Summary
Protein turnover is a critical regulatory mechanism for proteostasis. However, proteome-wide turnover quantification is technically challenging and, even in the well-studied E. coli model, reliable measurements remain scarce.
Here, we quantify the degradation of ~3.2k E. coli proteins under 12 conditions by combining heavy isotope labeling with complement reporter ion quantification and find that cytoplasmic proteins are recycled when nitrogen is limited. Furthermore, we show that protein degradation rates are generally independent of cell division rates, and we used knockout experiments to assign substrates to the known ATP-dependent proteases. Surprisingly, we find that none are responsible for the observed cytoplasmic protein degradation in nitrogen limitation, suggesting that a major proteolysis pathway in E. coli remains to be discovered. Thus, we introduce broadly applicable technology for protein turnover measurements. We provide a rich resource for protein half-lives and protease substrates in E. coli, complementary to genomics data, that will allow researchers to decipher the control of proteostasis.
Introduction
Protein degradation is central to protein homeostasis (proteostasis) and is critical in most cellular pathways (Cohen-Kaplan et al., 2016; Mahmoud and Chien, 2018). As environments change, modification of degradation rates can rapidly adapt protein abundances to desired levels. Even if protein levels are modulated via transcription or translation, a protein’s response time is set by its replacement rate (Alon, 2006; Rosenfeld and Alon, 2003). Unsurprisingly, many signaling and transcriptional regulatory proteins exhibit short half-lives (Auld and Silver, 2006; Liu et al., 2012; Pahl and Baeuerle, 1996; Wettstadt and Llamas, 2020). Protein degradation is important in health and disease, for example in cancer and neurodegenerative disorders (Jang, 2018; Labbadia and Morimoto, 2015). Additionally, protein degradation plays an important metabolic role. It has been shown that bacteria and yeast cells increase their proteome turnover rates under starvation conditions, presumably generating and recycling scarce amino acids (Borek et al., 1958; Halvorson, 1958; Kuroda et al., 2001; Mandelstam, 1957).
Quantitative models have been developed to describe the dependence of global protein expression on cells’ physiological characteristics, most notably cell doubling times. These models are probably the most well-developed in the model bacterium E. coli (Erickson et al., 2017; Klumpp et al., 2009). The cell cycle time in E. coli varies from 20 minutes in rich media to the cessation of division under starvation. Transcription rates typically increase with cell division rates, while translation rates remain constant (Liang et al., 1999; Liang et al., 2000). Knowing how global parameters scale with physiological cell states allows for remarkable quantitative predictions for gene expression changes across different growth conditions (Erickson et al., 2017; Klumpp et al., 2009). However, active protein degradation is typically ignored in these models. Instead, proteins are assumed to be completely stable and only diluted via cell growth and division. This simplification is likely due to a lack of reliable genome-wide degradation rate measurements under varying growth conditions. It is still unclear how active degradation rates scale with changing cell cycle times and how this affects global gene expression regulation. Presumably due to the lack of reliable genome-wide turnover resources, most systems biology studies have ignored the contribution of protein degradation rates in modeling protein homeostasis or dynamics (Balakrishnan et al., 2021; Belliveau et al., 2021; Scott et al., 2010). Knowledge of protein degradation rates and how they scale with the physiological characteristics of cells would improve predictive models of protein expression across various cell states.
Even for the proteins we know to be degraded, it is often unclear how these proteins are marked for destruction and which protease is responsible. Cells have developed sophisticated mechanisms to recognize and degrade specific proteins. In bacteria, selective proteolysis is executed by ATP-dependent proteases (Mahmoud and Chien, 2018). While many proteases can digest unfolded proteins and peptides, unfolding a protein for degradation requires energy. In E. coli, four ATP-dependent proteases are known: ClpP, Lon, HslV, and FtsH. Pulldown experiments with proteolytic-dead mutants or protein-array studies have allowed the proteome-wide identification of putative substrates (Arends et al., 2018; Flynn et al., 2003; Tsai et al., 2017; Westphal et al., 2012). Orthogonally, individual substrates have been assigned to the four proteases by measuring the degradation of individual proteins in protease knockout strains or via in vitro assays (Camberg et al., 2009; Schweder et al., 1996). Several proteins (e.g., RpoH, LpxC, and SoxS) have been shown to be degraded by multiple proteases, demonstrating remarkable redundancy (Biernacka et al., 2020; Griffith et al., 2004; Kanemori et al., 1997). On a proteome-wide scale, it is unclear which proteases degrade which substrates and to what extent those substrates might overlap.
