The ubiquitin proteasome system (UPS) has emerged from obscurity to be seen as a major player in all regulatory processes in the cell. The concentrations of key proteins in diverse regulatory pathways are controlled by post-translational ubiquitination and degradation by the 26 S proteasome. These regulatory cascades include growth-factor-controlled signal-transduction pathways and multiple points in the cell cycle. The cell cycle is orchestrated by a combination of cyclin-dependent kinases, kinase inhibitors and protein phosphorylation, together with the timely and specific degradation of cyclins and kinase inhibitors at critical points in the cell cycle by the UPS. These processes provide the irreversibility needed for movement of the cycle through gap 1 (G1), DNA synthesis (S), gap 2 (G2) and mitosis (M). The molecular events include cell-size control, DNA replication, DNA repair, chromosomal rearrangements and cell division. It is doubtful whether these events could be achieved without the temporally and spatially regulated combination of protein phosphorylation and ubiquitin-dependent degradation of key cell-cycle regulatory proteins. The oncogenic transformation of cells is a multistep process that can be triggered by mutation of genes for proteins involved in regulatory processes from the cell surface to the nucleus. Since the UPS has critical functions at all these levels of control, it is to be expected that UPS activities will be central to cell transformation and cancer progression.
The Nobel Prize for Chemistry was awarded in 2004 to the discoverers of the ubiquitin system of protein degradation. Intracellular proteolysis occurs by several mechanisms: the UPS (ubiquitin–proteasome system), autophagy (wrapping a double membrane around cytoplasm to form an autophagosome, with subsequent fusion with a lysosome) and the endosome–lysosome system (endocytosis of membrane components to form endosomal vesicles followed by fusion with a lysosome). It turns out that molecular events in these distinct cellular protein-degradation systems are controlled by the covalent attachment of ubiquitin or ubiquitin-like molecules to target proteins. This illustrates the diverse utilization of the ‘ubiquitin superfold’ in the course of evolution. The ubiquitin fold can be attached (covalently linked) to target proteins or can be genetically built into proteins for diverse downstream functions. A generic term for the complete family of covalently linked ubiquitin-like molecules and built-in ubiquitin folds is ‘ubiquitons’. The conjugatable ubiquitons are used for a variety of purposes in the cell, in addition to acting as a signal for degradation by the 26 S proteasome. The diverse uses of the ubiquiton in the life process is a tribute to the parsimonious nature of evolution and the reason for the award of the Nobel Prize .
It is worth noting that a mere 30 years ago, the widely accepted dogma was that the degradation of proteins in cells was improbable or did not occur, since such a process would ‘waste’ energy. It was thought that the ATP invested in protein synthesis was too valuable to expend by subsequently destroying the proteins. However, it has been known for 52 years  that intracellular proteolysis proceeds by mechanisms requiring ATP! That proteins should be degraded by energy-dependent reactions seemed unlikely, since proteolytic enzymes in the gut and blood did not require ATP hydrolysis for their activities. The requirement of ATP for intracellular proteolysis was considered theoretically unnecessary and therefore probably experimentally wrong! This is not the case: ATP is required by the UPS at several stages and is necessary for the use of ubiquitin and ubiquitin-like molecules for autophagy and the endosome–lysosome systems.
The process of protein ubiquitination is driven by three classes of enzymes: E1, E2 and E3. Ubiquitin is activated by a ubiquitin-activating enzyme (E1), transferred to a ubiquitin-conjugating enzyme (E2) and then handed to a ubiquitin ligase (E3) for covalent attachment to the ∊ amino group of a lysine residue of a target protein (Figure 1). A ubiquitin chain, consisting of at least four ubiquitins linked via Lys48 of the previous ubiquitin, acts as a signal for protein degradation. The polyubiquitinated protein is recognized and binds the 26 S proteasome. The 26 S proteasome consists of a 20 S catalytic barrel-like core with a multiprotein 19 S regulator at each end of the 20 S core. Each target protein is deubiquitinated, unwound by the hexameric ring of ATPases in the base of the 19 S regulator, and fed into the central chambers of the 20 S core for fragmentation into small peptides .
