Ming
Tong
,
Johanna M.
Smeekens
,
Haopeng
Xiao
and
Ronghu
Wu
*
School of Chemistry and Biochemistry and the Petit Institute for Bioengineering and Bioscience, Georgia Institute of Technology, Atlanta, Georgia 30332, USA. E-mail: ronghu.wu@chemistry.gatech.edu; Fax: +1-404-894-7452; Tel: +1-404-385-1515
First published on 10th March 2020
Proteins are continuously synthesized during cell growth and proliferation. At the same time, excessive and misfolded proteins have to be degraded, otherwise they are a burden to cells. Protein degradation is essential to maintain proteostasis in cells, and dysfunction of protein degradation systems results in numerous diseases such as cancer and neurodegenerative diseases. Despite the importance of protein degradation, the degradation pathways of many proteins remain to be explored. Here, we comprehensively investigated the degradation of newly synthesized proteins in human cells by integrating metabolic labeling, click chemistry, and multiplexed proteomics, and systematic and quantitative analysis of newly synthesized proteins first revealed the degradation pathways of many proteins. Bioinformatic analysis demonstrates that proteins degraded through two major pathways have distinct properties and functions. Proteins degraded through the ubiquitin-proteasome pathway contain more disordered structures, whereas those through the autophagy-lysosome pathway have significantly higher hydrophobicity. Systematic and quantitative investigation of the dynamics of newly synthesized proteins provides unprecedented and valuable information about protein degradation, which leads to a better understanding of protein properties and cellular activities.
There have been many reports studying the degradation of individual proteins.7,8 However, most of the existing methods rely on antibodies and fluorescence probes to investigate individual proteins, which limit their capacities for large-scale analysis of protein degradation. In recent years, mass spectrometry (MS)-based proteomics has provided an opportunity to comprehensively characterize proteins,9–21 including protein dynamics.22–25 For instance, Doherty et al. studied the stability of nearly 600 proteins in human A549 cells by using a dynamic stable isotope labeling by amino acids in cell culture (SILAC) approach.22 Mathieson et al. determined the half-lives of several thousand proteins in human cells using a SILAC-based method and found the architecture-dependent turnover of complex subunits.26 Martin-Perez et al. also employed a SILAC-based method to measure the turnover rates of over 3000 yeast proteins and found that actively used proteins had a faster turnover rate.27 However, some potential issues could exist in the SILAC-based methods. First, amino acid recycling exists, which impacts the quantification of protein dynamics, including protein degradation. Second, proteins are continuously synthesized during cell growth, and analysis of protein degradation could be complex. Furthermore, low-abundance proteins or proteins with high degradation rates may not be analyzed.
Both protein synthesis and degradation are extremely important to maintain proteostasis, and they need to be well-balanced in cells.1,25 Newly synthesized proteins are those synthesized within a relatively short period of time. Compared with existing proteins, newly synthesized ones may be more dynamic because a portion of them being misfolded and an excessive amount of newly synthesized proteins need to be promptly degraded.28,29 The regulation of newly synthesized proteins is especially important to maintain protein homeostasis in cells, including the degradation of unfolded or improperly folded proteins. In cells, there are many high-abundance proteins that mask signals from low-abundance proteins during MS analysis, and thus effective separation and enrichment of those newly synthesized proteins, many of which have low abundance, are critical for comprehensive analysis of their degradation by MS. In addition, they need to be distinguished from existing proteins. With the development of metabolic labeling30,31 and bioorthogonal chemistry,32,33 it is possible to selectively tag and separate newly synthesized proteins for further analysis.
Here we integrated metabolic labeling, click chemistry and multiplexed proteomics to comprehensively investigate the dynamics of newly synthesized proteins and their degradation pathways. This method allowed us to selectively enrich newly synthesized proteins and distinguish them from existing ones. Parallel experiments were performed to study protein degradation in MCF-7 cells and measure their half-lives with or without the inhibition of each major degradation pathway: bortezomib (Btz) for the proteasome pathway and 3-methyladenine (3-MA) for the lysosome pathway. Around 3000 newly synthesized proteins were quantified in each experiment, and a total of 4042 proteins were quantified and their half-lives spanned a wide range from a few minutes to over 200 hours. Systematic and quantitative analysis of the dynamics of newly synthesized proteins first revealed that 868 proteins were degraded through the ubiquitin-proteasome pathway while 228 were through the autophagy-lysosome pathway. Bioinformatic analysis demonstrates that proteins degraded through two major pathways have distinct properties and functions. This research provides unprecedented and valuable information regarding protein dynamics, protein degradation pathways and protein properties correlated with their degradation.