In most biological systems, protein degradation is balanced by the synthesis of new protein, making measurements of degradation rates challenging. An easy way to overcome this complication is by using translational inhibitors like cycloheximide or chloramphenicol (Belle et al., 2006; Biran et al., 2000; Li et al., 2021). Assuming that the addition of the drug does not perturb the cells aside from blocking the translation of new proteins, protein degradation can be conveniently measured by assaying changes in protein abundances over time via western blots or quantitative proteomics. However, when we performed such experiments in E. coli, we found that many proteins that seemed to be rapidly degraded were periplasmic (Fig. S1). Further investigation revealed that these periplasmic proteins were not degraded, but rather the addition of chloramphenicol led to the accumulation of periplasmic proteins in the bacterial growth medium (Fig. S1). Presumably, this was due to protein leakage through the outer membrane. We concluded that translation inhibitor experiments in E. coli could lead to major perturbations and interpreting such studies might be challenging.
A classic method to measure the unperturbed turnover of biological molecules is using radioactive isotope tracking or the combination of heavy isotope labeling and quantitative mass spectrometry (Arias et al., 1969; Foster et al., 1939). Isotopic labels can be introduced with heavy nutrients (e.g., ammonia, glucose, or amino acids) or by incubation in heavy water. Heavy ammonia, glucose, and water are comparatively cheap but result in overly complex MS1 spectra, which are difficult to interpret, particularly for lower abundance proteins (Cargile et al., 2004; Helbig et al., 2011; O’Brien et al., 2020). Most proteomic turnover studies have been performed with heavy amino acid labeling (dynamic SILAC) (Boisvert et al., 2012; Schwanhausser et al., 2011), but the small number of labeled residues limits sensitivity for short-time SILAC labeling, and missing values can hinder the coverage of multiple time points in complex systems. A further advance has been the combination of SILAC experiments and isobaric tag labeling (Savitski et al., 2018). However, these measurements tend to suffer from the inherent ratio compression of multiplexed proteomics (Pappireddi et al., 2019; Ting et al., 2011; Wenger et al., 2011). For a more detailed discussion of the advantages and limitations of various global protein turnover measurement techniques, please see the recent review by Ross et al., particularly Table S1 (Ross et al., 2021). Despite the central role of protein degradation in nearly every aspect of biological regulation, reliable and large-scale measurements are still scarce. Even fewer studies have compared degradation rates between multiple conditions (Christiano et al., 2020; Rao et al., 2008).
Here, we measure protein turnover in E. coli by combining heavy isotope labeling via 15N ammonia with the accurate multiplexed proteomics method TMTproC (Johnson et al., 2021). We provide a rich resource of protein degradation rates for ~3.2k E. coli proteins (77% of all proteins in E. coli) measured across 12 different growth conditions with replicates. We find that active degradation rates typically are constant irrespective of changing cell division rates. When comparing degradation rates among various nutrient limitations, we find that E. coli recycles its cytoplasmic proteins when nitrogen-limited, and we assign substrates to proteases by measuring the change of protein turnover in knockout strains.
