Proteomics Moves From Expression to Turnover

Update and Future Perspective

Mary K Doherty; Phillip D Whitfield


Expert Rev Proteomics. 2011;8(3):325-334. 

In This Article

Advancing Strategies for Protein Turnover Analysis

A number of interesting strategies, including both mass spectrometric and non-MS based approaches, have recently been developed to investigate the biological effects of changes in protein turnover. A summary of the different methods used to define protein synthesis and degradation can be found in Table 1.

Mass Spectrometric-based Approaches

Previous research has led to the development of a technique, termed 'dynamic-SILAC' in which the stability of proteins from a lung adenocarcinoma cell line was investigated and the degradation rates of over 600 proteins were determined experimentally.[42] These data were then used to investigate the effects of protein structure and function on inherent protein stability. By comparing protein size, isoelectric point, cellular location and function and other factors such as predominance of PEST sequence and the identity of the N-terminal residue, it was observed that the only weak correlation with protein stability was the predicted extent of 'order' of the protein tertiary structure. Other researchers have also adapted the SILAC methodology to address specific problems. Schwanhäusser et al. developed a variation on the dynamic SILAC method, which they have termed 'pulsed-SILAC'.[43] Different samples are grown in the presence of a different stable isotope labeled amino acid and the samples from each defined time point are combined after harvest. The relative intensities of the two labels are monitored, allowing the relative rates of synthesis between the two different states to be determined. This approach has been used to investigate a model of iron homeostasis, however, it does not take into account protein degradation, and again the absolute rates of protein synthesis are not determined. In another study, Yee et al. investigated the protein turnover of a recombinant IgG-secreting myeloma.[44] In addition, they looked at the secretion kinetics of recombinant IgG, with the half-life of secretion calculated as being 2.6 h.

A pulse-chase MS approach was used by Bunner and Williamson to explore E. coli 30S ribosome assembly dynamics.[45] In this method, 16S RNA was incubated with a pulse of 15N-labeled ribosomal proteins. This was followed by a chase with an excess of 14N-labeled proteins. The purified 30S particles contained a mixture of 14N- and 15N-labeled proteins. The fraction of proteins that were found to be labeled with 15N represents the fraction of protein bound when the chase period was initiated. This approach does not measure transiently bound proteins or rapid exchange. The deconvoluted data were fitted to a theoretical isotope distribution so that the amplitude of the labeled and unlabeled peaks could be calculated. The averaged relative quantities of 14/15N peptides from the range of time points were plotted and fitted to exponential curves to obtain the binding rate constants of each protein. The order of protein binding was found to be consistent with previous kinetic data, in that proteins that bind directly to RNA were the fastest binders, whereas tertiary binders (those that require at least one primary and one secondary bound protein) were the slowest binders.

Extended SILAC methodologies have also been utilized to investigate cells under non-steady-state conditions.[46] In this approach a combined SILAC/isobaric tag for relative and absolute quantitation (iTRAQ) strategy was used to determine protein degradation rates in Streptomyces coelicolor, a multicellular differentiating bacterium. Proteins were grown in the medium supplemented with labeled arginine and switched to light medium and isolated at defined time points. Proteins from each time point were then labeled with a specific iTRAQ reagent (mass 114–117 Da), and the time points were mixed and analyzed by MS. This allowed estimation of the change in labeled protein concentrations over time without the need for the steady-state assumption. It was possible to estimate degradation rates for 246 proteins using the SILAC approach, with 115 determined using the combined SILAC/iTRAQ approach. Of these, 69 were found in both analyses, allowing comparisons to be drawn between the methods. For proteins that were found to be 'less dynamic', good agreement was obtained between the two approaches, although the authors felt that there was divergence for 'more dynamic' proteins. In addition, as complete labeling of the proteome was not obtained before the chase period was initiated and this was not fully compensated for in the subsequent analyses, there should be considerable caution attributed to the absolute rates of degradation determined.

In 2008 Rao et al. investigated the global protein turnover profile under acid shock and iron starvation conditions, but in Mycobacterium smegmatis.[47] In this instance, protein turnover was defined using a synthesis/degradation ratio. Cells were labeled with 15N ammonium sulphate and samples were analyzed using LTQ-FTICR-MS. Samples were taken at one generation of growth after either the shock was applied or iron starvation initiated. A control flask was maintained under the same conditions. The isotopomer peak intensity was calculated from an averaged MS spectrum covering the chromatographic elution time window of the peptide. Protein turnover values were determined for 151 proteins, with 31 having significant changes in their protein turnover under the different conditions studied. Protein turnover was found to be increased under acid shock for 28 of the proteins but decreased under iron starvation for all proteins. Although this approach allows comparisons of protein turnover to be made between different experimental conditions, it does not allow protein synthesis or degradation rates to be defined absolutely.

An alternative approach to determining protein turnover rates of individual proteins is to use deuterated water as the label.[35] In a recent study, De Riva et al. used such an approach, which they referred to as stable isotope labeling of nonessential amino acids with heavy water.[48] Body water can be enriched to approximately 5% without any adverse effects on the physiology of the body. Deuterium is equilibrated across the body, reaching isotopic equilibrium rapidly. Once this is attained, the rate-limiting step of 2H incorporation is protein synthesis. Using 2H2O, it is possible to monitor protein turnover rates for extremely stable proteins. The proteins of interest are subjected to tryptic hydrolysis and the resultant peptides analyzed by liquid chromatography–MS/MS. In this study, MHC proteins were isolated from murine cells and analyzed using an LTQ Orbitrap system. Data were deconvoluted using mass isotopomer distribution analysis and the change in isotopomer abundance from baseline values are plotted over time for each peptide. From these data, fractional synthesis rates for MHC proteins were obtained. Importantly, it was found that this technique was viable in primary cells, which opens up new avenues of research. Rachdaoui et al. used both 2H2O and H218O to measure albumin synthesis in mice in vivo during steady-state and non-steady-state conditions.[49] This dual-labeling approach allows integration of the transitions between the fed and fasted state or during an acute perturbation. The use of heavy water has obvious appeal. Labeled water is cheap in comparison to stable-isotope labeled amino acids and is easy to administer. This offers the advantage of being able to perform longer-term labeling protocols, allowing the turnover of long-lived proteins to be measured with greater accuracy. However, the deconvolution of these data is complex and requires the development of specific algorithms. In addition, different amino acids have different incorporation efficiencies of 2H, favoring peptides with high levels of alanine or glycine.