This UPS has a central role in the degradation of cytosolic and nuclear proteins and therefore in the control of the cell cycle and cancer.
The cell cycle
The cell cycle can be divided into stages, as indicated in Figure 2. In rapidly growing and dividing cells, the cycle describes two major phases: interphase and mitosis. In the former process (interphase), a sequential unidirectional set of events occurs that can be divided into a gap (G1) between mitosis and where DNA synthesis and chromosomal replication take place (S phase), followed by another gap (G2) before mitosis (M) takes place. In the latter process (mitosis), chromosomes are segregated to the two daughter cells followed by cell division (cytokinesis). The cell cycle is a critical time for every cell since any mistake may lead to DNA damage, genome instability, cell death or cancer. At every step in the cycle there must be checks and balances to ensure that only normally replicated DNA is incorporated into the daughter cells.
The key questions are: first, what mechanisms ensure that one process is complete before the next process starts and, secondly, what mechanisms ensure the strict alternation of S-phase and M-phase? The answer to the first question is ‘checkpoints’, which are safe stopping points where progress can be halted and previous events subjected to biochemcial quality control. The answer to the second question is a carefully constructed set of regulator proteins that are controlled by phosphorylation and dephosphorylation and by ubiquitin-dependent degradation. Cells deemed to have failed the checkpoint scrutiny may be eliminated by apoptosis. If the checkpoints and apoptosis fail to operate properly, then a cell with damaged DNA may survive, eventually to become a cancer cell.
Accelerators and brakes
The orderly irreversible progression of the cell cycle is controlled by molecular machines. These molecular machines consist of many protein components, including Cdks (cyclin-dependent kinases) and kinase inhibitors (Figure 3). The concentration of every protein in the cell is determined by the balance between a rate of synthesis of the protein (Ks) that depends on transcription of the gene and translation of the mRNA, and a rate of degradation (Kd) of the protein. The concentration of each protein (P) in a steady state is given by: P=Ks/Kd
The complexity of the cell cycle requires both the ordered synthesis and degradation of proteins to ensure the unidirectionality of the process. A seminal finding supporting the overall notion was made in 1983, by Tim Hunt and colleagues, for which he was awarded a Nobel Prize. They found that several proteins in rapidly dividing sea-urchin embryos varied in concentration with a cyclical periodicity dependent on continuous synthesis and periodic degradation at the end of the cell cycle . These proteins were named cyclins and were subsequently found to be essential for the activity of the cell cycle by controlling kinases, hence the term Cdk. However, the ubiquitin story was still in its infancy and it took many years to realize that the cyclical disappearance of the cyclins was due to ubiquitin-dependent proteolysis. It is now known that the concentrations of these key cell-cycle regulators are controlled by both phosphorylation and ubiquitination, followed by proteasomal degradation.
Ubiquitin ligases involved in cell-cycle control
The interplay between the UPS, Cdks and inhibitors is shown in Figure 4. The devil is, as usual, in the detail. Two classes of related ubiquitin-protein ligases: the SCF (Skp1–Cdc53/Cul1–F-box protein) ligase and the APC/C (anaphase-promoting complex/cyclosome) are used to ubiquitinate cell-cycle proteins at defined transition points in the cycle. There are three groups of E3s: the single subunit RING-finger type, e.g. the p53 targeting Mdm2 (muring double minute clone 2 oncoprotein) E3; the multisubunit RING-finger type, e.g the SCF and APC/C E3s; and the HECT (homologous to E6-AP C-terminus)-domain type, e.g. the cellular E6-AP (E6-associated protein) ligase that is ambushed by the papilloma virus E6 protein and used to ubiquitinate p53 for proteasomal degradation as part of the process of causing cervical carcinoma.
Most of the multi-subunit RING-finger types of E3 ligases contain a cullin protein (Figure 5), so named because these proteins appear to be involved in a ‘cull’ of proteins by degradation. Each cullin (there are seven different cullins) acts as a molecular scaffold that interacts simultaneously with an adaptor protein Skp1 (S-phase associated protein-1) and a RING-finger protein (Rbx1–Roc1–Roc2) and a specific E2. In turn, Skp1 binds to one of many F-box proteins, the specificity factors that recognize phosphorylated proteins that are to be ubiquitinated by the SCF ubiquitin ligase complex. Different F-box proteins are used to identify different cell-cycle protein substrates at different stages of the cell cycle, adding to the regulatory power of the ubiqutin-ligase system to control the cell cycle.
