Cell cycle

Click to collapse Click to expand
Main | Save Edit | Discussion | History | Cube (0)

Cell Cycle and Cancer
 

Somatic cells are normally completely dependent upon signaling pathways in order to divide and proliferate (13). Only upon receiving specific mitogenic signals (from mitogens such as growth factors), indicating an appropriate environment for cells to grow in, will they begin to proliferate (13). Cells enter the cell cycle at the end of the G1 phase, after they have received sufficient mitogen exposure, and are checked for DNA damage to confirm that they possess all required machinery needed for successful division. The pathways that regulate the cell cycle and cell proliferation are disrupted in cancers (13). As we have previously learned, an example of an important checkpoint protein whose gene is deleted in certain cancers is the Retinoblastoma (Rb) protein (14). Rb acts as a gatekeeper protein, controlling entry into the S phase of the cell cycle, and when inactivated, the cell is able to proliferate uncontrollably (14). Below, we discuss various pathways critical to the proper functioning of the cycle and which are often disrupted in cancer.

 

Cell division: a cycle. Image retrieved from: www.australianscience.com.au/wp-content/uploads/2013/09/hela-spiral-copy.jpg

 

CDK and Cyclin
 

The cell cycle, also known as the cell-division cycle, is a series of events that ultimately leads to cellular division and duplication of a cell. Within this cycle, multiple checkpoints are present that regulate the progression through various stages, based on a series of complex biochemical reactions. This complex mechanism is controlled by a subfamily of cyclin-dependent kinases (CDKs). As their name suggests, the functionality of CDKs in regulating the cell cycle is dependent on their association with an activating subunit called cyclin, illustrated in Figure 14.2.1. These cyclins activate CDKs by binding to them. In terms of their enzymatic role, CDKs are serine and threonine kinases that phosphorylate their substrates to modify their functions.                                              

                              Figure 14.2.1. Eukaryotic cell cycle phases with respective cyclin-CDK complexes and inhibitors (CDKs) (Inspired by Fernandez et. al., 2009)
 

 

In eukaryotic cells, there are multiple CDK-cyclin complexes that play specific roles at various phases in the cell cycle. These complexes include three interphase CDKs (CDK2, CDK4, and CDK6), a mitotic CDK1 (also known as cell division control protein 2 (CDC2)), and ten cyclins belonging to four different classes (A-, B-, D-, and E-type cyclins) (1). This specificity is described in Figure 14.2.3, where mitogenic signals correlate with the increasing expression of D-type cyclins. By binding with CDK4 and CDK6, activation of the kinases during G1 results in the initiation of the cell cycle where the cell prepares for DNA synthesis (1). Cyclin degradation is a crucial process that corresponds to the inactivation of CDKs, which in turn affects the progression of the cell cycle (Figure 14.2.3). According to the classical model of cell cycle control, cyclin D-CDK4 and cyclin D-CDK6 stimulate the initiation of G1 phase and the start of the cell cycle (not). The progression towards the end of G1 phase is characterized by the increasing levels of cyclin E-CDK2, which in turn triggers the onset of S phase. As discussed previously, the reduction in cyclin E-CDK2 levels by degradation is important in initiating the S phase. As seen in Figure 14.2.3, cyclin A-CDK2 regulates the completion of the S phase and entry into G2 phase, where cyclin A-CDK1 is involved. The level of cyclin B increases during the start of mitosis and diminishes at end of the M phase. The inactivation of CDK1 due to decreasing cyclin B triggers the end of the cell cycle. Overall, the life cycle of a cell is orchestrated based on the presence of various CDK-cyclin complexes, which can affect the progression of cellular proliferation depending on their presence and levels. This selectivity within the cell cycle offers potential targets for clinical therapuetics for cancers.
 

                                            Figure 14.2.2. Active/Inactive forms of cyclin-CDK complexes. Complexes are inactivated by CKI (Cip/Kip & Ink4 family) (Inspired by Fernandez et. al., 2009)


 

Figure 14.2.3. Cyclin concentration throughout cell cycle. (Modification of work by "WikiMiMa"/ Wikimedia Commons)

 

Cyclin-CDK and Cancer
 

In the case of cancer, cell cycle defects are often generated by changes in CDK activity as a result of accumulated mutations. These mutations are characterized by unscheduled proliferation of cells due to constitutive mitogenic signaling and hyperactivation of various CDKs. The resulting occurrences are in favor of tumor formation due to amplified and unregulated cell division.


