The Ubiquitin Proteasome System Protein Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The ubiquitin proteasome system (UPS), of which the neuronal proteasome is a central part, is an important mechanism for protein degradation in eukaryotic cells. It is integrally involved in many cellular processes, and perhaps most importantly in the brain, is involved in regulating the development of synaptic connections and synaptic plasticity. Recent studies (Djakovic et al., 2009; Bingol et al., 2010; Tai et al., 2010) have shown that the proteasome can be regulated by synaptic activity. This study investigated whether glutamate receptor stimulation was involved in the regulation of the proteasome by treating cultured NG108-15 cells with various intracellular second messenger system agonists and antagonists and measuring the effect they had on three types of proteasome activity: chymotrypsin-, trypsin- and PGPH-like. This study found that the three best characterised catalytic activities of the neuronal proteasome were differentially affected by glutamate receptor stimulation and intracellular signalling pathways. Cultured cells were exposed to glutamate receptor agonist NMDA and within 4 hours this resulted in a significant decrease in proteasome activity. Further investigations suggest that regulation of the proteasome is partly mediated by the interaction of several NMDA receptor mediated, calcium activated, intracellular signalling pathways. The PKG pathway has a possibly stimulatory effect on proteasome activity and while the roles of the CaMKII and PKC pathways are not entirely clear, they possibly have a role in suppression of proteasome activity. Although the results are not entirely conclusive they are consistent with the hypothesis that the neuronal proteasome can be regulated by intracellular second messenger systems.


The proteasome is an integral part of the ubiquitin proteasome system (UPS) which is responsible for the majority of cellular protein degradation (De Martino and Slaughter, 1999). It is involved in a number of cellular processes and serves many important functions: not least, it is heavily involved in the function and plasticity of synapses. Understanding how the neuronal proteasome is regulated is of vital importance, as it is clear that proteasome activity, and hence the regulated proteolysis of many synaptic proteins can be controlled in an activity-dependent manner. Previous studies have demonstrated specific examples of intense regulation of the neuronal proteasome. This study aims to investigate how chymotrypsin-, trypsin- and PGPH-like proteasomal activities are regulated by glutamate receptor stimulation and intracellular second messenger systems.

Ubiquitin Proteasome System

When ubiquitin, a 76 amino acid protein, is covalently attached to a substrate protein, it tags it for proteasomal degradation. This is a very complex and highly regulated process. The activity of the UPS can be summarised in six main steps, as seen in Figure 1. The first four steps are a dynamic E1-E2-E3 enzymatic cascade that result in the binding of the activated ubiquitin to the substrate protein by an isopeptide linkage.

Firstly, ubiquitin is activated by an ubiquitin-activating enzyme (E1). The E1 enzyme generates a high-energy thioester intermediate, E1-S~ubiquitin, in an ATP-dependent reaction.

This, secondly, encourages the association of the E1 enzyme with an ubiquitin carrier protein, known as an E2 conjugating enzyme, by causing a conformational change in the E1 enzyme. The activated ubiquitin is transferred to the E2 enzyme which then dissociates from the E1 enzyme (Huang et al., 2007).

Thirdly, the activated ubiquitin is transferred to a substrate specific E3 ligase enzyme. In mammals there are hundreds of different E3 enzymes which are largely responsible for ubiquitination specificity, due to the large variety of E2/E3 combinations. Each separate combination will target a specific group of proteins (Ciechanover, 1998). The E2/E3 combination determines the number and type of ubiquitin chain linkages that are possible (Kim et al., 2007).

Fourthly, an isopeptide linkage is created between the carboxyl group of the ubiquitin molecule and the lysine of the target protein and this facilitates the transfer of the ubiquitin to the substrate protein. This is part of the one reaction between the E3 ligase and the substrate protein-E2 ubiquitin complex.

Fifthly, the substrate protein with the attached polyubiquitin chain is bound to the ubiquitin receptor subunit in the 19S complex of the mammalian 26S proteasome. The mammalian proteasome, 26S, is made up of a central, catalytic component (20S) and two regulatory complexes (19S) (Glickman and Ciechanover, 2002) although these can be substituted by other regulatory complexes. Once attached to the proteasome the protein can be degraded to shorter peptides by the 20S complex. The proteasome is almost exclusively recognises proteins that have been tagged with ubiquitin by the UPS, as described above.

The 20S proteasomal core is a 700 kDa complex, that consists of two α rings and two β rings, each with 7 subunits making a total of 28 subunits. The α and β subunits are heptameric rings, stacked axially in the order: α1-7β1-7β1-7α1-7 (Haas and Broadie, 2008). Within the β subunits lie three important catalytic sites responsible for chymotrypsin- trypsin- and post-glutamyl peptidyl hydrolytic (PGPH) like activity. Although the α rings are catalytically inactive they are important for stabilisation of the inner β rings and in the binding of the 20S complex to the 19S regulatory complexes (Ciechanover, 1998). The 26S proteasome exists in several forms as several of the β subunits can be replaced by inducible subunits. For example, after the cell has been stimulated with interferon-γ. β1, β2 and β5 can be replaced by LMP2, MECL-1 and LMP7 respectively (Eleuteri et al., 1997).