The APC/C ubiquitin ligase is a more complex variant of the SCF ubiquitin ligase. The catalytic core of the APC/C consists of a cullin-like protein (Apc2), a RING-finger protein (APC/C11) and an E2. In addition to an F-box protein, the APC/C contains ten proteins that appear to create a hollow particle that functions as a scaffold for the cullin-like protein, and probably controls protein-substrate entry to the ligase complex. Another distinguishing feature of the APC/C is that the complex has two alternative substrate selection subunits, Cdc20 (cell-division cycle 20) and Cdh1, that are critical for substrate selection at different stages of the cell cycle. The comparative structures of the complexes are shown in Figure 5.
Kinase regulation and the cell cycle
As shown in Figure 4, each phase of the cell cycle is controlled by a complex, yet beautifully orchestrated, set of interactions between Cdks and E3s . The kinases set the pace of the cyclic activities, with the ligases acting as executioners of key players to maintain irreversible progression. The kinases have an extra tier of regulation, i.e. the kinase inhibitors, p21 and p27. These inhibitors are crucial for kinase regulation and are, in turn, substrates for ubiquitination and degradation, so that kinases become active on demand as required for cell-cycle progression. The key kinases, Cdk1 and Cdk2, are present, but inactivated by these inhibitors in G1. At the G1–S checkpoint, DNA synthesis starts and duplication of the centrosome takes place. The centrosome is the microtubule-organizing centre adjacent to the nucleus that controls the microtubules of the mitotic spindle. These events are accompanied by a large increase in the activity of Cdk2 and subsequent phosphorylation of Cdk2 substrates, peaking during DNA synthesis and centrosome duplication in G2, after which Cdk2 is inactivated. The activities, of Cdk2 can be complimented by low Cdk1 and possible further kinase activities i.e. cell-cycle progression can continue in the absence of Cdk2. Redundancy of kinase activities is a feature of cell-cycle control, as in other critical regulatory kinase cascades in the cell.
Coincident with the decline in Cdk2 activity in G2, there is a large increase in Cdk1 activity to create the phosphorylated proteins necessary for mitosis. Anaphase (where each duplicated chromosome separates and moves apart) is defined by a sudden decline in Cdk1 activity. The activities of Cdk1 and Cdk2 are maintained at low levels until the next G1–S transition. The known roles of the APC/C and SCF ubiquitin ligases in these transitions are shown in Figure 3. A key feature is that the ligases ubiquitinate not only key protein substrates, but also each other, to attenuate respective ligase activities in the cycle. The interplay of the APC/C and SCF ligases is at the heart of the regulation of the cell cycle. The cell-cycle proteins are synthesized in anticipation of their eventual need for cycle progression. However, the cycle can idle because of the ubiquitin-dependent elimination of cell-cycle-protein targets to keep cycle proteins at low levels. In this way, cell-cycle progression is kept in check, at least in normal cells. In cancer cells it is a different story.