CDK

An aspect of cancer formation due to cell cycle aberration is based on the deregulation of interphase CDKs (1). In melanoma patients, a miscoding mutation (Arg24Cys) in the CDK4 gene results in an altered CDK4 that is unable to bind with Ink4 inhibitors. As a result, CDK4 levels are amplified which subsequently leads to the onset of deregulated cell division (1). Furthermore, certain leukemias such as pro-B acute lymphocytic leukemia possess an amplified level of CDK6 due to chromosomal translocation (3). In certain human pancreatic ductal adenocarcinomas, CDK5 is greatly expressed, resulting in the migration and invasion of pancreatic cancer cells (4).  Similarly, high levels of CDK1 and CDK2 have been observed in subsets of colon adenomas (9).


Cyclin

As mentioned previously and illustrated in Figure 14.2.1, D-type cyclins functions as a growth sensors that connects mitogenic stimuli with the cell cycle (6). Cyclin D1 and D3 translocations, amplifications, missense mutations, and elevated protein levels have been observed in multiple cancer types. These mutations and abnormities increase cyclin D activity, resulting in enhanced cell cycle progression to S-phase and cell proliferation (7). In B-cell and mantle cell lymphoma, cyclin D1 overexpression has been observed due to cyclin D1 gene translocation and the formation of a fusion oncogene with the immunoglobulin heavy chain gene (9).Overexpression of cyclin D1 tied to gene amplification has been studied in breast, esophageal, bladder, lung, and squamous cell carcinomas (9). Elevated levels of cyclin D proteins in cancer have also been attributed to defective mechanisms of degradation. Phosphorylation of cyclin D threonine283 targets it for nuclear export and proteasomal degradation; therefore, a missense mutation in threonine283 has been implicated in elevated cyclin D levels by attenuation of cyclin D degradation (10). Alternatively, in androgen sensitive prostate cancer, low levels of cyclin D1 and D3 have been observed. It is hypothesized that cyclin D interferes with androgen receptor signalling preventing cell cycle progression (11). While cyclin D appears to be involved in many types of cancers, it is not the only mis-regulated cyclin. cyclin E have also been reported to be overexpressed in breast and colon cancer (9), while cyclin A and cyclin E are amplified in some cases of lung carcinomas.


CKI

In normal cells, CDK activity is regulated by two types of inhibitors: INK4 proteins (INK4A, INK4B, INK4C, INK4D) and Cip/Kip family proteins (p21, p27, and p57) (1). Together, these cell cycle inhibitors function as a brake system that inhibit proliferation in multiple tissue types. Due to the inhibitory activity of CKI, which leads to the suppression of growth through pRb, CKIs possess a tumour suppressor functional role. The importance of these cylin-dependent kinase inhibitors can be explained by the presence of mutant p27 proteins unable to bind to CDK-cyclin complexes. Knock-in mice with the mutant genotype display an amplified stem and progenitor cell population, which is further indicated by the presence of a wide range of tumors. The p16Ink4a gene is observed to be altered in many human tumours by either deletion, point mutation, or hypermethylation (9). The loss of the p16 CKI leads to unrestrained procession of G1. In G2 phase, loss of p27 expression is observed in lung, breast, and bladder tumours. Due to the interconnectivity between the various CKIs within the cell cycle, the detrimental effect of a single aberrant CKI is not limited to one stage but rather to the entire proliferative cycle.

Cell cycle-targeted therapeutics

 

Cell cycle-targeted therapeutics
 

The complex pathways involved in oncogenic signaling are an obscurity against target therapies. Due to genetic heterogeneity acquired from accumulated mutations, cancer cells are able to circumvent and evade clinical therapies that target specific proteins or pathways. In the case of this section, anti-cancer drugs will be discussed based on inhibiting CDKs, considering its crucial role in modulating cell cycle. There are currently a number of therapeutic strategies that aim to control the level of CDK activity and they can be categorized into two groups: indirect strategy and direct strategy. In the case of indirect strategy, regulators of CDK activity such as cyclins are targeted, while direct strategy involve the inhibition of CDK kinases (9). The approaches that are indirect include CKI overexpression, synthetic peptides with CKI activity, cyclin inhibitor, and modulators of CDK-phosphorylated state. For therapeutics pertaining to direct inhibition of CDK activity, the concurrent inhibitors so far function via competitive inhibition of ATP binding to CDK, which leads to the disruption of CDK activation and cell cycle arrest. One of the major concerns in regards to intervention of the cell cycle pathway is that the inhibition will not discriminate between normal dividing cells and tumour cells, thus yielding detrimental and systemic side effects (9).