The binding of the 19S (PA700) regulatory complexes is an ATP dependent reaction. The complex contains 20 subunits, 6 ATPase (Rpt1 to Rpt6) and 14 non-ATPase (Rpn1 to Rpn14). Rpn10 binds to polyubiquitin via an ubiquitin interaction motif (Pickart and Cohen, 2004) and is responsible for substrate recognition in the proteasome, although it is believed that a supplementary, presently unknown, subunit may also be involved (Ciechanover, 1998). 19S complexes are additionally involved in the regulation of substrate entry into the proteasome (DeMartino and Gillette, 2007). It is thought that substrate entry is dependent on an ATP-dependent structural rearrangement of the 20S proteasome complex subsequent to the association with the 19S complex, as there are no obvious entry pores in the 20S proteasome complex unlike in the Thermoplasma proteasome which contains an entry pore at each end of the cylinder (Ciechanover, 1998).

Several 11S (PA28) regulatory complexes may also replace the 19S regulatory complexes present to associate with the 20S proteasome: PA28α, PA28β and PA28γ (Hill et al., 2002). They associate in an ATP-independent reaction. However, the 11S-20S-11S complex will only digest peptides as it does not bind to polyubiquitin chains and so cannot digest the ubiquitin-conjugated intact proteins. It acts downstream to the 26S proteasome and further breaks down large peptides created by the 26S complex. A hybrid, 19S-20S-11S proteasome also exists and has higher levels of proteolytic activity as it is able to perform proteolysis on polyubiquitinated proteins and then further break them down, as described above (Hendil et al., 1998; Cascio et al., 2002).

Finally, there is a recycling of ubiquitin. Importantly, this releases ubiquitin and restores the cellular ubiquitin pool. Ubiquitin is released from the Lys residues of proteolytic products by de-ubiquitinating enzymes which can be divided into two main classes: ubiquitin C-terminal hydrolases (UCH) and ubiquitin-specific proteases (UBP), isopeptidases. UCHs are generally involved in the release of ubiquitin from small molecules: for example, amines. UBPs are much larger enzymes, ~100 kDa (as opposed to ~25 kDa) that catalyse the release of ubiquitin from larger cellular proteins or free polyubiquitin chains.

Neuronal Proteasome

Relatively little is known about the neuronal proteasome and how its activity and regulation differs to that in the rest of the body. Tai et al. (2010) purified 26S proteasomes from rat cortex and indentified the standard 26S subunits and 28 proteasome-interacting proteins, thought to be regulators or cofactors. These differed to those found in other tissues - for example, rat muscle. Proteasome degradation serves a significantly different function in the neuron than in muscle and protein composition is also different in the two tissues which is consistent with the differences in proteasome-interacting proteins found in the two tissues. The level of doubly-capped 26S proteasomes was found to be significantly higher in the cortex than in other tissues such as the kidney and liver.

Furthermore, there is differential proteasome composition even within the brain. The ratio of doubly-capped 26S proteasomes to singly-capped (or hybrid) 26 proteasomes was significantly higher in cytosolic tissue than in synaptic tissue (Tai et al., 2010).

Proteasome activity decreases with age. Keller et al. (2000) showed that all three types of proteasomal activity (chymotrypsin-, trypsin- and PGPH-like) decrease with age. A study by Zeng et al. (2005) studied catalytic activity of the proteasome within the brain and found differential decreases in proteasomal activity between different brain regions with age. Proteasomal activity was most significantly decreased in the substantia nigra which may be linked to a higher level of basal oxidative stress (Zeng et al., 2005). The striatum, globus pallidus and frontal cortex also showed a significant decrease in proteasomal activity. Catalytic inactivation is an age-related event, but some tissues are evidently more vulnerable to alterations in proteasomal activity than others.

The effect that regulatory factors have on the neuronal proteasome is not well understood but it is believed they may stimulate the breakdown of specific substrates. For example, PA28γ is thought to stimulate the ubiquitin-independent breakdown of the SRC-3, a transcriptional coactivator (Li et al., 2006). It is possible that the singly-capped 26S proteasome when associated with various regulatory complexes has a more specialised role within different subcellular locations - for example, the synapse (Tai et al., 2010).


Posttranslational modifications have been shown to have an effect on the proteasome. For example, glycosylation or phosphorylation have an effect on the proteasome in muscle cells, amongst others, but it is unknown whether the same effect is seen in the neuronal proteasome (Glickman and Raveh, 2005; Schmidt et al., 2005). Bose et al. (2004) found that phosphorylation of the α subunit, C8, by protein kinase CK2 was essential for the formation and also the stability of the 26S proteasome. Phosphorylation is also believed to affect the activity of the proteasome within the cell. Bardag-Gorce et al. (2004) demonstrated that ethanol ingestion causes inhibition of chymotrypsin-like activity (amongst others) of the purified 26S proteasome in the liver which resulted in the hyperphosphorylation of α subunits α3/C9 and α7/C8. Evidently, phosphorylation and other metabolic processes have a clear role to play in the regulation of the proteasome in the kidney, liver and other tissues (Zhang et al., 2007) and possibly, therefore, in the neuronal proteasome.

The proteasome is responsive to the needs of any one cell and concentrations of proteasome are localised to concentrations of substrate. Furthermore, protein degradation and the UPS have been show to regulate changes in synaptic strength, underlying synaptic plasticity. It is probable, therefore, that the UPS itself is regulated by synaptic activity.

