TMT-Based Quantitative Proteomics Analysis of the Fish-Borne Spoiler Shewanella putrefaciens Subjected to Cold Stress Using LC-MS/MS

College of Food Science and Technology, Shanghai Ocean University, Shanghai 201306, China National Experimental Teaching Demonstration Center for Food Science and Engineering Shanghai Ocean University, Shanghai 201306, China Shanghai Engineering Research Center of Aquatic Product Processing and Preservation, Shanghai 201306, China Shanghai Professional Technology Service Platform on Cold Chain Equipment Performance and Energy Saving Evaluation, Shanghai 201306, China School of Public Administration and Services, Shanghai Urban Construction Vocational College, Shanghai 201415, China


Introduction
Fresh fish is very perishable due to endogenous enzymes and microbial activities, which can result in large economic losses [1]. Indeed, the intrinsic properties of fresh fish are favourable for growth and enzymatic activity of spoilage bacteria, with the consequent off-flavours, off-odours, discoloration, textural changes, and slime formation [2]. Cold storage is widely used to maintain the quality of fish and prolong the shelf life [3,4]. Shewanella putrefaciens is considered as the common specific spoilage organisms (SSOs) in fish during cold storage [5,6]. S. putrefaciens has been reported to be able to use electron acceptors, such as TMAO, instead of oxygen to survive under oxygen or hypoxia conditions. It can produce proteolytic and lipolytic enzymes broken down proteins and produce various flavour defects to lower the fish quality [7]. In addition, S. putrefaciens is a cold-adapted bacterium in refrigerated fish and exhibits many special characteristics and molecular mechanisms that allow them to adapt to the cold stress environment [8,9].
Some bacteria show a variety of physiological adaptation mechanisms, in order to cope with cold stress, survive, and grow in the cold stress environment. e mechanism is as follows: (i) Increased fluidity of cell membranes (ii) e freezing point of the aqueous phase in the cytoplasm decreased (iii) Macromolecules with enhanced stability (iv) Under the effects of cold shock and cold acclimation, the reaction protein of cells to temperature decreases (v) Peroxidase, catalase, redox, and superoxide dismutase protect reactive oxygen species (vi) Whether the catalytic efficiency is maintained under cold stress and under cold stress [10][11][12][13][14].
Proteomics can provide advanced information on microbial metabolism and mechanisms of adaption to the cold stress environment, and this knowledge could be useful to reveal the cold-adaptation mechanisms in S. putrefaciens. Proteomics techniques have been widely used in microbiology, among which two-dimensional electrophoresis and protein identification are commonly used [15]. Due to its high technical reproducibility, improved proteome coverage, and more confident peptide identification and quantification, proteome analysis based on the tandem mass tag (TMT) is suitable for analyzing the abundance of thousands of proteins in complex biological samples [16,17]. Some research studies have been performed to determine the proteomics changes of bacteria under cold stress. For example, proteomics methods were used to investigate the quantitative proteomics of Edwardsiella tarda in the midexponential growth phase at the optimal temperature of 37°C for 24 h and then through the hatch at 4°C for two weeks without vibration. Several key proteins related to DNA synthesis and transcription were significantly upregulated [18]. Similar comprehensive studies for S. putrefaciens have yet to be carried out. To provide insight into potential mechanisms underlying the ability of S. putrefaciens to grow at a temperature of 4°C, we investigated the whole proteome response of S. putrefaciens exposed to cold stress using mass spectrometry.

Bacterial Strain and Growth
Conditions. Broth cultures of S putrefaciens (ATCC 8071) were prepared as follows: 1 mL aliquots of logarithmic phase grown broth cultures were transferred to 250 mL Erlenmeyer flasks containing 100 mL medium. e flasks were incubated aerobically agitating at 200 rpm, at 30 and 0°C, respectively, until an absorbance (OD600) of 0.4 was attained. Six independent replicates were collected for each sample. e cell pellets prepared by centrifugation of the bacterial culture were resuspended and washed three times with phosphate-buffered saline (PBS).

Protein Extraction and Quantification.
e spoilage cells were resuspended in a 600 L lysis buffer and subjected to high-intensity probe ultrasound in a 200 w ice bath (UP-250S sonicator, Scientz, Ningbo, China). e mixtures were centrifuged at 16000 g at 4°C for 5 min. e supernatant was collected, and the protein concentration was quantified using the bicinchoninic acid method. 10 μg protein samples were added to 5X loading buffer at a rate of 5 : 1 (V/V), and then, the mixture was put in the boiling water for 5 min. e purity of proteins was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on the basis of Hou et al. [19].

