the identity of the oncogene, the consequences of oncogene inactivation are similar: cellular differentiation, apoptosis and arrest of cell proliferation. Following only a brief inactivation, the reactivation of some oncogenes, including MYC, results in apoptosis — not resumption of tumorigenesis. Although the experimental data have demonstrated the reversibility of tumorigenesis induced by the inactivation of oncogenes, the mechanism(s) are still not clear. Several mechanisms have been suggested based upon the functions of the oncogenes and the various steps in oncogenesis. These include: (i) preventing the initiation of cancer; (ii) decreasing proliferation; (iii) inducing differentiation; (iv) decreasing angiogenesis; (v) increasing the efficacy of DNA repair/regaining genomic stability; (vi) decreasing the resistance to cytotoxic drugs or ionizing radiation; (vii) inducing apoptosis/allowing normal apoptosis to occur; and (viii) decreasing the invasion into surrounding tissues and decreasing metastasis. Another factor that should not be overlooked is the tumor microenvironment. In addition to signals coming from the affected cell, surrounding cells may enhance the tumor proliferation. One final consideration is the systemic response to the tumor. Decreasing the expression or inhibiting the function of oncogenes may also allow for enhanced immune system surveillance that can result in eradication of the tumor.

Finally, oncogenes can also be developed as biomarkers to provide pharmacodynamic endpoints and to serve as diagnostic and prognostic parameters in determining the effect of drugs on their molecular targets and in selecting subjects for clinical trials.

A better understanding of the molecular basis of carcinogenesis and tumor progression helps to identify and validate new therapeutic targets, and allows effective application of the newly discovered agents in the context of tumor biology.[150,151] For example, depending on the role of their target genes in the early and late stages of cancer, drugs may be utilized as chemopreventive agents or as treatment for established tumors. There are still challenges. In order to be able to create more effective therapies, it is necessary to gain a more thorough understanding of the fundamental differences between normal and tumor cells, particularly the differences in cell growth and death. Other areas of high importance for investigation are: (i) the basis of drug selectivity; (ii) the underlying mechanisms of the combination therapies used to treat cancers that are characterized by multiple molecular abnormalities; (iii) the optimal clinical use of molecular therapeutics in combination with conventional therapy; (iv) when and how to target mechanism-based agents against particular subgroups of cells characterized by expression of a particular genes; and (v) how to apply genetic-based drugs for chemoprevention and chemotherapy. The future of cancer therapy appears to lie in the development of drugs that target molecular abnormalities and deregulated pathways.