Djakovic et al. (2009) demonstrated rapid and dynamic regulation of the neuronal proteasome by synaptic activity using drugs such as tetrodotoxin (TTX) to put a blockade on action potentials in hippocampal neurons, resulting in an inhibition of proteasomal activity. Bicuculline, however, up-regulated action potentials and resulted in significantly increased proteasomal activity. This was a result, at least in part, of the entry of calcium through NMDA receptors and L-type voltage-gated calcium channels (VGCCs). This in turn, was dependent on the activity of calcium/calmodulin-dependent protein kinase II (CaMKII). CaMKIIα acts to phosphorylate Rpt6, a subunit of the 19S regulatory complex of the 26S proteasome. Proteasome function can be significantly increased, therefore, by neuronal activity and it is apparent that CaMKII is a key regulator in this process.

Neuronal activity affects the activity of the UPS in the dendrite as well as the synapse (Bingol and Schuman, 2006). As a result of synaptic stimulation there is redistribution of proteasomes, from dendritic shafts to synaptic spines. This is thought to be dependent on the activity of NMDA receptors. Therefore, synaptic activity regulates proteasomal activity locally, within the dendrites as it stimulates both the recruitment and also the sequestration of proteasomes which results in a redistribution of the protein composition of the synapse.

Tai et al. (2010) show that the UPS activity of a neuron, as a whole, can be affected by neuronal activity. Whole-cell lysates of neurons treated with NMDA showed a decrease in the level of ubiquitin conjugates. This is possibly a result of either decreased levels of ubiquitination or increased levels of substrate degradation. A further result of NMDA treatment was decreased levels of 26S proteasomes, thought to be a result of degradation of 19S complexes. When 19S particles dissociate from the 20S catalytic core, they can then be degraded by the proteasome, when induced by NMDA. This is a possible mechanism for suppression of proteolysis. Tai et al. (2010) also demonstrate selective degradation of the 19S complex which results in a shift to 20S proteasome complexes from 26S complexes but the detailed mechanism is not fully understood. However, it is thought that the 20S proteasome complex alone cannot degrade ubiquitinated proteins in vivo as entry of protein substrates into is proteolytic chamber is blocked by the N-terminal residues of the α-subunits which make a "gate" (Smith et al., 2007). There is important evidence to suggest that the degradation of unfolded/oxidant-damaged proteins is dependent on a ubiquitin requiring mechanism that involves the 26S proteasome and VCP/p97 complexes (Medicherla and Goldberg, 2008). It is therefore unclear whether an increased number of individual 20S proteasome complexes improves the proteolytic function of a neuron.

There is an increasing evidence to suggest that the composition and, presumably, function of the neuronal proteasome may be regulated by levels of closely associated proteasome-interacting proteins. Perhaps most important are three E3s (KCMF1, HUWE1 and UBE3A) and five DUBs (USP5, USP7, USP13, USP14 and UCH37) (Tai et al., 2010) that are thought to improve the efficiency of proteasome function by determining substrate specificity or ensuring that there is rapid removal of released ubiquitin chains as they inhibit proteasomal activity (Finley, 2009).

There are hundreds of E3s and yet only a very few of them are associated with the neuronal proteasome and so can be considered to be important in its regulation. Knockout mice with UBE3A mutations show aberrations in the morphology of the synapse, glutamate receptor endocytosis and long-term potentiation (Greer et al., 2010). UBE3A mutations are also linked to Angelman syndrome, a genetic disorder associated with mental retardation. HUWE1 mutations are linked to synaptic dysfunction and X-linked mental retardation. Synaptic defects in both UBE3A and HUWE1 mutations are possibly caused by altered function of the neuronal proteasome (Tai et al., 2010). When the 26S proteasome is exposed to NMDA, both UBE3A and HUWE1 dissociate and are degraded - a possible mechanism for regulation of the ubiquitin proteasome system.

In mammals both USP14 and UCH37, DUBs have been shown to influence protein degradation as a result of their interaction with the 26S proteasome, catalysing the degradation of the ubiquitin chain and also the recycling of free ubiquitin (Koulich et al., 2008). USP14 also has an important regulatory role in "gate" opening in the 20S proteasome with the assistance of the ATPases (Peth et al., 2009). Animal models have shown that mutations of USP14 lead to defects in synaptic ubiquitination - these mice are called ataxia (ax1 mice) (Wilson et al., 2002). Chen et al. (2009) reported ax1 mice with reduced levels of free ubiquitin and ubiquitin conjugates and changes in short-term synaptic plasticity - which may be due to the altered functioning of the proteasome which degraded the protein but did not recycle the ubiquitin.

The neuronal proteasome is clearly the subject of intense regulation and the increasing numbers of proteasome interacting proteins that have been identified suggest that the regulation of the neuronal proteasome is even more complex and diverse than previously thought. In view of the evidence presented, the hypothesis that the activity of the neuronal proteasome may be regulated by glutamate receptor stimulation and intracellular second messenger systems was tested by treating NG108-15 cells with glutamate, NMDA and various signalling pathway agonists and antagonists to investigate the effect of specific second messenger systems on proteasome activity.