 

Flavopiridol

 

Flavopiridol is a synthetic flavonoid that follows the direct strategy as a CDK-inhibitor. Originally extracted from an indigenous plant from India, the CDK inhibitor demonstrates potent and specific in vitro inhibition of CDK1, CDK2, CDK4, CDK6 and CDK 7 (12). Preclinical studies have shown that Flavopiridol is capable of inhibiting angiogenic processes, inducing apoptosis, and promoting differentiation. The specific mechanism of Flavopiridol works by preventing phosphorylation of CDKs by down-regulating cyclin D1 and D3 expression (12). This occurrence leads to G1 cell cycle arrest and apoptosis as mentioned previously. Flavopiridol is currently under clinical trial phase 2 under the name alvocidib (12).

 

Signaling Pathways Regulating The Cell Cycle
 

ATM-Chk2 and ATR-Chk1 Pathways

One pathway regulating the progression of the cell cycle is the ATM-Chk2 pathway, which can arrest the cell cycle at numerous checkpoints if DNA damage is detected (15). In normal cells, when a double-strand break (DSB) is detected, the primary activator of the response pathway is the ataxia telangiectasia-mutated (ATM) protein kinase. ATM phosphorylates downstream proteins possessing a consensus sequence containing either a serine or threonine, immediately preceding a glutamine residue (15). Nbs1 augments the activity of ATM to enhance the response. One of the proteins phosphorylated at a threonine residue at position 68 by ATM is the kinase Chk2, which propagates the signal by also phosphorylating downstream proteins (15). A target of Chk2 is Cdc25, which upon phosphorylation is either targeted for degradation or sequestration in the cytoplasm, thus causing cell cycle arrest at the G1, S, and G2/M stages (15). Chk2 is implicated in transcriptional regulation as well, by phosphorylating E2F-1 and p53, promoting transcription of genes in their response pathway (15).

 

A similar signaling pathway that responds to the presence of single-stranded DNA is the ATR-Chk1 pathway. ATR-Chk1 is activated when DNA replication is obstructed such as when there is a deficiency of nucleotides or UV-induced dimers, which impede the replication fork, leading to the formation of ssDNA (16). This pathway is commonly activated in conjunction with the ATM-Chk2 pathway (16). TopBP1 enhances ATR activity by an unknown mechanism, which then phosphorylates Claspin, which in turn binds to Chk1. Chk1 is then recruited to ATR so it can be phosphorylated at serine residues, and consequently inhibits Cdc25 by phosphorylation and targets it for degradation (16). Thus, this pathway has a similar function of arresting the cell cycle.

 

These two pathways can delay progression from G1 to S phase, act within the S phase to slow the process of DNA replication, or prevent progression from G2 to mitosis (16). Thus, both pathways can act at any point before mitosis to prevent division if there is DNA damage. There are also checkpoints within the DNA replication process, whose responses can be controlled by both pathways. These responses include stabilizing stalled replication forks, suppressing replication at the origin, and delaying entry into mitosis until DNA replication is fully complete (16).

 

A defective response system to DNA damage involving both the ATM-Chk2 and ATR-Chk1 pathways is frequently found in many cancers (15). This results in a failure to halt the cell cycle when needed by loss of regulation at cell cycle checkpoints, and cells carrying DSBs and other DNA abnormalities are able to progress through the cycle and divide, passing down their mutations. A germ-line loss-of-function mutation in ATM predisposes individuals to lymphoma (16). ATM is considered to be a partially penetrant cancer susceptibility gene, since mutations in this gene increases incidence of cancer associated with environmental mutagen exposure (16). Related mutations in Nbs1 are also associated with a predisposition to cancer. Chk2 is considered to be a moderate to low cancer susceptibility gene, whose mutations are observed in breast and prostate cancer, among others (16). On the other hand, it is hypothesized that Chk1 is required for the proliferation of specific cancer stem cells, such as epidermal cells giving rise to skin tumors induced by exposure to carcinogens. It is interesting to note that deletion of Chk1 in mouse epidermal cells actually suppresses carcinogen-induced skin tumorigenesis (16). Disruptions in both signaling pathways have contrasting consequences. A cell or organism can tolerate a loss of ATM-Chk2 signaling, or a disruption in its pathway; however, it would be predisposed to cancer (16).  Contrastingly, the ATR-Chk1 pathway is needed for cell proliferation and survival in many types of cells. Thus, therapeutic strategies to combat tumorigenesis can target either ATR or Chk1. However, mouse models have also shown that a partial loss of Chk1 function can promote tumorigenesis, but further research is needed in this area.  