The experimental procedure is shown in Fig. 1, and a detailed description is provided in ESI.† Briefly, around 3 × 107 MCF-7 cells were equally passaged to six flasks. After 24 hours, cells were cultured in the medium containing heavy lysine (Lys8, 13C6 and 15N2) and AHA for 4 hours. After metabolic labeling, we switched the media to normal Dulbecco's Modified Eagle Medium (DMEM) with Lys0 and methionine. Simultaneously, either Btz, 3-MA or dimethyl sulfoxide (DMSO, as a control) was added to the culture medium, respectively, for each experiment. The numbers of cells were kept as similar as possible before switching the media. Then cells were harvested at 0, 2, 4, 6, 8, and 10 h, respectively, after the media switch. Newly synthesized proteins labeled with AHA were selectively enriched through the copper-free click reaction with dibenzocyclooctyne (DBCO)-derivatized beads; heavy lysine labeling allowed us to unambiguously distinguish them from existing proteins. The beads were stringently washed to remove non-specific binding proteins. Enriched proteins were reduced with 5 mM dithiothreitol (DTT) and alkylated with 15 mM iodoacetamide. After on-bead digestion with Lys-C, peptides from the samples at different time points were each chemically labeled with the sixplex tandem mass tag (TMT) reagents. The reaction was quenched by adding hydroxylamine. Samples were mixed, purified and then separated into 20 fractions by high-pH reversed-phase HPLC. Each fraction was further purified, and then analyzed by LC-MS/MS.
Fig. 1 Experimental procedure for quantifying the degradation of newly synthesized proteins and measuring their half-lives in parallel experiments. |
An example of peptide identification and quantification is displayed in Fig. 2(a). The peptide SEASSEFAK* (* refers to heavy lysine) was confidently identified with an Xcorr of 4.1. This peptide is from PRP16, a pre-mRNA-splicing factor ATP-dependent RNA helicase, which may be involved in pre-mRNA splicing and mRNA exporting from the nucleus. The intensities of the reporter ions enabled us to accurately quantify the relative abundance change as a function of time.
Protein half-lives were calculated by fitting six ratios (the ratio is set to 1 for the first time point) to an exponential decay equation, as reported previously (Fig. 2(b)).25 The median intensity of the reporter ions were calculated from four unique peptides for PRP16, and the ratio at each of the six time points was 1, 0.84, 0.69, 0.55, 0.42, or 0.35, respectively. Those ratios were then fitted to the following equation, F(t) = e−0.101t, and the half-life was calculated to be 6.8 (ln(2)/0.101) hours. The half-life of FIS1 was determined by the same process. Overall, eleven total and six unique peptides from this protein were quantified, and its half-life was 14.9 hours. In addition, as an example, the half-life of each unique peptide was also calculated and compared with the parent proteins (Fig. 2(c)). The majority of peptides are within ±10% of the half-life of the parent protein, which shows the relatively high reproducibility of this method.
Gene Ontology analysis with Database for Annotation, Visualization and Integrated Discovery (DAVID)42 was performed to determine the cellular localization of those proteins with short (<3 hours) or long (>20 hours) half-lives. As shown in Fig. S2,† among short-lived proteins, those related to the nuclear lumen, proteasome and ribosomal subunits were highly enriched. For instance, the half-lives of RPL15 (60S ribosomal protein L15), PSMD6 (26S proteasome non-ATPase regulatory subunit 6) and JAK1 (tyrosine-protein kinase JAK1) were 1.2, 1.8 and 1.8 hours, respectively. It has been reported that ribosomal proteins, especially those located in the nucleolus, have fast turnover rates.23 JAK1 was reported to be a highly unstable protein, and the current result corresponds well with the half-life of 1.5 h in previous research.43 On the contrary, among long-lived proteins, those located in the extracellular exosome, endomembrane system and mitochondrion were over-represented (Fig. S2(b)†). Moreover, the half-lives of histone proteins are relatively long: 23.9 h for HIST2H2BC, 24.1 h for HIST1H1B and 20.0 h for HIST1H1D.