Synthetic proteasomal substrates were purchased commercially: succinyl (Suc-LLVY-AMC) and butyloxycarbonyl (Boc-LAA-AMC) from Enzo Life Sciences (UK) Ltd. (Exeter, UK) and acetyl (Ac-GPLD-AMC) from Biomol International, L.P. (Exeter, UK). AMC was purchased from Calbiochem (California, USA). For the protein assay Bio-Rad Protein Assay Reagent (Bio-Rad Laboratories GmbH, München, Germany) and BSA (Roche Diagnostics GmbH, Mannheim, Germany) were used. EDTA and HEPES used to make up the assay buffer were from Sigma Chemical Co. (Poole, UK); ATP was from Enzo Life Sciences (UK) Ltd. (Exeter, UK); MgCl2 from VWR International Ltd. (Poole, UK). Proteasome activity inhibitor, MG-132, was purchased from (Alexis Biochemicals, Lausen, Switzerland). NG108-15 cells (ATCC, Harwell) were treated with glutamate, kainate, A23187 and MK801 (Sigma Chemical Co., Poole, UK); NMDA, AMPA and ACPD (Tocris, Bristol, UK); KN62, KT5720, KT5823, PD98059, GF109203X, SNAP and PMA (Merck, Nottingham, UK).

Cell Culture

All cell culture reagents were from Invitrogen (Paisley). Neuroblastoma x glioma 108-15 (NG108-15) cells (ATCC, Harwell) were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% foetal bovine serum, penicillin/streptomycin and glutamine. Every four days, cells were dislodged using 1% trypsin, centrifuged at 1200g for 5 minutes, and resuspended in DMEM. A proportion of NG108-15 cells were allowed to grow in supplemented DMEM as above, while the majority of cells were seeded into culture plates in differentiation medium: DMEM supplemented with 1.5% dimethyl sulfoxide (DMSO)/0.5% foetal bovine serum, penicillin/streptomycin and glutamine. As reported (Seidman et al., 1996; Muller et al., 2010) under these conditions, NG108-15 cells stopped proliferating and extended neurites.

Cell Treatment

NG108-15 cells in 12 well or 24 well culture plates were treated for either 10 or 30 minutes at 36.4°C with 100μM glutamate which was injected directly into the cell culture medium; for 30 minutes with 100μM NMDA, 100μM AMPA, 100μM ACPD or 100μM kainate; for 240 minutes with 100μM glutamate or 100μM NMDA; for 240 minutes with 100μM NMDA pre-treated with 10μM KN62, 1μM KT5720, 1μM KT5823, 40μM PD98059 or 1μM GF109203X; for 240 minutes with 5μM A23187, 100μM SNAP or 140nm PMA; for 240 minutes with 100μM NMDA pre-treated with 50μM MK801. In each case an equal number of cell wells were treated or pre-treated with an equal volume of vehicle (deionised H20), where appropriate, as a control measure.

Proteasomal Enzyme Assay

NG108-15 cells were placed on ice and homogenised in ice-cold extraction buffer (100μL per well) containing 25mmol/L HEPES (pH 7.6), and 0.5mmol/L EDTA. The homogenate was centrifuged at 13,000rpm at 4°C for 10 minutes to form a supernatant of tissue extract. Protein content of the supernatant was determined using 4μL of each sample and the Bio-Rad micro protein assay system using Bio-Rad Protein Assay Reagent, based on the method of Bradford (Bradford, 1976) and used to calculate results / mg protein.

Proteasomal activity was assessed in the lysates of NG108-15 cells by monitoring the accumulation of the fluorescent cleavage product 7-amino-4-methylcoumarin (AMC) from synthetic proteasomal substrates Succinyl-Leu-Leu-Val-Try-AMC for chymotrypsin-like activity; Butyloxycarbonyl-Leu-Arg-Arg-AMC for trypsin-like activity and Acetyl-Gly-Pro-Leu-Asp-AMC for peptidyl-glutamyl-peptide (PGPH)-like activity.

10µL of each freshly made supernatant was incubated in a 96-well plate at 37°C for 30 minutes with 10µL of 400µM of succinyl-LLVY-AMC substrate (or 400µM of butyloxycarbonyl-LAA-AMC or 400µM of acetyl-GPLD-AMC) and 80µL of assay buffer containing 25mmol/L HEPES (pH 7.6), 5.0mmol/L MgCl2 and freshly added 5.0mmol/L ATP. Each sample was assayed in duplicate.

Release of fluorescent AMC was measured with in a Fluoroskan Ascent microplate fluorometer (Thermo Labsystems) at 460nm with an excitation wavelength of 350nm. The fluorescence of the released AMC was measured every 60 seconds for 30 minutes and the concentration of the liberated products was calculated using a standard curve for AMC (AMC calibration, ranging from 0.125µM to 1µM. Proteasomal activity was determined as the increase in fluorescence of the reaction products and expressed as pmole AMC/min/mg protein.

The specificity of the proteasomal assay was confirmed by the ability of 50µM MG132 to near-totally inhibit fluorescence change.

Data Analysis

Results were expressed as mean ± S.E.M. Statistical analysis was carried out by ANOVA (General Linear Model) followed by a post hoc Tukey test on normally distributed data and the Kruskal-Wallis Test on non-normally distributed data, using Minitab software. Statistical significance was taken at p<0.05 but results approaching significance with p<0.1 have also been highlighted in some cases.


To establish the specificity of the proteasomal assay, MG132, a proteasome activity inhibitor was used resulting in a significant and near total inhibition of changes in fluorescence (see Figure 1).

Glutamate Treatment Stimulates Proteasomal Activity

Cells were first treated with 100µM glutamate for 10 and 30 minutes respectively. Although there was no significant change in activity after 10 minute treatment, there was an increase in proteasomal activity after 30 minutes. The increase was statistically significant overall (ANOVA - F(1,32)=7.60, p<0.05) when compared to vehicle treated NG108-15 cells and was specifically significant for chymotrypsin-like activity where there was an increase of 143.5 ± 15.1% relative to the vehicle control (see Figure 2). There was also a near significant increase in PGPH-like activity of 137.5 ± 18.3% relative to the vehicle control.