Results
Combining heavy isotope labeling with complement reporter ion quantification enables high-quality measurements of protein turnover
We wanted to measure protein degradation rates and evaluate how these rates vary across growth conditions. To simplify our measurements, we grow E. coli in chemostats, where we can control the cell doubling time and enforce steady state (Fig. 1A). After cells reach steady state, we change the medium from unlabeled nutrients to 15N-labeled ammonia. Over time, the 15N-ammonia concentration in the medium increases, and newly synthesized proteins incorporate more heavy isotopes. We used 15N-ammonia rather than labeled amino acids because of the higher signal we obtain with a low labeling fraction. For example, when 10% of the nutrient pool is labeled: If using heavy arginine, 90% of newly synthesized tryptic peptides ending in arginine are unlabeled. In contrast, when using heavy ammonia, a peptide having 15 nitrogens, only (90%)15=21% will be unlabeled. Over time the isotopic envelope of peptides will shift towards heavier forms (Fig. 1B). With the knowledge of a peptide’s chemical composition and the fraction of heavy isotopes over time, we can calculate the degradation rate of the corresponding protein. However, in practice obtaining such measurements of isotopic envelopes in the MS1 spectrum is quite challenging, particularly at later time points when all the isotopic envelopes spread out and overlap with others. Additionally, missing values between measurements of various time points are a severe limitation of such approaches (O’Brien et al., 2020). To overcome these limitations, we labeled each of the acquired eight time-points with TMTpro isobaric tags and combined them for co-injection into the mass spectrometer (Johnson et al., 2021; Thompson et al., 2019). Analyzing these extremely complex samples with standard low m/z reporter ion quantification would lead to severe ratio distortion and measurement artifacts (Ow et al., 2009; Pappireddi et al., 2019; Ting et al., 2011). We overcame this limitation by programming the instrument to isolate the pseudo-monoisotopic peak (M0) and quantify the complement reporter ions in the MS2 (Fig. 1C). Figure 1D shows quantification for peptides from the expected stable protein OmpF and the expected degrading protein RpoS from carbon-limited chemostats with 6h doubling times (Schweder et al., 1996). Proteins without active degradation are expected to follow the theoretical dilution curve (dotted curve) based on the chemostat dilution rate. Fitting the measured signal of OmpF peptides with a model for the expected decay of the M0 peak (solid curve) results in a turnover half-life similar to this expected value. In contrast, the fit and deduced half-life for RpoS is much shorter than the cell doubling time. We obtain half-lives for ~2.6k E. coli proteins with a median standard deviation of 0.3 h (Fig. 1E, Table S1). Having established this technology, we acquired similar measurements for 12 different growth conditions, each with two biological replicates (Table 1, Table S1). We then used this resource to investigate how E. coli adapts protein turnover under various growth conditions.
Active protein degradation rates typically do not scale with division rates
With a method to measure protein degradation among various conditions, we determined how protein degradation scales with cell cycle time. The total turnover rate (ktotal) is a combination of active degradation (kactive) and dilution (kdilution) due to cell division (Fig. 2A). We consider two simple and reasonable models of the relationship between these two parameters. In the first model, kactive scales with kdilution, i.e., the protein half-life remains a constant fraction of the cell cycle time. In the second model, active degradation rates are independent of growth rate, i.e., the active degradation rate of each protein remains constant regardless of cell doubling time. The two models have distinct predictions on how the total protein replacement half-time (T1/2, total) should scale with changing cell cycle times. In the scaled model, T1/2, total for each protein linearly increases with cell cycle time (Fig. 2B). In contrast, in the constant model, the dilution rate dominates for rapidly dividing cells while the contribution from active degradation becomes more relevant for slower dividing cells. To test the models’ predictions, we grew E. coli cells with a range of doubling times, including rapidly doubling cells with unlimited growth in minimal medium (0.7h) and slower doubling cells in carbon-limited chemostats (3h, 6h, and 12h). We asked how T1/2, total changes among these conditions (Fig. 2C). The likelihood-ratio test clearly favors the constant model regardless of which cell-cycle times we compare. This indicates that active protein degradation rates typically remain constant regardless of cell division rates. However, while we are confident that this model describes most of the proteome well, we noted interesting exceptions, particularly when comparing slower-growing cells in the chemostat to cells growing without nutrient limitation. For example, RpoS degrades faster in unlimited growth condition than the non-scaled model would predict based on chemostat measurements. This is consistent with the previous finding that RpoS is rapidly degraded in exponentially growing cells but becomes stabilized when nutrient-limited (Zhou and Gottesman, 1998). Because kactive rates are generally constant across cell division rates, we can more accurately measure kactive when kdilution is small. For the remainder of this study, we will focus on measurements with 6h doubling times. These conditions are experimentally easy to access, provide good resolution for active degradation, and can be extrapolated to other conditions using our insight into the constant scaling of active degradation rates.