Among quantified proteins, those in the extracellular exosome had the longest median half-life. Proteins in the mitochondrion, the cytosol and the nucleus had a median half-life close to the overall median of 9.8 h (Table S2† and Fig. 3(a)). The half-lives of proteins in the proteasome and ribosome complexes were relatively shorter and their median half-lives were 6.8 and 6.3 h, respectively. This trend is in good agreement with previous studies in both mammalian and yeast cells.26,27
Protein half-lives in different protein complexes and their subunits were also assessed and compared (Table S3† and Fig. 3(b)–(d)). Here we looked into three protein complexes, i.e. the proteasome, the ribosome, and the nuclear pore complex (NPC). Differences between different subunits of the complexes were observed. For instance, the half-lives of proteins in the 20S core subunit were dramatically shorter (p < 0.001) than those of proteins in the 19S regulatory subunit of the proteasome (Fig. 3(b)). Protein half-lives in the peripheral structure of the NPC were longer (p < 0.05) than those in the scaffold part (Fig. 3(c)), which differs from a previous study.26 We reason that proteins in the complexes are typically synthesized in excessive amounts and degraded continuously.28,44 According to a previous study about the dynamic organization of NPC, the residence periods of the peripheral proteins Nup153 and Nup50 were 20 s and 1 min, respectively, while the residence periods of the scaffold proteins Nup133 and Nup205 were more than 35 h.45 Therefore, newly synthesized peripheral proteins Nup153 and Nup50 may be more rapidly assembled to NPC, contributing to their longer half-lives (16.4 and 19.5 h).
For the ribosome complexes, the median half-life of proteins in the mitochondrial ribosome was slightly longer compared to proteins in the cytosolic ribosome, and the median half-life of proteins from the small subunits was longer than that from the large subunits (Fig. 3(d)).
After cells were treated with Btz, a total of 3023 proteins were quantified and their half-lives were calculated, which are listed in Table S1† (Fig. 4(a)). It is expected that many proteins have longer half-lives with the Btz treatment, especially for those degraded through the proteasome. The median half-life for quantified proteins was 21.7 h, which is more than twice as long as untreated cells (9.8 h); the half-life distributions of overlapped proteins (1975) in three parallel experiments are displayed in Fig. 4(b). As shown in Fig. 4(a), 2249 proteins were quantified in cells with or without the Btz treatment. The half-lives of >60% proteins increased more than two-fold while the half-lives of very few proteins (only 9) decreased over 2-fold (Fig. 4(c)).
For instance, ARPC5 functions as a component of the Arp2/3 complex which is involved in the regulation of actin polymerization, and it was reported to be degraded through the ubiquitin-proteasome pathway.7 ARPC5 had a half-life of 3.7 h in untreated cells while its half-life increased >3-fold (11.9 h) in proteasome-inhibited cells. CDK5RAP3 is a tumor suppressor which controls cell proliferation and inhibits the NF-kappa-B-mediated gene transcription.8 In MCF-7 cells, CDK5RAP3 had a half-life of 10.8 h, which increased to 54.7 h after the Btz treatment.
In order to further understand protein degradation, we also quantified protein dynamics after the inhibition of the other major protein degradation pathway, i.e. the autophagy-lysosome pathway. 3-Methyladenine was often reported to effectively inhibit the autophagy-lysosome pathway.48,49 In this experiment, 2895 proteins were quantified and their half-lives were obtained for 3-MA-treated cells (Fig. 4(a)). The median half-life of these quantified proteins is 13.4 h, which is 3.6 h longer than that from untreated cells.
Among 2305 proteins quantified in cells with or without the treatment of 3-MA, the half-lives of 630 proteins increased more than 2-fold (Fig. 4(d) and S3†). For example, RAB5A, a small GTPase Rab protein, is a key regulator of intracellular membrane trafficking and autophagosome assembly. VAMP7 is required in the transport of proteins from early endosome to the lysosome. Both proteins are known to be degraded through the lysosome after the fusion between the autophagosome and the lysosome. Their half-lives increased 6.3- and 7.5-fold after the 3-MA treatment, indicating the effective inhibition of the autophagy-lysosome pathway.
Taken together, a total of 4042 proteins were quantified across three parallel experiments (Table S4†); as shown in Fig. 4(a), 1975 proteins (∼65% in each experiment) were quantified in all three experiments. The half-life distributions of overlapped proteins in untreated, Btz-treated and 3-MA-treated cells are displayed in Fig. 4(b). With the treatment of Btz, the distribution shifted dramatically towards longer half-lives.
The up-regulated proteins were further analyzed using DAVID and Protein Analysis Through Evolutionary Relationships (PANTHER);50 selective results are included in Fig. 5(a). For Btz-treated cells, proteins located in the proteasome regulatory particle, ribosomal subunit, cytoskeleton and nucleolus were highly enriched. Proteins related to translational initiation and ubiquitination were also overrepresented. For 3-MA-treated cells, the clustering results were dramatically different from Btz-treated cells. Proteins located in the plasma membrane, endocytic vesicle and mitochondrial matrix were highly enriched. Proteins related to transport, unfolded protein binding and hydrolase activity were enriched as well.