Due to the significant increase in proteasomal activity after glutamate treatment, further investigations were performed. NG108-15 cells were treated with several glutamate agonists: NMDA, AMPA, ACPD and kainate, to see if any one of them had a particular effect. There were no significant effects of treatment (see Figure 3) although 30 minute treatment with 100µM ACPD caused a slight decrease in trypsin-like activity (71.5 ± 10.2% relative to vehicle control). This was inconsistent with its effects on chymotrypsin and PGPH-like activity however. Similarly 30 minute treatment with 100µM AMPA caused a small decrease in PGPH-like activity (77.4 ± 9.9% relative to vehicle control). There was also a smaller decrease in chymotrypsin- and PGPH-like activity as a result of 30 minute 100µM kainate treatment (89.3 ± 12.6% and 83.2 ± 8.0% relative to the vehicle control, respectively). Further investigations into the effect of AMPA, ACPD and kainate showed no significant activity (data not shown).

NMDA Treatment Suppresses Proteasomal Activity

As treatment of NG108-15 cells at 30 minutes had an insubstantial effect, the effect of 4 hour 100µM glutamate and 100µM NMDA treatment was investigated to see if the longer time course would have an effect in line with the results of Tai et al. (2010). The increase in proteasomal activity as a result of 4 hour glutamate treatment was consistent with our earlier results and was seen in both chymotrypsin- and trypsin-like activity where there were increases of 131.2 ± 32.6% and 134.1 ± 19.2% respectively, relative to the vehicle control. However, these results did not achieve statistical significance.

Interestingly however, there were decreases in all three types of proteasomal activity after 4 hour 100µM NMDA treatment. The decrease was significant for chymotrypsin- (58.1 ± 14.4% relative to the vehicle control) and PGPH-like activity (18.4 ± 4.6% relative to the vehicle control) (see Figure 4).

Mechanisms of Action for NMDA Receptor Activity

To confirm the decrease in proteasomal activity as a result of NMDA treatment we investigated the action of ionotropic NMDA receptors and calcium activated intracellular signalling pathways more closely. Ionotropic NMDA receptors will only open after stimuli of sufficient strength or frequency depolarise the membrane sufficiently expel the magnesium block in the NMDA channel. When it opens it admits large amounts of sodium and calcium. Calcium acts as a second messenger and activates intracellular signalling cascades (see Figure 5). The mechanisms of several of these pathways are not fully understood as yet. NG108-15 cells were treated with various antagonists specific to the different pathways (see Table 1) to investigate whether any intracellular signalling pathways had a specific effect on the regulation of the neuronal proteasome.

Results confirmed that 4 hour 100µM NMDA treatment resulted in a significant decrease in chymotrypsin-like proteasomal activity (83.4 ± 4.9% relative to the vehicle control) (see Figure 6). However, NMDA treatment caused only a very slight decrease in trypsin-like (92.5 ± 8.2% relative to the vehicle control) and PGPH-like (97.5 ± 7.6% of vehicle control) proteasomal activity.


Concentration Used


Mechanism of Action




Potent and selective protein kinase C inhibitor (Toullec et al., 1991).




Selective, cell-permeable inhibitor of CaM kinase II. Binds directly to the calmodulin binding site of the enzyme (Tokumitsu et al., 1990).




Potent, selective inhibitor of protein kinase A. Has no effect on PKG or PKC. (Cabell et al., 1993)




Selective inhibitor of protein kinase G (Gadbois et al., 1992; Smolenski et al., 1998).




Specific inhibitor of mitogen-activated protein kinase kinase (MAPKK / MEK). Acts by binding to the inactivated form of MEK, thereby preventing its phosphorylation by cRAF or MEK kinase (Dudley et al., 1995).

Pre-treatment with 100µM KN62 appeared to inhibit the effect of the NMDA treatment and restore proteasomal activity back to the vehicle level, even increasing proteasomal activity slightly for trypsin-like proteasomal activity (116.8 ± 19.4% relative to the vehicle control). However, it is possible that part of this effect is due to the action of KN62 alone, as pre-treatment with KN62 followed by vehicle treatment also resulted in an increase in all three types of proteasomal activity, especially noticeable in chymotrypsin-like (126.4 ± 19.7% relative to the vehicle control) and trypsin-like (162.3 ± 27.9% relative to the vehicle control) activity. This suggests that the CaMKII pathway has a possible regulatory role in the suppression of proteasomal activity.

Pre-treatment with 1µM KT5823 suppresses proteasomal activity slightly and has an overall significant effect on chymotrypsin-like activity where it reduces activity to 78.5 ± 7.4% relative to the vehicle control when followed by NMDA treatment and 61.0 ± 5.2% when followed by vehicle treatment. However, for both trypsin-like and PGPH-like activity pre-treatment with KT5823 when followed by NMDA treatment seems to have no significant effect on activity relative to the vehicle control, and decreases in proteasomal activity were only seen when pre-treatment with KT5823 was followed by vehicle treatment (75.5 ± 12.4% and 45.2 ± 13.4% relative to the vehicle control, respectively). Despite this, the PKG pathway has a possible role in stimulation of proteasomal activity.