Escherichia coli recycles its cytoplasmic proteins under nitrogen-limitation
Next, we wanted to compare protein degradation rates under various nutrient limitations. We compared carbon (C-lim), phosphorus (P-lim), and nitrogen (N-lim) limitation measurements from chemostats with 6h doubling times. A histogram of C-lim protein half-lives indicates that most proteins are stable with a measured total half-life close to the theoretical dilution time (Fig. 3A). Using biological replicates to identify confidently degrading proteins (Fig. 3B), we found that 15% of the proteome is actively degraded in C-lim (p-values < 0.05). The histogram of protein half-lives under P-lim shows a similar distribution and percentage of proteins that confidently degrade as in C-lim. However, the N-lim measurements indicate that many proteins shift from stable to slowly degrading (Fig. 3A). Consistent with this, the number of confidently degrading proteins increases to 43% (Fig. 3B, p-value < 0.05). We found that the proteins’ subcellular localization is a major predictor of the differential half-lives between N-lim vs C-lim or P-lim (Fig. 3C, D). Histograms and scatter plots of half-lives for the proteins partitioned by subcellular localization indicate that the mode for membrane and periplasmic proteins in all three conditions is very close to the theoretical dilution limit. In contrast, the mode of half-lives for cytoplasmic proteins is similar to the dilution half-life in C-lim and P-lim but is significantly shorter under N-lim, indicating that most of the cytoplasm is actively degrading under N-lim. It is difficult to confidently identify active degradation for the proteins with half-lives close to the dilution half-life. Therefore, our estimates of 43% (containing 56% of cytoplasmic, 13% of membrane, and 4% of periplasmic proteins) of the proteome actively degrading in N-lim is a very conservative estimate.
We then tested whether our finding of cytoplasmic recycling under N-lim chemostats extends to the more physiologically relevant case of batch starvation. We grew E. coli cells in minimal medium and switched the exponentially growing cells into medium depleted of nitrogen (Fig. 3D). Once again under nitrogen starvation, cytoplasmic proteins are degraded, and membrane/periplasmic proteins are stable. Thus, we find that E. coli cells slowly degrade their cytoplasmic proteins when nitrogen is scarce in both chemostats and batch culture. About 2/3 of the cell’s nitrogen is stored in proteins (Milo and Phillips, 2016). The slow degradation of proteins upon nitrogen starvation likely allows the regeneration and recycling of scarce amino acids and enables E. coli to produce new proteins to adapt to new environments.
Measuring protein turnover in knockout mutants allows the identification of protease substrates
Next, we wanted to investigate protease-substrate relationships on a proteome-wide level. We were particularly interested in discovering the protease(s) responsible for the large-scale turnover of cytoplasmic proteins in N-lim. Since unfolding and degrading stably folded cytoplasmic proteins requires energy, we focused on assigning substrates to the ATP-dependent proteases. In E. coli, there are four known ATP-dependent protease complexes: ClpP (in complex with ClpX or ClpA), Lon, HslV (in complex with HslU), and FtsH (Mahmoud and Chien, 2018). We can identify putative substrates for these proteases by comparing the protein half-lives in protease knockout (KO) with wild-type (WT) cells (Fig. 4A). For example, the unfoldase ClpA is completely stabilized on knocking out clpP. ClpA’s C-terminus has previously been shown to be a signal for degradation by ClpP (Maglica et al., 2008). Similarly, Tag and UhpA are identified as the substrates of Lon and HslV because their half-lives confidently increase upon knocking out the respective proteases. However, many proteins still degrade in the three protease KO lines, e.g., the phosphatase YbhA required in Vitamin B6 homeostasis still rapidly turns over with a half-life of ~1 h in each knockout strain (Sugimoto et al., 2018). Surprisingly, even the proteins that increase their half-lives in single KOs often do not get completely stabilized. Additionally, bulk cytoplasmic proteins are still degraded in all three single KOs.