To determine which pathway is responsible for the degradation of each protein, we divided all proteins into five groups (Fig. 5(b)). Group I contains 313 proteins, whose half-lives increased at least two-fold due to treatment with Btz or 3-MA. The inhibition of either the ubiquitin-proteasome or autophagy-lysosome pathway dramatically affected (log2 ratio > 1) the degradation of those proteins. For instance, RPL19 is a component of the ribosomal large subunit, and its half-life increased 3.3- and 2.1-fold after Btz or 3-MA treatment, respectively. Group II has 868 proteins, and their half-lives increased more than two-fold after Btz treatment, but not 3-MA treatment, indicating that those proteins were degraded only through the ubiquitin-proteasome pathway. On the contrary, Group III contains 228 proteins which were degraded only through the autophagy-lysosome pathway. The median half-life in untreated cells were 8.9 h for Group II and 12.7 h for Group III, and the difference in half-life distributions between Group II and III is significant with a P value < 0.0001 (Fig. 5(c)). This is consistent with the common belief that short-lived proteins are typically degraded through the ubiquitin-proteasome pathway while proteins degraded through the autophagy-lysosome pathway are relatively longer-lived.51,52
Five-hundred and sixty-three proteins in Group IV were defined as uncertain because their half-lives stayed relatively consistent (−1 < log2 ratio < 1) with the treatment of Btz or 3-MA, compared with those in untreated cells. They might be relatively more stable during the inhibition of either major pathway, and/or these proteins may be degraded through other pathways. Surprisingly, three proteins, i.e. histone H4, lanosterol 14-alpha demethylase and POTE ankyrin domain family member E, were found to have dramatically shorter (log2 ratio < −1) half-lives after the Btz or 3-MA treatment. More studies need to be performed to understand this interesting phenomenon.
To better understand protein degradation, we performed further analysis on the physiochemical properties between proteins degraded only through the ubiquitin-proteasome or autophagy-lysosome pathway. Selected results are displayed in Fig. 5(d) and S4.† For instance, hydrophobic amino acid residues more frequently existed in proteins degraded through the lysosome pathway, including the percentages of valine, isoleucine and phenylalanine being significantly higher. However, basic amino acid residues were more frequently found in proteins degraded through the proteasome pathway, especially lysine (Fig. S4†). This can be explained by the fact that ubiquitination typically occurs on the lysine residue before protein degradation by the proteasome. As shown in Fig. 5(d), protein length and molecular weight did not have notable differences between proteins degraded through either pathway. However, proteins degraded through the proteasome pathway contained more disordered structures, i.e. the lower numbers of beta sheets and helix. Consistent with the frequencies of amino acid residues, the hydrophobicities (GRAVY score) of proteins were significantly higher in proteins degraded through the lysosome pathway. Interestingly, higher isoelectric points were found in proteins degraded through the proteasome pathway, which matches very well with the higher frequencies of basic amino acid residues.
In untreated MCF-7 cells, proteins in the 20S core subunit had much shorter half-lives compared with those in the 19S regulatory subunit. As shown in Fig. 6(a) and (b), different degradation pathways were found for proteins in the 20S and 19S subunits (Table S5†). Most proteins in the 19S regulatory subunit were degraded through the proteasome, except Rpt1, Rpn6, Rpn12 and Rpn13, whose ratios are higher than 1.0 (1.5, 1.1, 1.8 and 1.3, respectively) with the inhibition of the proteasome. This suggests a high likelihood that they were also degraded through the proteasome pathway. The half-life of Rpn8 was 6.9 h in untreated cells and increased to 12.5 and 15.3 h under the Btz or 3-MA treatment, respectively. This protein was likely degraded through both pathways. On the contrary, only two proteins in the 20S core subunit were degraded by the proteasome. The responses of 19S regulatory subunit proteins to the Btz or 3-MA treatment were dramatically different (P-value < 0.001), while no significant difference was found for 20S core subunit proteins (Fig. 6(c)).
The cytosolic and mitochondrial ribosomes are responsible for protein synthesis in cells. Overall, we quantified 82 ribosomal proteins from the 60S and 40S ribosome subunits and the 28S and 39S mitochondrial ribosome subunits. Thirty-seven proteins were degraded through the proteasome when using >2-fold increase of the half-life as the threshold (Fig. 6(d) and S5†).