In G1, the levels of the p21 and p27 inhibitors are high and levels of cyclin A (that controls both Cdk1 and Cdk2) and cyclin B (that controls Cdk1) are low. The reason is that the APC/CCdh1 ligase (the F-box substrate receptor of a ligase, e.g. Cdh1, is always represented by a superscript) is active, and destroys the SCF ubiquitin ligase, preventing degradation of p21 and p27 Additionally, APC/CCdh1 degrades cyclin A and cyclin B to keep Cdk1 activity low. When the cycle is triggered by growth factors to move into G1–S, S and G2, the p21 and p27 inhibitors are degraded and cyclins A and B accumulate to drive up the activity of Cdk2. The concentration of Cdk1 is kept low. A key event is the transcription and translation of the Emi1 (early mitotic inhibitor-1) protein. This is driven by the adenovirus E2 promoter-binding transcription factor E2F (see Figure 6), which also up-regulates the synthesis of cyclins A, B and E. The Emi1 protein inhibits APC/CCdh1 activity. This decrease in APC/C activity increases the concentration of SCFSkp2 and cyclins A and B, which results in the degradation of p21 and p27 and the activation of Cdk2. This raises the question of why Cdk1 does not accumulate. The reason is that another ubiquitin ligase, SCFβ-TrCP (SCF β-transducin repeat-containing protein), enters the scene. This ligase ubiquitinates the Cdc25a phosphatase, which is degraded. Therefore, Cdc25a cannot dephosphorylate Cdk1, which is maintained inactive by the kinase Wee1. The end result is accumulation and high activity of Cdk2–cyclin E and Cdk2–cyclin A to drive DNA synthesis and chromosome replication. The G2–M transition is accompanied by a decrease in the concentration of cyclin E after ubiquitination of cyclin E by another ubiquitin ligase, SCFFbw7. The APC/CCdh1 is still inhibited, as well as APC/CCdc20, by Emi1, but the concentration of Cdk1 now increases because of an increase in the activity of Cdc25a, which is only degraded in S and G2 by SCFβ-TrCP and by the demise of Wee1 kinase, again through the ubiquitination activity of SCFβ-TrCP. In late mitosis, both Cdk1 and Cdk2 are attenuated. The block on both APC/Cs is now lifted by the ubiquitination of Emi1 by SCFβ-TrCP and its degradation. This results in activation of APC/CCdc20, which ubiquitinates securin, an inhibitor of the separase protease, that cleaves the cohesins and facilitates sister-chromatid separation. The Cdc20 protein needed for securin ubiquitination is predominately expressed late in the cell cycle, which is consistent with Cdc20 functioning in the metaphase–anaphase transition, during which chromosomes become arranged at the cell midline (metaphase) and then separate (anaphase). Significantly, Cdc20 becomes ubiquitinated and degraded once its functions are over, limiting the essential functions of Cdc20 to this part of the cell cycle .
A crucial point to emphasize is that the APC/C and SCF ligases are activated in different ways to integrate with the kinase activities of the cell cycle. The APC/C ubiqutin ligase is activated by phosphorylation, whereas the SCF-ubiquitin-ligase substrates are activated by phosphorylation . No doubt, other protein cofactors are involved in the activities of the ligases, and these cofactors will contribute to the exquisite control of the cell cycle, which is so necessary to avoid genomic instability and cancer.
The APC/C promotes cyclin degradation in order to exit mitosis. The re-accumulation of cyclin A causes the inactivation of the APC/C and entry into S-phase. However, it was not clear until recently, how cyclin A could accumulate in the presence of active APC/C. This problem has been solved by the demonstration that, during G1, APC/CCdh1 autonomously becomes inactivated to allow cyclin A accumulation . This is achieved by auto-ubiquitination and degradation of the E2 UbcH10 (ubiquitin-conjugating enzyme H10), that serves APC/CCdh1. It appears that other (possibly numerous) protein substrates of APC/CCdh1 can prevent the ubiquitination of UbcH10, as long as these protein targets remain in G1. After these substrates are degraded, UbcH10 becomes ubiquitinated and degraded, and cyclin A accumulates and causes entry into S phase. These data are interpreted to mean that alternating and self-perpetuating destruction of the E2 is central to the down-regulation of APC/C, allowing cyclin A accumulation and perpetuation of the cell cycle .
Note, however, that new data on the kinases regulating the cell cycle emerge continuously. For example, in the last few years, protein kinase B/Akt has come to centre stage as a cell-cycle regulator . No doubt, interactions between protein kinase B/Akt and the UPS will emerge in the future.