Deleting ftsH is more complicated than the other protease genes. One of its substrates, LpxC, catalyzes the committed step in the lipid A synthesis pathway. Lipid A is the hydrophobic anchor of lipopolysaccharides (LPS), a key outer membrane component. Deletion of ftsH leads to increased levels of LpxC, causing an accumulation of LPS, making the cells nonviable (Ogura et al., 1999). ftsH null mutant cells can be rescued with a mutation of FabZ (L85P), which slows LPS synthesis and compensates for the increased LpxC levels (Ogura et al., 1999). These ΔftsH fabZ (L85P) cells are viable, though unfortunately they grow too slowly on minimal media and are washed out of the chemostat. We therefore could not measure protein turnover in a ftsH mutant similar to the other proteases. Rather, we repeated the batch nitrogen starvation experiments (Fig. 3E). Similar to the WT cells, cells lacking ftsH degrade cytoplasmic proteins. In contrast, membrane proteins are mostly stable (Fig. 4B). This indicates that none of the four known ATP-dependent proteases in E. coli are individually responsible for the large-scale cytoplasmic recycling under nitrogen scarcity.
We then asked if proteases might act redundantly, i.e., multiple proteases share a substrate, which could mask the effects of deleting individual proteases. To this end, we measured protein turnover in a triple KO line (ΔhslV Δlon ΔclpP). We observe that many more proteins get stabilized in the triple KO than in any individual KO line (Fig. 4C). Strikingly, YbhA is almost completely stabilized in the triple KO, indicating that at least two of these proteases act redundantly. Nevertheless, many proteins still degrade even in the triple KO. LexA, a SOS repressor, cleaves itself under stress and unperturbed growth (Jones and Uphoff, 2021; Little, 1991). Consistent with this, LexA is still degrading in the triple KO. YoaC, an uncharacterized protein, is the shortest-lived protein in the triple knockout with a half-life of 1.2 h. It will be interesting to investigate if YoaC and many other proteins with short half-lives in the triple KO are auto-degrading, degraded by FtsH, or if other mechanisms are at play. Furthermore, we find that the bulk of cytoplasmic proteins is slowly degraded in the triple KO under nitrogen limitation.
A quantitative comparison of protein degradation rates between the triple KO and the individual KOs allows us to assign the contribution of each protease in turning over a substrate (Fig. 4D). We can classify the substrates into six groups: those being degraded predominantly by an individual protease, those where the effects of the individual proteases are additive, those that are stabilized more in the triple KO than the combined effect of individual KOs, and those that are still actively degraded in the triple KO. We identified 64 and 14 substrates to be predominantly degraded by ClpP and Lon, respectively. We only assigned one substrate to HslV: UhpA, a transcriptional regulator that activates the transcription of genes involved in transporting phosphorylated sugars (Weston and Kadner, 1988). 81 proteins are degraded additively, a notable example of which is IbpA, a small chaperone. Previous studies have proposed that Lon either degrades free Ibps or both Ibps and bound client proteins (Bissonnette et al., 2010). We find that both ClpP and Lon contribute approximately equally to the degradation of IbpA, and their contribution is additive. We classify 59 proteins as degraded redundantly. Interestingly, 72 proteins are still confidently degraded in the triple-protease-KO line (Fig. 4E).