 

The therapeutic drug Tamoxifen is capable of deleting Chk1 in epidermal cells, reducing the conversion of skin papillomas to carcinomas (11). Most therapeutic strategies have been focused on Chk1, since it is a strong effector of DNA damage and replication checkpoints and is active in most tumors. Chk1 inhibition increases damage within the cells by increasing DSBs, thus activating apoptotic pathways. If normal cells prematurely enter mitosis (i.e. containing unreplicated DNA) due to failure to meet the requirements of various cell cycle checkpoints, they will die (16). Inhibiting Chk1, ATM, and ATR also augments the effects of genotoxic agents including radiation and nucleoside analogs. XL-844 is a Chk2 inhibitor being developed that will override the S phase checkpoint, leading to premature entry into mitosis, and reduced survival (16).

 

p21-Activated Kinase 1 (PAK1)  Pathways

The signaling pathways that characterize cancer cells involve several signaling kinase proteins. The p21-activated kinases (PAK) are another such group, and are associated with regulating several cellular processes including growth factor and steroid receptor signaling, reorganization of the cytoskeleton, oncogenic transformation, and cell survival (17). PAK1 is activated by external stimuli detected by activator proteins at the plasma membrane leading to a cascade resulting in the phosphorylation of PAK1, which changes from an inactive to an active conformation (17). PAK1 does not get activated in cancer cells as a result of a mutation, rather it acts as an oncogene if its expression is highly upregulated (17). This can occur as a result of increase copy number of the gene, inhibition of its negative regulators, or hyperactivation of upstream GTPases including Rac and Cdc42 (17).

 

PAK1 increases tyrosine kinase activity after being recruited to the plasma membrane by the adapter protein Nck upon activation of tyrosine kinases by growth factors (17).  Thus, it is an important component for tumorigenesis, which is regulated by growth factor signaling, and it is involved with cellular growth, proliferation and differentiation, all of which are processes of the cell cycle. Disruption of PAK1 mediated growth factor signaling is a common characteristic of cancer cells.

 

The cytoskeleton is involved not only in the structure and motility of cells, but is also a crucial player in the cell cycle (17). PAK1 specifically affects the cell cycle through the M phase. A study performed in breast cancer cells showed over-expressing PAK1 resulted in multiple spindle poles being formed leading to irregular separation of chromosomes, an event known as aneuploidy, which is illustrated in Figure 14.2.5 (17). Although PAK1 has several long signaling cascades for other cellular processes, the signaling pathway it initiates to regulate the cell cycle is quite short in comparison. PAK1 is normally localized to the spindle poles in the nucleus as a result of its kinase activity, where it can either phosphorylate spindle pole resident molecules Aurora A or Arpc1b at a threonine residue at position 21 (17). Arpc1b can also be phosphorylated by Aurora A, after the latter has been phosphorylated by PAK1. Aurora A kinases help the dividing cell pass on its genetic material to its daughter cells by controlling chromatid segregation (18). It functions during prophase of mitosis and is necessary for proper functioning of centrosomes (18). Defects in its function cause genetic instability leading to tumorigenesis. Additionally, PAK1 phosphorylates the cytoskeletal substrate tubulin cofactor B, a protein responsible for microtubule destabilization, which is important for changes in the cytoskeleton that occur during division (17). The involvement and localization of PAK1 in the nucleus is important for successful completion of mitosis as well as in the proliferation of cancer cells.

 

Figure 14.2.4. Effect of over-expressed PAK1 on mitosis in cancer cells resulting from a disrupted signaling pathway. Multiple spindle poles are formed during division and the end result is a genetic alteration.