Protein Enzymatic Hydrolysis and Peptide Desalting.
300 μg of each sample and 0.1 M DTT were mixed together. e mixture was heated in the boiling water for 5 min and then cooled to room temperature. en, 200 μL UA buffer (8 M urea, 150 mM Tris-HCl, pH 8.0) was added to the mixture, and the protein was collected by centrifugation with 10 kDa ultrafiltration centrifuge tube at 12,000 ×g for 15 min. is process was repeated twice. Subsequently, the protein was added with 100 μL IAA (50 mM IAA in UA) and oscillated at 600 rpm for 1 min and centrifuged at 12,000 ×g for 10 min at room temperature in dark. e protein was collected by centrifugation with 10 kDa ultrafiltration centrifuge tube at 12,000 ×g for 15 min, and this procedure was repeated twice.

TMT Labelling.
Desalted peptides were reconstituted in 0.1% FA, and the concentration of peptide was determined with the total protein assay kit (BCA method, Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Peptides were reconstituted in 50 mM 2-hydroxyethyl (pH 8.5) and TMT zero reagent ( ermo Fisher, Waltham, USA) was added from stocks dissolved in 100% anhydrous ACN. e peptide-TMT mixture was incubated for 1 h at 25°C and 400 rpm, and the labelling reaction was stopped by the addition of either 5% hydroxylamine to a final concentration of 0.4% or 8 μL of 1 M Tris, pH 8.0, and incubation for 15 min at 25°C and 400 rpm. Peptide solutions were acidified with 45% (v/v) of 10% FA in 10% ACN prior to drying or directly frozen at 80°C and dried by vacuum centrifugation. For in-depth proteome analyses, peptides derived from Lys-C/trypsin digests of luminal and basal PDX tumors were processed as described in Fang et al. [4] but following the optimized TMT labelling protocol.

LC-MS/MS Measurements.
Tryptic peptides for oneshot analyses were analyzed on an EASY-nLC 1200 ( ermo Scientific) coupled to a Q Exactive Plus mass spectrometer ( ermo Fisher Scientific). After reconstitution in 0.1% FA, an amount corresponding to 500 ng peptides was injected. Peptides were separated on an analytical column (EASY column, 75 μm × 45 cm, ermo Fisher Scientific) applying a flow rate of 300 nL/min and following elution program: 0-2 min: from 5 to 8% solvent B (0.1% formic acid + 98% acetonitrile); 2-42 min: from 8 to 23% solvent B; 42-50 min: from 23 to 40% solvent B; 50-52 min: from 40 to 100% solvent B; and 52-60 min: 100% solvent B, respectively. Mass spectrometers were operated in data-dependent and positive ionization mode. On the Q Exactive Plus, MS1 spectra were recorded at a resolution of 70 k using an automatic gain control (AGC) target value of 1e6 charges and maximum injection time (maxIT) of 50 ms. After peptide fragmentation via higher energy collisional dissociation, MS2 spectra of up to 10 precursors were acquired at 17.5 k resolution using an AGC target value of 1e5 and a maxIT of 50.

Identification of Proteins by Quantitative Proteomics
Analysis.
e results of spectrometry in the present research included protein identification, peptide identification, protein quantification, and differential protein classification analysis. A total of 266670 peptide spectrum matching (PSM) numbers, 19483 unique peptides, and 2292 quantified proteins were obtained after data analysis. e intensity histogram for each sample is shown in Figure 1(a). Figure 1(b) was the box plot of normalized density and represented the box plots of log 2 protein intensity average for each sample.

Identification of Proteins and eir Total and Differential
Abundances. TMT-based quantitative proteomic analysis was developed to identify, quantify, and statistically verify quantitative differences in protein abundance from S. putrefaciens grown at 4°C versus 30°C. e use of this approach ensured that the quantitative differences recorded were reliable, irrespective of the magnitude of the quantitative differences, and provided a robust means of interpreting the biological relevance of the data. A total of six experiments were analyzed by LC-MS/MS. Of these proteins, 274 were significant DE proteins with abundances that changed >1.5-fold (cultivated at 30°C/cultivated at 4°C) and P values of <0.05. A total of 189 proteins were upregulated and 85 proteins were downregulated (red and green background colors, respectively, in Table 1).