To validate our classifications, we compared our protease-substrate relationships with previous proteome-wide measurements. We see a large overlap (p-value = 6E-9) of our identified ClpP substrates when compared with the previous annotation of substrates via trap mutants (Fig. 4F)(Feng et al., 2013). However, we do not observe an overlap of our putative Lon-substrates with a previous Lon-trap experiment (p-value = 0.22) (Arends et al., 2018). This lack of overlap is most likely caused by separating the Lon trap substrates into the different classifications, indicated by a larger overlap with the substrates from all the categories combined (p-value = 0.05). Interestingly, the putative substrates of HslV identified through a microarray study show a strong overlap with the proteins we classify as additive or redundant (p-value = 0.002) (Tsai et al., 2017). This is consistent with the previous reports that HslV substrates are shared with other proteases (Kanemori et al., 1997; Seong et al., 1999). We also find mild enrichment (p-value = 0.08) between substrates identified in a previous FtsH trap (Arends et al., 2016) study and additive or redundant substrates, consistent with findings that FtsH often degrades proteins that are also substrates for other proteases (Biernacka et al., 2020). The lack of overlap between the proteins still degrading in the triple KO and FtsH trap substrates implies that FtsH is likely not involved in the degradation of these substrates.
Surprisingly, 40% of the active protein degradation in WT remains upon knocking out the three canonical ATP-dependent cytoplasmic proteases (Fig. 4G). We could not generate a viable quadruple KO with ftsH deletion, so we cannot completely rule out the possibility that all four proteases act redundantly as an explanation of the remaining protein degradation. However, the results from the individual ftsH knockout (Fig. 4B) and the lack of overlap between degrading proteins and the FtsH-trap experiment (Fig. 4F) are evidence against FtsH being responsible for the remaining degradation. Regardless, a major pathway for degrading proteins in E. coli remains to be discovered: either FtsH has a much bigger role to play than is currently believed, or a completely new mechanism degrades cytoplasmic proteins under nitrogen starvation.
Discussion
This paper introduces a technique for the global measurement of protein turnover on a gene-by-gene basis by combining complement reporter ion quantification with heavy isotope-labeled nutrients (Fig. 1). When we compare protein turnover across E. coli growth rates, we find that active protein degradation remains constant (Fig. 2). Based on this finding, the relative contribution of active degradation compared to dilution due to cell division must change as a function of growth rate. Therefore, relative protein levels of actively degrading proteins must change with differing cell growth rates, or cells must compensate by adjusting transcription and/or translation rates. Protein expression regulation combines gene-specific effects with such global parameters changes (Klumpp et al., 2009). Our insight will help to improve genome-wide protein abundance regulation models and could help to better engineer gene expression circuits with desired properties.
Applying our method to measure protein turnover across multiple nutrient limitations, we find that most cytoplasmic proteins slowly degrade in nitrogen-limited conditions (Fig. 3). However, in phosphorus-limited and carbon-limited conditions, proteins are mostly stable. We observe this phenomenon in both a nitrogen-limited chemostat and in nitrogen-starved batch culture. The slow degradation of cytoplasmic proteins is likely a strategy E. coli has developed to keep scarce amino acids available, which could be critical to various metabolomic processes, including the ability to synthesize new proteins and adapt the proteome to changing environments. Bulk protein turnover measurements in the 1950s showed that Saccharomyces cerevisiae also increases overall protein turnover when starved of nitrogen (Halvorson, 1958), suggesting that a similar strategy might apply even to eukaryotes.