 

There has been considerable effort to develop therapeutic agents against cancer cells with overexpressed PAK1. An endogenous therapeutic approach would be to increase expression of the tumor suppressor gene NF2, which inhibits activation of PAK1 by binding to its p21-binding domain (17). MicroRNA-7 can also inhibit PAK1 by binding to its 3’ untranslated region. CEP-1347 is a small molecule inhibitor that blocks the function of PAK1 through direct bindng (17). Finally, allosteric inhibitors such as IPA-3 can also be used, since they covalently bind the PAK1 regulatory domain, blocking its ability to bind and be activated by the upstream Cdc42. Other approaches including the use of organometallic compounds are being tested as well (17).

 

The MAP Kinase signalling pathway

The Mitogen-Activated Protein (MAP) Kinase pathways are composed of evolutionarily conserved kinase proteins, which transforms extracellular signals to fundamental cellular processes including proliferation, differentiation, migration and survival/apoptosis (19). As seen in the figure below, there are several different MAP kinase pathways, which are stimulated by different extracellular factors and in turn regulate different cellular processes. The MAP kinase pathways typically comprise three successive MAP kinases: MAP kinase kinase kinase (MAPKKK), which phosphorylates and activates MAP kinase kinase (MAPKK), which then phosphorylates and activates MAP kinase (MAPK). The MAPK then phosphorylates and activates a variety of transcription factors which act on DNA and either enhance or repress the transcription of their target genes.

 

                                                    

                                                                                      Figure: The four major MAP kinase pathways in mammals (21)

There are three main MAP Kinase groups: Extracellular signal-Regulated Kinase (ERK), Jun N-terminal Kinase (JNK) and p38. The MAP kinase cascades function downstream of cell surface receptors and cytoplasmic scaffold proteins (eg. Grb and Sos) which are specific for the different MAP kinase cascades. The ERK pathway is typically stimulated by mitogens/growth factors, while JNK and p38 pathways are typically stimulated by cellular stress (21).

The MAP kinase pathways are often dysregulated in cancer and several of the involved proteins have been identified as oncogenes, which when activated can lead to increased proliferation, cell survival and migration (19, 20, 21). One MAP kinase pathway has been more intensively studied than others due to its continuous dysregulation in tumors: The Raf-MEK-ERK pathway.

 

                                                   

                                                             Figure: The Raf-MEK-ERK pathway and the typical mutations found in different kinds of cancers

 

After stimulation of the EGF receptor by growth factors, it dimerizes enabling Grb2 to bind on its intracellular domains. Grb2 then recruits Sos which is capable of activating Ras, which then initiates a phosphorylation cascade of MAP kinases leading to phosphorylation and activation of ERK (MAPK). Phosphorylated ERKs can form homodimers and translocate to the nucleus where they activate various transcription factors leading to altered gene expression. Often the Raf-MEK-ERK pathway leads to increased cellular proliferation (19, 21). As it can be seen on the figure, both Ras and Raf are often found to be mutated resulting in their constitutive activation in various kinds of tumors (21). Ras proteins function as GDP/GTP-regulated switches with intrinsic GTPase activity. The exchange of GDP to GTP leading to activation of Ras is caused by guanine nucleotide exchange factors (RasGEFs e.g. Sos), while the hydrolysis of GTP to GDP causing inactivation is performed by Ras itself through its GTPase activity. Mutated Ras proteins often loose the ability to hydrolyse GTP, leaving Ras in a stimulus-independent activated state, while Raf is known to be a potent retrovirus oncogene (21). 

Several inhibitors of Ras, Raf and MEK have been developed and shown great promise. Due to the high occurrence of drug resistance in cancer cells, two inhibitors (e.g. one against Raf and one against MEK) are often used in combination (21). To date, no successful inhibitor of ERK has been developed (21). 

 

 

Review Questions

1. Discuss a potential cancer therapeutic strategy to target a component in a cell cycle signaling pathway with reference to the following questions. What pathway would you target? What protein/component of the pathway would you target? What would be the downstream consequences as a result of applying this therapeutic? Explain how the cell cycle would be affected.
 

2. During a co-op work placement at UBC, a MEDG421 student discovers that his culture of breast cancer cells have a mutation in the p21Ink4a gene. This mutation causes the p21 protein to be aberrant in functionality. Explain the significance of this mutation in relation to carcinogenesis and cell cycle, while also indicating how he may mitigate the phenotyical effects of this mutated protein. 
 