Pre-treatment with 1µM KT5720 has a very similar but smaller effect to that of KT5823 and so is not significant for any of the three types of proteasomal activity. In almost every case it causes a decrease in proteasomal activity, with the largest being 54.6 ± 11.1% relative to the vehicle control for PGPH-like activity when KT5720 pre-treatment was followed by NMDA treatment.

The results following pre-treatment with 40µM PD98059 are conflicting. For both chymotrypsin-like and PGPH-like activity pre-treatment with PD98059 resulted in a decrease in proteasomal activity and the effect of pre-treatment had an effect approaching significance for PGPH-like activity where there was a decrease of 52.1 ± 9.4% relative to the vehicle control following pre-treatment with PD98059 and treatment with NMDA. However, the effect of PD98059 on trypsin-like activity was quite different, causing a highly significant increase in proteasome activity when followed by NMDA treatment. Although, there was no reason to discount the results, they lie far out of the normal range for the rest of the data and it would be important to increase the n number on this experiment to reduce the variability and to see if further experimentation confirmed earlier results.

Similarly, pre-treatment with 1µM GF109203X resulted in a slight decrease in chymotrypsin-like activity but variable increases in both trypsin-like and PGPH-like activity. This proved to be significant in the case of trypsin-like activity when followed by vehicle treatment. Again, the variability in results would suggest a need to increase the n number to reduce variability and confirm as to whether the first results are truly representative of proteasomal activity following GF109203X treatment.

In an attempt to confirm the results found, cultured NG108-15 were treated with various agonists of the NMDA-receptor, calcium mediated intracellular signalling pathways (see Table 2 and Figure 5). Three pathways were investigated further: the CaMKII, PKG and PKC as results in these pathways were the most promising and possibly significant.

Table 2 | Agonists used to investigate regulation of proteasome activity by NMDA-receptor, calcium mediated intracellular signalling pathways.


Concentration Used


Mechanism of Action




Ionophore highly selective for Ca2+. Potentiates responses to NMDA. In cell culture, stimulates nitric oxide production by calmodulin-dependent constitutive nitric oxide synthase.





Nitric oxide donor that mimics the actions of nitric oxide. Markedly activates soluble guanylate cyclase. 





Activates protein kinase C in vivoand in vitro, even at nM concentrations. Activates Ca2+-ATPase and potentiates forskolin-induced cAMP formation.

The results for chymotrypsin- and trypsin-like activity are broadly consistent and resulted in significant decreases in proteasomal activity following treatment with all three agonists (see Figure 7). Results for PGPH-like activity have not been shown as the signal recorded was too low and so only a couple of accurate results were recorded. Treatment with 5µM A23187, used to recreate stimulation of the CaMKII pathway resulted in a slight decrease in chymotrypsin-like proteasomal activity and a highly significant decrease in trypsin-like activity which is consistent with the results found after treatment with 10µM KN62, suggesting that the CamKII pathway is involved in suppression of neuronal proteasome activity. However, as can be seen in Figure 8 it is possible that the decrease in activity is, at least in part, a result of A23187 induced toxicity. Indications that the cultured NG108-15 cells weren't healthy after A23187 treatment include retraction of neurites and shrunken cell bodies.

Treatment with 100µM SNAP resulted in a highly significant decrease in proteasome activity (46.5 ± 7.7% and 16.1 ± 2.0% relative to the vehicle control for chymotrypsin- and trypsin-like activity, respectively). Treatment with 140nM PMA also resulted in a highly significant decrease in proteasomal activity (46.0 ± 10.5% and 9.5 ± 2.5% relative to the vehicle control for chymotrypsin- and trypsin-like activity, respectively).

Cultured NG108-15 cells were finally treated with 50µM MK801, a highly potent and selective non-competitive NMDA glutamate receptor antagonist that acts at the NMDA receptor-operated ion channel as an open channel blocker. Results for chymotrypsin- and trypsin activity conflicting in relation to the vehicle control and results for PGPH-like activity have not been shown as the signal recorded was too low in almost every case (see Figure 9). Results for a significant decrease in trypsin-like proteasomal activity following NMDA treatment (64.1 ± 6.5% relative to the vehicle control) are consistent with earlier results, unlike the results for chymotrypsin-like proteasomal activity following NMDA treatment.

Treatment with 50µM MK801 does appear to at least partly inhibit the effects of NMDA treatment, bringing the value for proteasomal activity towards the vehicle control value. However, pre-treatment with MK801 followed by vehicle treatment has a similar, though not as strong, effect and so the results are difficult to interpret. As the n number is relatively small it would be worthwhile repeating the experiment to reduce the variability in results.


Several recent studies have demonstrated the fact that neuronal activity influences the activity of the ubiquitin proteasome system, both in dendrites and synapses (Bingol and Schuman, 2006; Djakovic et al., 2009). The present study confirms this.

Glutamate Mediated Stimulation of Proteolytic Activity

Treatment of NG108-15 cells with glutamate resulted in a significant increase in chymotrypsin and PGPH-like activity. However, further investigation into which glutamate receptor mediated this reaction proved inconclusive and treatment of NG108-15 cells with NMDA resulted in a opposing and significant decrease in proteasome activity.

One possible explanation for this effect is that NG108-15 cells have glial characteristics and so are likely to contain large amounts of glutamate uptake sites, and so when the cells were treated with glutamate for 4 hours, this could result in large amounts of glutamate being taken up into the cell. After the induction of LTP, for example, glutamate uptake is rapidly increased and this increase can last for 3 hours or more (Pita-Almenar et al., 2006). The large amounts of glutamate being taken up into the cell could possibly give a false positive result.