The tumour suppressor gene product pRb (retinoblastoma protein)
Cell-cycle progression from G1 through S and M clearly is controlled by protein phosphorylation and ubiquitination. There are several extracellular and intracellular cues that orchestrate these cyclical events. The replication of DNA and division of cells needs to be tightly regulated to prevent anomalous cell division and cancer. One method of regulation is through tumour suppressors. The first tumour suppressor to be discovered was pRb. Germ-line mutation in one allele of the pRb gene, together with somatic mutation in the other allele, causes retinoblastoma.
pRb is a key regulator upstream of the phases of the cell cycle, as it controls the E2F transcription factors (Figure 6) that control the expression of DNA-synthesis genes and collaborating gene products, e.g. the cyclins. The E2F transcription factors are complexed with active pRb. pRb can be hyperphosphorylated by Cdk4 and Cdk6. Normally, the activities of these kinases are blocked by p21 and 27, plus the inhibitors of kinases (INKs). Hyperphosphorylation of pRb causes inactivation of pRb, releasing E2F, with subsequent transcription of genes for proteins involved in DNA synthesis and cell-cycle progression . Several viral proteins can interfere with the tumour-suppressor function of Rb and trigger DNA synthesis, and therefore facilitate viral replication and cancer, e.g. simian virus 40. Once the p21, p27 and INK blocks on Cdk4/6 are relieved and E2F activation of the expression of cyclin E–Cdk2 is commenced, cells are sometimes said to have crossed the ‘restriction point’. The cells will proceed with DNA synthesis and chromosome replication with little upstream control, as is apparent in cells with mutant pRb that have lost the ability to restrain E2F. This results in cancer in the eye.
DNA repair and p53
The synthesis of DNA is not without error, particularly when subjected to environmental interference, e.g. radiation and carcinogens. Cells have elaborate DNA-repair mechanisms to repair different sorts of DNA damage, e.g. single-strand breaks or deletions. Protein ubiquitination and sumolylation (SUMO is a cousin of ubiquitin) are involved in the molecular machines that repair DNA damage. Additionally, protein ubiquitination is involved in the regulation of transcription, e.g. in chromatin remodelling through histone ubiquitination. The ubiquitination of proteins involved in DNA repair and gene expression again places ubiquitin at the hub of activities involved in genome stability, which becomes deranged in tumour cells .
The cell cycle pauses during DNA repair. If the DNA cannot be repaired then DNA synthesis and cell-cycle progression are aborted and apoptosis (programmed cell death) takes place. The DNA-repair process is regulated by a transcription factor called p53 . The expression of p53 is dramatically increased in cells subjected to physical or chemical insults, affecting DNA quality. The p53 transcription factor activates the transcription of genes for proteins involved in DNA repair and apoptosis, e.g. caspases. If the DNA cannot be properly repaired then the apoptotic cascades are activated and cells with irrevocably damaged DNA are eliminated. Approximately 50% of human tumours have mutated p53 which is unable to cause apoptosis. Even with normal (wild-type) p53, any compromise of the transcription of DNA repair and apoptotic genes will be deleterious for the cell. The concentration of p53 is regulated by transcription and translation in response to DNA damage and other toxic cell stresses, and also by degradation. The ubiquitination of p53 is tightly controlled by several ubiquitin ligases, but the major ligase is called Mdm2 . This solitary RING ligase is mutated in many tumours. Clearly, if p53 cannot function properly, then apoptosis will be impaired and cells will lose the option of p53-dependent apoptosis. Tumour cells, through genomic instability and alterations in response to intracellular and extracellular regulatory cues, progressively become autonomously dividing cells independent of any method of preventing cell division. In the absence of functioning p53 they are deprived of one mechanism of apoptosis. If the cells cannot deliberately die, they may acquire more genomic damage and become rampantly dividing cells. Such cancer cells may move throughout the body causing metastases. Metastatic cell growth is a characteristic of cancer progression. Many current treatments, e.g. chemotherapy and radiation, seem to activate p53-dependent apoptotic mechanisms in tumour cells, retaining active p53.
Growth factors and tumorigenesis
Cell division is controlled by extracellular growth factors in multicellular organisms. There are several signal-transduction pathways activated by the growth factors. The MAPK (mitogen-activated protein kinase) system is a major regulator of cell division . The MAPK system consists of several parallel biochemical pathways (modules) that can be activated by growth-factor receptors. The general principle is that a MAPKKK (MAPK kinase kinase) phosphorylates a MAPKK (MAPK kinase), which phosphorylates a MAPK to drive cell-cycle events in the nucleus. For example, Raf kinase (MAPKKK) phosphorylates MEK1/2 [MAPK/ERK (extracellular-signal-regulated kinase) kinase] (MAPKK) that phosphorylates ERK1/2 (MAPK) to drive cell division.