Bioinformatics Analysis of DE Proteins Identified by TMT.
e upregulated and downregulated DE proteins were annotated by Gene Ontology (GO) with Fisher's exact test to better understand the roles that these proteins may play in cold adaptation. e significantly upregulated and downregulated DE proteins were classified into three categories using GO terms: biological process (BP), cell component (CC), and molecular function (MF). e downregulated DE proteins were clustered into 50 BP terms (the most representative term was "organic substance metabolic process"), 20 CC terms (the most representative term was "cell"), and 18 MF terms (the most representative term was "binding"). Each of the first ten terms in BP, CC, and MF determined based on P values is listed in Figure 2.

Categorization of Differentially Downregulated Ribosomal
Proteins (RPs). Ribosomes are thought to act as sensors for the heat and cold shock response networks in bacteria and are involved as signals linking environmental stimulus (temperature) with the increased heat shock gene expression [20]. Numerous studies have shown that RPs have a strong functional role, especially in regulating protein synthesis and maintaining the stability of ribosomal complexes [21]. Among the quantified proteins, 44 RPs were identified including 23 30S RPs and 21 50S RPs. Of the 183 significant DE proteins, 31 RPs were upregulated (14 30S and 17 50S, Figure 3) and 21 RPs were downregulated (6 30S and 15 50S).
is result is consistent with the fact that RPs may be important for the correct assembly of rRNA under cold stress. All of these RPs had a significant score based on FC, and the very large number indicates that more RPs were likely downregulated in the cold stress environment.

Main Energy Source Metabolism Network Analysis.
e metabolic network of the main energy sources was established, including the citrate cycle, glycolysis/gluconeogenesis, fatty acid degradation, and main amino acid (valine, leucine, and isoleucine) degradation, to reveal the energy change profiles of S. putrefaciens in cold temperature ( Figure 4). Some of these energy change profiles contained several proteins, and the preferred upregulated proteins are shown. Overall, 13 proteins were identified and quantified in the constructed energy metabolism network. In total, 6 proteins (46.15%) showed a downregulated trend (FC <1), including 2 tricarboxylic acid cycle, 3  were significantly upregulated (FC ≥1.5, P < 0.05). In addition, these two enzymes are involved in other metabolic processes, such as fatty acid degradation and degradation of valine, leucine, and isoleucine. In addition, retinal dehydrogenase 1 and 3-ketoacyl coenzyme A thiolase are also involved in the degradation of fatty acids and metabolic processes such as valine, leucine, and isoleucine [22]. In summary, amino acid and fatty acid degradation were decreased while glycolysis/gluconeogenesis was activated in S. putrefaciens under cold stress.

Interaction Network of Upregulated DE Proteins.
Studies have shown that proteins in living cells do not exist as a single entity, but rather as functional associations within the cell [23]. It was of great significance to reveal the qualitative characteristics of proteins through the interaction between the formations of the network [24]. By using Cytoscape software against the S. oneidensis database, we explored protein interaction networks altered in the cold stress of S. putrefaciens by extracting a putative PPI network. According to the filter criteria of score >400, 23 proteins in PPIs showed significantly differential abundance between S. putrefaciens cultivated at 4°C and 30°C ( Figure 5). e proteins were mainly associated with the cellular metabolic process, organonitrogen compound biosynthetic process, plasma membrane ATP synthesis coupled proton transport, protein metabolic process, ATP synthesis coupled proton transport, and ATP biosynthetic process. Among these proteins, 30S and 50S RPs were upregulated, including those encoded by rpl and rps. Proteins linked to ATP synthase, such as atpA and atpG, were downregulated significantly. We speculated that these proteins had a pivotal role in the network.
Bacteria may encounter a variety of physiological threats under low temperature stress, such as less and less membrane fluidity, less and less enzyme activity, irregular protein folding, lower and lower ice formation and transcription rate in cells, nutrient transport, translation, cell changes, and     even division [25,26]. To overcome these challenges, bacteria employ several cold-adaptation strategies such as ensuring nutrient uptake, maintaining the integrity of membrane structure, retaining ribosome functionality, and facing inefficient, slowing protein folding, decreasing the ability in DNA replication and transcription and in RNA translation [27,28]. Moreover, cold stress affects the membrane fluidity, thus leading the cells to counteract by crafting their membrane lipid composition [26]. Furthermore, bacteria under cold stress protect their DNA, and superoxide dismutase (SOD) and catalase (CATALase) are produced in response to oxidative stress in lipids and proteins due to damage at low temperatures [29].