One other explanation is that the concentrations of NMDA, AMPA, ACPD and kainate that were used to investigate the mechanism of glutamate action were possibly too low to cause any change in proteasome activity that we would be able to record. However, although the changes noticed were slight, there were changes in proteasome activity relative to the vehicle control and there were also differential rates of activity after treatment with the various glutamate agonists which would suggest that our recordings were accurate and that the concentrations were sufficient to induce change in proteasomal activity. Secondly, the concentrations of glutamate agonists used in this experiment were carefully researched beforehand and the concentrations used were in accordance with other recent studies. It is possible that simultaneous stimulation of a combination of glutamate receptor types is required for the enhancement of proteasome activity 30 minutes later. Further experimentation could test this hypothesis using combinations of agonists.

NMDA Mediated Suppression of Proteasome Activity

The present study demonstrated a significant decrease in chymotrypsin- and trypsin like proteasomal activity following 100µM NMDA treatment which is consistent with and expands upon the work of Tai et al. (2010) who demonstrated a significant decrease in chymotrypsin-like activity following NMDA treatment. They also showed a similar decrease in ubiquitin-conjugate levels after NMDA treatment which would suggest that there is a coordinated reduction in the rate of proteolytic activity during NMDA-induced plasticity. Djakovic et al. (2009) demonstrated similar decreases in proteasomal activity after activity-blockade with tetrodotoxin and Ehlers (2003) demonstrated decreases in ubiquitin-conjugate levels. It is suggested that the decrease in proteasome activity after NMDA treatment is a result of NMDA-induced destruction of proteasomes as a mechanism for suppressing overall proteolysis (Tai et al., 2010). Although the mechanism is not clear, NMDA treatment appears to increase the degradation of 19S particles which reduces proteasome activity as the 20S proteasome itself cannot degrade ubiquitylated proteins. Although the role of the 20S proteasome itself is not clear, it appears that entry of substrate proteins into the proteolytic chamber is restricted by a "gate" which is thought to be formed by the N-terminal residues of the α-subunits (Smith et al., 2007; Tanaka, 2009). There is strong evidence to suggest that the degradation of almost all proteins requires the complete 26S proteasome - i.e. the 20S proteasome with associated regulatory complexes.

CaMKII Mediated Suppression of Proteolytic Activity

The present study found that inhibiting the NMDA receptor, calcium mediated CaMKII pathway stimulated proteasomal activity, suggesting that the CaMKII pathway has a possible regulatory role in the suppression of proteasomal activity.

CamKII is a calcium-dependent protein kinase and is believed to play an important role in neuronal behaviour, development and plasticity (Wayman et al., 2008). Several studies indicate that calcium signalling is very important in the regulation of the neuronal proteasome. However, the data reported here is inconsistent with early work on the proteasome (Aizawa et al., 1996) which suggested that the release of calcium from intracellular stores activated the proteasome, if transiently and also inconsistent with recent studies (Djakovic et al., 2009; Bingol et al., 2010) which suggest that the CaMKII stimulates proteasome activity by phosphorylating a 19S subunit, Rpt6. Activity of 26S proteasomes was shown to be enhanced by the addition of recombinant CaMKIIα in the presence of calcium/calmodulin and ATP which suggests CaMKII mediated phosphorylation of the 19S complex. Mass spectrometry demonstrated strong phosphorylation of the 19S subunit, Rpt6 by CaMKIIα in vitro kinase reactions. Calcium/calmodulin binding and the ensuing autophosphorylation of CamKIIα occurs as a result of calcium influx at stimulated synapses (see Figure 5). When calcium and calmodulin bind, CaMKIIα undergoes a conformational change that leads to its autophosphorylation, which renders he enzyme constitutively active (Lisman et al., 2002). This is a mechanism that provides a mechanism whereby proteasomes could be recruited specifically to activated synapses, where they can then mediate local protein degradation (Bingol et al., 2010).

Although this study would need to repeat the experiment and increase its experiment number to reduce variability and increase the significance of its results, one possible explanation for the CaMKII mediated suppression of proteasome activity is the fact that the CaMKII pathway inhibitor, KN62, appears to have some effect on proteasome activity itself, without interacting with the NMDA receptor-stimulated pathway. Again, it is possible that our results represent a false positive reading.

NO/cGMP/PKG Mediated Stimulation of Proteasome Activity

Inhibiting the PKG pathway resulted in an inhibition of proteasomal activity, suggesting that the PKG pathway may be involved in the regulation of and stimulation of proteasomal activity.

A study by Kotamraju et al. (2006) repeats this result in endothelial proteasomes and demonstrates the upregulation of immunoproteasomes by nitric oxide. They suggest that NO signalling-mediated proteasomal upregulation is due to the induction of LMP2 and LMP7 subunits that are specific to the imunoproteasome. This is in line with other studies that suggest that phosphorylation and other metabolic processes have a clear role to play in the regulation of the proteasome (Zhang et al., 2007). It is also thought that high levels of endogenous NO will account for non-immune functions of the proteasome (Kotamraju et al., 2006). They also presented data to show a significant decrease in both chymotrypsin- and trypsin-like proteasomal activity in the aorta or iNOS-/- mice. This was correlated with a significant decrease in LMP2 and LMP7, immunoproteasomal subunits.