The receptors that drive the MAPK pathways include the epidermal growth factor receptor and the platelet-derived growth factor receptor, and several other RTKs (receptor tyrosine kinases). The activity of these receptors must be switched off (down-regulated) to control cell division. One mechanism involves the ubiquitination of the cytosolic tails of the receptors as well as several of the downstream adaptor proteins that relay the activation signal to the nucleus to cause cell division. Receptor ubiquitination is part of the signalling system that causes receptor-mediated endocytosis to bring the receptors and associated proteins into the cell for degradation in endosome–lysosomes . The ubiquitination of the cytosolic tails of the receptors is a crucial event in down-regulating the potent activity of these growth factor receptors. The ligands for these receptors often induce receptor oligomerization, which then activates the kinase pathways. Oligomerization of mutant forms of the receptors in the absence of ligand can trigger activation of the kinases and cause uncontrolled cell division, as occurs in several tumours, including breast cancer. Mutant ubiquitin ligases, with decreased catalytic activity directed at receptor tails, can prevent receptor internalization into endosomes and cause continuous kinase signalling and cancer. This occurs with mutated Cbl, a RING-finger ubiquitin ligase involved in RTK down-regulation . Receptor ubiquitination is additionally controlled by deubiquitination. Mutations in the DUB (deubiquitinating enzyme) and tumour suppressor CYLD, which deubiquitinates proteins in the IKK [IκB (inhibitory κB) kinase complex] and negatively regulates NF-κB (nuclear factor κB) signalling, inhibits apoptosis and causes cylindromatosis, an autosomal-dominant skin-tumour condition . A twist, emphasizing the importance of protein ubiquitination in signal transduction pathways, is that Raf-regulated MEKK1 (MEK kinase 21) is not only a kinase but also a ubiquitin ligase .
The attachment of a chain of ubiquitins to a target protein via Lys63 of each ubiquitin is not a degradation signal, but a signal for kinase activation, e.g. in the NF-κB regulatory pathway involved in activating the immune and inflammatory responses. The kinase pathway adaptor enzyme A20, a potent inhibitor of the NF-κB pathway, contains an N-terminal Lys63-linked ubiquitin-chain deubiquitinating activity to deubiquitinate the RIP (receptor-interacting protein), and a C-terminal ubiquitinating activity to synthesize Lys48-linked ubiquitin chains on RIP . This catalytic strategy ensures the removal of Lys63-linked ubiquitin chains and the construction of the degradation-specifying Lys48-linked ubiquitin chains (perhaps simultaneously) to ensure efficient control of the NF-κB pathway. Tight regulation of the NF-κB pathway is again needed, since, for example, activation of the NF-κB pathway prevents apoptosis and could therefore contribute to cancer progression.
UPS defects in tumour cells
The UPS has multiple roles in the regulation of growth-signalling pathways, from cell-surface receptors and MAPK pathways, to the control of DNA replication and repair, chromosomal separation and cytokinesis. Therefore, mutations in genes of the UPS are increasingly seen to contribute to tumour progression and metasteses . For example, mutation in the F-box-related receptor, the VHL (von Hippel–Lindau) tumour-suppressor protein that is a component of a CDL (cullin-dependent ligase), is causative of renal cell carcinoma . The motif in a target protein recognized by F-box-like protein receptors for ubiquitin-ligase activity is not always a phosphorylated residue: the VHL protein recognizes a hydroxylated proline residue in a target protein. The most well-characterized protein substrate of the VHL receptor is called HIFα (hypoxia-inducible factor α). Mutations in the VHL protein prevent recognition of the motif in the target protein by the CDLVHL, preventing regulated degradation of the HIF transcription factor. This results in continuous activation of the expression of HIF target genes, including that encoding VEGF (vascular endothelial growth factor), causing dysregulated endothelial cell proliferation. The pathological hallmark of renal cell carcinomas and their metasteses is high vascularization, because HIF cannot be degraded properly. Mutations in the F-box protein Skp2, and in other components of the ubiquitin ligases controlling the concentration of cyclins, e.g. cyclin E, are also oncogenic .