While stable isotope labeling approaches have now been employed in cellular systems and animals to study protein turnover, few studies have focused on plants to date. The use of amino acids is problematic in plants as they are autotrophic and actively synthesize amino acids de novo. Moreover, amino acids are not transported equally to all cells in all tissues, which creates the possibility of introducing bias into the analyses. In comparison, heavy water has the advantage that it is totally invasive – it rapidly enters cellular compartments and equilibrates with the water environment. However, it should be noted that there is a limit to how much 2H2O an organism can tolerate. Yang et al. used heavy water to investigate protein turnover in Arabidopsis thaliana.[50] Intact seedlings were used and the turnover rates for 15 amino acids and two proteins involved in the auxin signaling pathway were determined following treatment with 30% 2H2O. As heavy water can inhibit seedling/plant growth and development, these data were compared with microarray analysis. The microarray analyses indicated that the vast majority of genes were unaffected by the heavy water, with only 122 increased in expression and 99 decreased after 7 days of continuous heavy water treatment. However, the expression of a greater number of genes was altered after only 4 h of supplementation with heavy water, although this change was recovered after prolonged exposure, suggesting both short- and long-term effects. Many of the proteins whose expression changed were stress-response genes such as heat-shock proteins. A hierarchy of incorporation half-life was observed for the 15 amino acids monitored.

Non-mass Spectrometric Methods

Alternative strategies have also been developed to define global protein turnover. Belle and colleagues integrated data from large-scale measurements of mRNA levels, translation rates, protein abundances and protein half-life measurements to probe the link between transcriptional regulation and protein half-life in yeast.[51] The program was initiated with the generation of approximately 4200 tandem affinity purification-tagged strains for which a protein product could be detected by Western blotting. The abundance of each protein was monitored over time by Western blotting following inhibition of protein synthesis by cycloheximide. Tagging of the protein was found to have no effect on the intrinsic stability. The error in measurement using this approach was that it was possible to group proteins as either kinetically stable or unstable but nothing more refined. Various clustering approaches were then used to look for functional links. The data fell into two main groups: efficient production (stable proteins) and regulatory flexibility (rapidly degraded proteins). However, as proteins with similar protein production rates can have very different degradation rate constants, the authors suggest that transcriptional regulation may be used by cells to buffer the half-life differences among proteins encoded by a group of coregulated genes.

These data were subsequently analyzed by Tompa et al. looking for trends in protein stability.[52] The half-lifes of 3750 proteins from the dataset, which represents 65.2% of the known yeast proteome, were studied. Quantitative measures of physical and sequential features of proteins were plotted against logarithms of their half-lifes and possible correlations assessed by linear fitting. It was found that the degree of disorder correlates inversely with half-life and there is a positive correlation with helical content and a negative correlation with the propensity for coil formation. Therefore, in agreement with the proteomic study in human cells discussed previously, protein disorder was found to play the major, although not deterministic, role. The authors concluded that protein degradation is not determined by a single characteristic but is a multifactorial process that shows large protein-to-protein variation.

A final approach that has been used for high-throughput studies is global protein stability analysis.[53] Developed in mammalian cells, this method is based on a fluorescent assay. A retroviral reporter construct is built, in which the expression cassette contains a single promoter that, with an internal ribosome entry site, permits the translation of two fluorescent proteins from one mRNA transcript. The first fluorescent protein (red fluorescent protein) serves as an internal control whereas the second (enhanced green fluorescent protein) is expressed as a fusion with the protein of interest. The enhanced green fluorescent protein/DsRed ratio of cells represents the stability of the protein in the system and is quantified by FACS. The ratio is not affected by transcriptional regulation. This technique allows real-time protein stability detection at the level of individual living cells and can be multiplexed by combining it with microarray technology. Using this approach, the stability of almost 8000 human proteins was calculated. Half-lifes were determined by extrapolation of the data from the known half-lifes of three proteins. Correlations showed that longer proteins were more stable; however, as with other global studies, there was no evidence to support the PEST hypothesis.

Advances in Bioinformatic Analysis

One of the key hindrances to global studies of protein turnover by MS has been the limitation of software to process the large volumes of data generated. It is feasible to identify thousands of proteins in a single experiment, leading to the requirement for quantifying the changes in expression of those proteins. As such, there is a requirement for fast and effective bioinformatic resources to accompany this increase in data. Although commercial packages are available, many of the bioinformatic advances have been led by academic researchers, who have developed 'in-house' solutions. These include programs such as RELEX,[54,55] MSQuant,[56] XPRESS and MaxQUANT,[57,58] among others (Table 2). In addition to quantifying changes in a large cohort of samples, many of these software packages, which in general are freely available, link to downstream applications for gene ontology annotation and pathway analysis. As such, these tools are now evolving to fully integrate data analysis from mass spectral processing to understanding the underlying biological processes.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.