We could assign protein substrates to proteases by measuring the change in protein turnover rates of protease knockout strains (Fig. 4). We were surprised by how little protein degradation changed in the knockout strains, particularly when deleting the canonical proteases ClpP and Lon. We show that in these knockout strains, only a few proteins have a lower degradation rate, and the observed degradation of cytoplasmic proteins continues. Even when we knocked out clpP, lon, and hslV simultaneously, 40% of total protein turnover remains, including the cytoplasmic recycling and the degradation of many short-lived proteins. However, we observe remarkable additive and redundant effects when comparing protein turnover rates in the individual knockouts with the triple knockout. This suggests that many proteins are substrates for more than one protease. We could not extend these approaches to identify substrates for FtsH, as its deletion is lethal due to the accumulation of LPS. However, when combined with a fabZ mutation (Ogura et al., 1999), we could show that the degradation in nitrogen-starved batch culture continues when ftsH is deleted. So far, we have not been able to generate viable quadruple knockout cells for all four known ATP-dependent proteases in E. coli. We, therefore, cannot completely rule out that FtsH is responsible for the remaining cytoplasmic degradation when the other three proteases are deleted. Regardless, a major protein degradation pathway in E. coli still needs to be discovered: either FtsH plays a much bigger role than generally anticipated, or there is a completely different pathway outside the four known ATP-dependent proteases. While protein degradation itself is energetically favorable, unfolding a protein requires energy. E. coli encodes many non-ATP-dependent proteases (Rawlings et al., 2016), but the rapidly turning over proteins in the triple knockout line and the fact that most cytoplasmic proteins are structured, suggest that some ATP-dependent unfoldase is involved. Perhaps, an adapter like ClpX or a chaperone unfolds proteins and allows those substrates to be degraded by one of the proteases that are believed to be energy independent (Rawlings et al., 2016).
The discoveries in this manuscript have been enabled by introducing a new method to measure protein turnover. We chose to use heavy ammonia and glucose to label newly synthesized proteins to avoid the use of mutants and to boost the signal for short labeling times. The resulting MS1 and MS2 spectra are extremely complex, making standard quantification approaches challenging (O’Brien et al., 2020). We have overcome these challenges by taking advantage of the exquisite ability of the complement reporter ion strategy (TMTproC) to distinguish real signals from the chemical background (Johnson et al., 2021). The introduced methods are applicable widely beyond E. coli. Our ability to use comparatively cheap heavy isotope labels opens up the possibility of performing similar studies on larger animals e.g., after D2O intake, which would be cost-prohibitive with heavy amino acid labeling (O’Brien et al., 2020; Sadygov, 2021). Unlike many other cutting-edge multiplexed proteomics approaches, the applied technology is compatible with comparatively simple and widely distributed instrumentation such as quadrupole-Orbitrap instruments, as we avoid the need for an additional gas-phase isolation step. The required analysis software is available on our lab’s GitHub site (https://github.com/wuhrlab/TMTProC).
We have generated a broad resource of protein turnover rates in 12 different growth conditions, each with biological replicates. The investigated conditions include varying cell cycle times from 40 minutes to 12 hours, nitrogen-, carbon-, phosphorus-limitation, and various protease knockout strains. We expect this resource to allow researchers to complement their datasets with protein turnover information. Our measurements of how protein turnover rates change in protease knockout strains will help refine protease substrate relationships. Unlike studies relying on trap experiments or protein microarrays, we could start to deduce the redundant nature of these connections. We have shown the power of the provided resource by showing cytoplasmic recycling in nitrogen-limitation and by finding a scaling law for active protein degradation rates with varying cell cycle times. Thus, we advance protein turnover measurement technology, provide a resource for ~3.2k E. coli protein half-lives under various conditions, and provide fundamental insight into global protein expression regulation strategies.
Acknowledgements
We would like to thank Markus Basan for the gift of triple-protease knockout strain. We thank Michaela Eickhoff, Yihui Shen, Josh Rabinowitz, Joanthon O’Brien, Joe Sheehan, Irina Mikheyeva-Bridges for advice and discussions. This work was supported by NIH grants R35GM128813 (MW), R35GM118024 (TJS), and grant T32-GM007388 (to Princeton University [EMH]), the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under award number DE-SC0018420 and DOE grant DE-SC0018260. This work was supported in part by the National Science Foundation, through the Center for the Physics of Biological Function (PHY-1734030 [NSW]) We gratefully acknowledge support by the American Heart Association predoctoral fellowship 20PRE35220061 (TN), Princeton Catalysis Initiative (MW), Eric and Wendy Schmidt Transformative Technology Fund (MW), Harold W. Dodds Fellowship (MG), NSF Graduate Research Fellowship (ERC), Princeton University’s Summer Undergraduate Research Program (EC).