The subtleties of the UPS regulation of the cell cycle are illustrated by the roles of the first liver oncoprotein, gankyrin. This protein is a newly discovered 26 S proteasome subunit, but is also present in other protein complexes, including a complex containing Cdk4 that regulates the phosphorylation status of pRb and controls its degradation. Gankyrin was discovered in a yeast two-hybrid screen, as a specific interactor with one of the six ATPases (S6b) in the base of the 19 S regulator of the 26 S proteasome , and by subtractive hybridization, to detect genes overexpressed in hepatocellular carcinoma . Gankyrin is a promiscuous ankyrin-repeat protein and, besides being present in complexes with pRb, is also in a complex with the ubiquitin ligase Mdm2. Consequently, gankyrin, via Mdm2, regulates the ubiquitination of p53. Overexpression of gankyrin causes ubiquitination and degradation of p53. The reduced concentration of p53 prevents apoptosis of liver cells that are experiencing growth stimulation and genomic instability. This eventually results in hepatocellular carcinoma . Gankyrin, by binding to the 26 S proteasome via the S6b ATPase, may bring complexes containing Mdm2/ubiquitinated p53 (and pRb and other cell-cycle regulators) to the proteasome to facilitate the rapid unfolding of these proteins and their degradation. There are a number of other ‘go-betweens’ that can ferry ubiquitinated proteins to the 26 S proteasome for degradation , as illustrated in Figure 7.
Conclusion: the UPS and novel therapeutic intervention
As they say, ‘the proof of the pudding is in the eating’: could potent drugs that inhibit specific components of the UPS have a role in the treatment of cancer? In a rather counterintuitive manner, an inhibitor (peptide boronic ester) of the catalytic activity of the proteasome named bortezomib (Velcade) is very efficacious in the treatment of intractable multiple myeloma . This drug probably works by preventing the degradation of ubiquitinated IKK in myeloma cells and, therefore, preventing the anti-apoptotic activities of NF-κB in myeloma cells.
Intuitively, inhibitors of specific SCF ubiquitin ligases, e.g. enzymes targeting the p21 and p27 inhibitors of the Cdks, may have much more generalized effects on tumour-cell growth. The principle behind drug discovery (so far) is to find small compounds (less than approx. 1 kDa) that will bind to small structural pockets in proteins, such as the ligand binding site of a receptor, and prevent some biological activity, e.g. enzyme catalytic activity. The next challenge for the UPS is to determine whether small molecules can be found to block the activities of SCFs where the transfer of ubiquitin from the E2 to the target protein substrate requires a relatively large surface of the Rbx–Cul1–Skp1 F-box complex. Recently, compounds have been discovered that can block protein–protein interactions, e.g. Mdm2–p53. However, these compounds (Nutlins) rely on a structurally defined rigid area on Mdm2 . The F-box binding sites for post-translationally modified residues in substrates might be suitable targets for interaction with small compounds, e.g. sites that bind phosphorylated residues, hydroxylated prolines and N-glycosylated residues in target proteins.
There is certainly the need for a chemical gear-change in drug discovery to find such drugs, since much of the intracellular chemistry of life involves protein–protein interactions generated by relatively weak non-covalent bonding over long distances on the protein surface.
The involvement of the UPS in probably all regulatory processes in the cell, including cell division and therefore cancer, has opened up a totally new view of cell physiology. It has been known for decades that exquisitely controlled extracellular proteases regulate vascular homoeostasis — blood clotting and complement activation. It is interesting that highly regulated intracellular proteases — the UPS and caspases — control cell life (division) and cell death!
The regulated cell cycle is at the heart of normal cell division.
Dysfunctional regulation of the cell cycle is at the heart of cancer.
The regulation of cell division from growth factor receptors to DNA replication, repair and chromosomal separation is controlled by phosphorylation/dephosphorylation and by ubiquitination/deubiquitination.
Mutations in kinases/phosphatases and ubiquitin ligases/DUBs cause cancer.
Drugs that target kinases and the UPS are effective in treating cancers.
The Biotechnology and Biological Sciences Research Council, European Union, Wellcome Trust and Alzheimer's Research Trust are thanked for support of some of the work described.
- © 2005 The Biochemical Society