Treatment with SNAP, a nitric oxide donor, thought to activate the PKG pathway, resulted in a highly significant decrease in proteasome activity. Although this is inconsistent with earlier results (following treatment with 1µM KT5823) there are several possible reasons for this. Figure 5 is a very simplified summary of intracellular signalling pathways and in fact the various pathways are extremely complex and interacting. It is difficult, therefore, to find an agonist specific to one pathway only. One possible explanation of the results, therefore, is that the increase in nitric oxide (caused by SNAP treatment) results in stimulation not only of the PKG pathway but another pathway that is involved in the suppression of proteasomal pathway. Alternatively, it is possible that treatment with SNAP induced cell apoptosis which has been recorded, although at much higher concentrations of SNAP (Chew et al., 2003). This, in turn, would be consistent with the extremely low values for trypsin-like proteasomal activity, with the very small SEM despite the small n number of results per group.

Many of the cellular effects of NO are thought to be cGMP dependent (Yan et al., 2003) (see Figure 5). However, intracellular cGMP not only targets PKG but also regulates cAMP levels (via phosphodiesterase modulation) which in turn modulates PKA activity - it is even thought that high concentrations of cGMP may directly stimulate PKA activity (Vandecasteele et al., 2001; Sausbier et al., 2000). This is consistent with the present study which showed both PKG and PKA mediated increases in proteasomal activity although only the data for PKG mediated increases was significant.

Although the data presented by Kotamraju et al. (2006) was in endothelial tissue and there is evidently differential activity between different tissues (Zeng et al., 2005; Tai et al., 2010) it is clear that NO-mediated signalling pathways play a significant role in modulating the up-regulation of proteasomal activity. Several studies show that protein S-nitrosylation can be functionally coupled to ubiquitination and can therefore modulate protein degradation (Yao et al., 2004; Lee et al., 2008). In fact, Yao et al. (2004) demonstrate the link between S-nitrosylation and ubiquitination in Parkinson's disease suggesting that NO has a role to play in the neuronal proteasome as well.

PKC Mediated Regulation of Proteasome Activity

This study also found some evidence that inhibiting the PKC pathway resulted in an increase in proteasome activity. This was consistent with the results of treatment with PMA, an activator of PKC, which resulted in a highly significant decrease in proteasomal activity. Treatment with a PKC inhibitor caused an increase in trypsin-like and PGPH-like proteasome activity which suggests that the PKC pathway might be involved in suppression of proteasomal activity. This is consistent with research into the proteasome in skeletal muscle and endothelial cells, where PKC activation is linked to a down regulation of proteasome activity (Nakashima et al., 2005; Leontieva and Black, 2004). However, the results were highly variable and it is possible that PMA also stimulates the PKC pathway, for example (see Table 2 and Figure 5) which has a possibly very different role to play in proteasomal regulation.

Finally, although our study did not present particularly convincing evidence for the effect of the NMDA-receptor specific antagonist, MK801, Djakovic et al. (2009) used another NMDA receptor antagonist, APV, and a L-type voltage gated calcium channel antagonist, nimodipine, to demonstrate that external calcium influx through postsynaptic NMDA receptors and L-type voltage gated calcium channels is important for activity dependent proteasome function in neurons. The results this study found are broadly consistent with their findings.


In conclusion, the present study found that the three best-characterised catalytic activities of the neuronal proteasome were differentially affected by glutamate receptor stimulation and intracellular signalling pathways. Although the results found are not entirely conclusive they are consistent with the hypothesis that the neuronal proteasome can be regulated by intracellular second messenger systems. Proteasomal activity was significantly affected by the CamKII and PKG and PKC pathways which appear to be important in the regulation of the proteasome. Further investigations will determine the exact mechanisms by which this occurs.


Although the proteasome and its activity have been described for some time, investigations into the differential activity of the neuronal proteasome and its regulation have only taken place relatively recently. I was fascinated by the neuronal proteasome and its importance in the brain for protein degradation and synaptic plasticity amongst other things. Understanding how the neuronal proteasome is regulated is of vital importance, as it is clear that proteasome activity, and hence the regulated proteolysis of many synaptic proteins can be controlled in an activity-dependent manner.

Throughout my investigation I treated cultured NG108-15 cells and then ran proteasomal assays to record levels of proteasome activity. I also ran Bradford protein assays which enabled me to standardise my results.

To locate appropriate references I used information sources on the web such as ScienceDirect, ISI Web of Knowledge and PubMed to search for literature related to my project. Once I had found and read several important papers that were the basis of my background research I was able to start reading papers they had cited and this expanded my reference base. My supervisor also supplied me with several important reference papers at the beginning of my project.

Practically, my supervisor was involved in the culturing of the NG108-15 cells. My supervisor also guided me in the use of GraphPad Prism and Minitab - neither of which I had used to any great extent before - and in the analysis of my results and statistics.

My dissertation is a unique piece of work. Recent studies have investigated the effect of chymotrypsin-like activity whilst looking at the regulation of the proteasome but no study, to date, has investigated the effect of regulatory pathways on all three types of proteasomal activity. Furthermore, whilst current studies have focused one important NMDA-receptor mediated, calcium activated intracellular signalling pathway, the CaMKII pathway, this study aimed to look at the effect of a variety of different intracellular signalling pathways to see if regulation of the proteasome was achieved by several specific pathways, or if the regulation was achieved through the summative effect of many intracellular signalling pathways.