However, the half-lives of nearly all ribosomal proteins increased with the inhibition of the proteasome. For example, the half-lives of MRPL37 and RPL27 increased 2 fold, and proteins RPS25, MRPS2, RPL10, RPS9, RPL29, MRPL41, RPS18 and RPL13A had a 1.9 fold increase in their half-lives with the inhibition of the proteasome, but their half-lives remained similar after the 3-MA treatment (Table S5(b)†). These proteins very likely were also degraded through the proteasome pathway. The current results indicate that ribosomal proteins were mainly degraded through the proteasome.
In this work, the protein half-lives were much shorter than those from the SILAC-based proteomic experiments.23,25,56 One contributing factor is that newly synthesized proteins are more dynamic than existing proteins. Previous studies illustrated that up to 30% of newly synthesized proteins were defective ribosomal products, which were rapidly degraded after their synthesis.29 Furthermore, using SILAC-based methods, amino acid recycling is an inherent issue, which results in longer protein half-lives. By using the current method, AHA and heavy lysine may still be recycled, but the possibility of recycling both AHA and heavy lysine in a single newly synthesized peptide and protein is very low.
Using the TMT method, the quantification of peptides could result in ratio suppression due to potential interferences.57 In order to minimize potential interferences, we took several preventative measures. First, each sample was fractionated into twenty fractions using HPLC, which made each fraction much less complex. Second, a long gradient during LC-MS analysis allowed peptides to be further separated. Third, a narrower isolation window was used, i.e. 1.2 D instead of the commonly used 2.0 D. As a result, potential interference may not be an issue in this work, especially considering the enriched samples (newly synthesized proteins) are less complex compared to the whole proteome. If the ratio suppression is a serious problem, the half-lives of quantified proteins would be much longer. However, the protein half-lives are obviously shorter than those obtained using SILAC-based methods,25,58 and they correspond well with those measured from an MS-independent method.41
Here, we studied protein degradation through investigating the dynamics of newly synthesized proteins, and the half-lives of newly synthesized proteins were calculated from six time points. Proteins at the six time points were quantified simultaneously using the multiplexed proteomics method, which further increased the quantification accuracy. Compared with SILAC methods typically based on two or three time points, the current results are more reliable. In this work, MCF-7 cells were used to study protein degradation, and cell-specific degradation of some proteins may not be excluded, which needs to be further studied.
After the inhibition of the ubiquitin-proteasome or autophagy-lysosome pathway through the treatment of Btz or 3-MA, respectively, the median half-life of newly synthesized proteins increased to 21.7 or 13.4 hours from 9.8 hours in cells without any treatment. Comprehensive analysis of the degradation of newly synthesized proteins with and without the inhibition of either major protein degradation pathway revealed that 868 proteins were degraded through the proteasome while 228 proteins were through the autophagy-lysosome pathway. In addition, 313 proteins were degraded through both pathways. Interestingly, some proteins were found to have shorter half-lives after the inhibition of either degradation pathway, which may be due to the cross-talk between the ubiquitin-proteasome and autophagy-lysosome pathways. Although the mechanisms of the two pathways are dramatically different, several studies reported that there are some correlations between these two major degradation pathways.59,60 The inhibition of the proteasome may result in the induction of the unfolded protein response and lead to the activation of the transcriptional factors, ATF4 and IRE1, increasing the activity of the autophagy-lysosome pathway.61 In the current work, we identified 41 proteins with longer half-lives after the inhibition of the autophagy-lysosome pathway, but their half-lives decreased with the Btz treatment, including NRAS and 7 Ras-related proteins, which could be due to cross-talk between the two pathways. These proteins may serve as examples to further study the mechanisms of the cross-talk between the ubiquitin-proteasome and autophagy-lysosome pathways.
The target-decoy method was employed to evaluate and control the FDRs of peptide and protein identifications.64 Linear discriminant analysis (LDA) was exploited to distinguish correct and incorrect peptide identifications using numerous parameters such as XCorr, ΔCn, and precursor mass error.65 After scoring, peptides fewer than seven amino acids in length were discarded and peptide spectral match was filtered to a less than 1% FDR based on the number of decoy sequences in the final data set.
The TMT reporter ion intensities in the MS/MS spectra were used to quantify peptides. The isotopic information provided by Thermo was utilized to calibrate the ion intensities. Only peptides with the reporter ions found at all six channels were further analyzed for quantification. The median intensity at each time point was calculated from all unique peptides for each protein. The degradation rate and the standard deviation were obtained by fitting six ratios (the ratio at the first time point is set as 1) to the exponential decay equation based on the orthogonal distance regression algorithm. The protein half-lives were then computed based on the degradation rates.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9sc06479f |
This journal is © The Royal Society of Chemistry 2020 |