References

1.     Schwartz, G.K. and Dickson, M.A. (2009). Cell cycle, CDKs and cancer: a changing paradigm. Nature Reviews Cancer, 9(3): 153-66

2.     Bloom, J. and Cross, F.R. (2007). Multiple levels of cyclin specificity in cell-cyle control. Nature Reviews Molecular Cell Biology, 8: 149-160.

3.     Kuo, T.C., Chavarria-Smith, J.E., Huang, D., and Schlissel, M.S. (2011). Forced expression of cyclin-dependent kinase 6 confers resistance of pro-B acute lymphocytic leukemia to Gleevec treatment. Molecular Cell Biology, 31(13): 2566-2576

4.     Eggers, J.P., Grandgenett, P.M., Lewallen, M.E., Tremayne, J., Singh, P.K., Swanson, B.J., Andersen, J.M., Caffrey, T.C., High, R.R., Ouellette, M., and Hollingsworth, M.A. (2011). Cyclin-dependent kinase 5 is amplified and overexpressed in pancreatic cancer and activated by mutant K-Ras. Clinical Cancer Research, 17(19): 6140-6150.

5.     Williams, G.H. and Stoeber, K. (2012). The cell cycle and cancer. Journal of Pathology, 226: 352-364.

6. Choi Y.J., Anders L., (2014). Signaling through cyclin D-dependent kinases. Oncogene, 33(15):1890-903.

7. Ashgar U., Witkiewicz A.K., Turner N.C., Knudsen E.S., (2015). The history and future of targeting cyclin-dependent kinases in cancer therapy. Nat Rev Drug Discov, 14(2):130-46.

8. Musgrove E.A., Caldon C.E., Barraclough J., Stone A., Sutherland R.L., (2011). Cyclin D as a therapeutic taret in cancer. Nat Rev Cancer, 11(8):558-72.

9.   Vermeulen, K., Bockstaele, D.R.V. and Berneman, Z.N. (2003). The cell cycle: a review of regulation, deregulation and therapeutic targets in cancer. Cell Proliferation, 36: 131-149

10. Lahne H.U., Kloster M.M., Lefdal S., Blomhoff H.K., Naderi S., (2006). Degradation of cyclin D3 independent of Thr-283 phosphorylation

11. Olshavsky N.A., Groh E.M., Comstock C.E., Morey L.M., Wang Y., Revelo M.P., Burd C., Meller J., Knudsen K.E., (2008) Cyclin D3 action in androgen receptor regulation and prostate cancer. Oncogene, 27(22):3111-21.

12.     National Cancer Institute (NCI). Alvocidib. Retrived from http://www.cancer.gov/drugdictionary?cdrid=42068

13.     Evan IG, Vousden KH. 2001. Proliferation, cell cycle and apoptosis in cancer. Nature. 411:342-348.

14.     Weinberg RA. 1995. The Retinoblastoma protein and cell cycle control. Cell. 81:323-330.

15.     Buscemi G, Carlessi L, Zannini L, Lisanti S, Fontanella E, Canveari S, Delia D. 2006. DNA damage-induced cell cycle regulation and function of novel Chk2 phosphoresidues. Molecular and Cellular Biology. 26:7832-7845.

16.     Smith J, Tho LM, Xu N, Gillespie DA. 2010. The ATM-Chk2 and ATR-Chk1 pathways in DNA damage signaling and cancer. Advances in Cancer Research. 108:73-112. 

17.     Eswaran J, Li DQ, Shah A, Kumar R. 2012. Molecular pathways: targeting p21-activated kinase 1 signaling in cancer: opportunities, challenges, and limitations. Clinical Cancer Research. 18:3743-3749.

18.     Keen N, Taylor S. 2004. Aurora-kinase inhibitors as anti-cancer agents. Nature Reviews Cancer. 4:927-936.

19.     Dhillon A. S. et al. 2007. MAP kinase signalling pathways in cancer. Oncogene. 26; 3279-3290.

20.     Wagner E. F. and Nebreda A. R. 2009. Signal integration by JNK and p38 MAPK pathways in cancer development. Nature Reviews Cancer. 9; 537-549

21.     Roberts P. J. and Der C. J. 2007. Targeting the Raf-MEK-ERK mitogen-activated protein kinase cascade for the treatment of cancer. Oncogene. 26; 3291-3310.