Energy Metabolism.
We detected 6 downregulated proteins and 7 upregulated proteins related to energy metabolism. A significantly higher level of enzymes involved in glycolysis, which may suggest that S. putrefaciens under implicates the production of high energy intermediates, such as pyruvate (Figure 3). Upregulation of the key enzymes phosphoglycerate kinase and enolase showed similar results in glycolysis. Phosphoglycerate kinase is a housekeeping gene referred to carbon metabolism and energy production [26]. ere are eight DE proteins participating in phosphoglycerate kinase (PGK, EC 2.7.2.3), glycolysis, and enolase, which were more abundant in S. putrefaciens cultivated at 4°C than that at 30°C (S. putrefaciens cultivated at 4°C/S. putrefaciens cultivated at 30°C >1.5) (Figure 3). Glycolysis was a more active energy-producing pathway in S. putrefaciens cultivated at 4°C under nonfavorable temperature, which was also demonstrated by Sun et al. [30]. PGK is found in every domain and in almost all living organisms [31]. It is coded by pgk and PGK activity relates to the conversion of ADP + 3-phospho-d-glyceryl phosphate to ATP + 3-phospho-d-glycerate. PGK is one of the oldest "housekeeping" enzymes, which is found in the most ubiquitous three-carbon portion of the best studied and probably the most ancient metabolic pathway-glycolysis, the Embden-Meyerhof-Parnas cycle (fermentation). PGK activity is also present in several other biochemical processes (e.g., Calvin-Benson-Bassham CO 2 fixation cycle � "CBB") and is often characterized as metabolically essential [32]. Enolase is a rich expression of cytoplasmic protein in many organisms. It plays a very important role, especially at the end of the catabolic glycolytic pathway, it is a key glycolytic enzyme, and its main role is to catalyze the dehydration of 2phosphate-d-glycerol (2-PGA) to produce phosphoenolpyruvate (PEP). In the presence of magnesium ions, it provides energy for organisms, so it is an important energy obtaining way for cells [33]. e upregulation of PGK and enolase was accompanied by the increase of glycolysis rate, which indicates the accumulation of pyruvate.
3.8. RPs. As mentioned above, RPs are one of the most abundant proteins. e effect of cold stress significantly increased their expression (Table 1), indicating their important role in the response to cold stress. Cold stress made the structure of ribosome subunits [34] incomplete, which stalled the translation process and reduced the number of polymers, but this reduction was temporary, accompanied by an increase in the number of a single 70 ribosome and 50 and 30 subunits [15]. Consistent with this, quantities of RPs increased when S. putrefaciens was cultivated at 4°C in the current research. For L. monocytogenes strains, the strongest and most active genomes are associated with ribosomal genes under cold stress [35]. We observed that nearly all of 30S and 50S RPs were upregulated, including those encoded by rpl, rps, and rpm. e RPs themself adapt to the new stress conditions [36]. In vitro experiments showed that the E. coli translation apparatus in the cold environment can preferentially act on mRNAs from coldinduced genes [37]. e drop in temperature reduces cell membrane fluidity at the cellular level, and the active transport and secretion of proteins are also affected. e role of RNA helicase in different desiccants was studied. Helicase helps to unlock the secondary structure of RNA so that efficient transcription and translation processes can be achieved under cold stress [38]. Some studies on coldloving and mesothermal prokaryotes have shown that external ribonuclease and helicase (dead box protein) have this effect; chaperonin proteins have important roles in RNA degradation, nucleotide excision repair pathways, and cold adaptation [39]. ere was evidence that RNA degradation under cold stress helps the cells to adapt its RNA metabolism for subsequent growth under cold stress [40]. With the stability of the secondary structure of DNA and RNA, in the case of reduced transcription and translation efficiency and low protein folding efficiency, RPs need to adapt to cold stress in order to function normally [41]. RPs are responsible for ribosome biogenesis and protein translation and play important roles in controlling cell growth, division, and development [42]. In the current research, 47 PRs were identified as being upregulated. e kinetic properties of RPs are different at 4°C than at 30°C, and a large increase in ribosome proteins may lead to the improvement of translation efficiency. Upregulated RPs   under cold stress could enhance the appropriate translation or function of ribosome assembly in response to growth requirements. Under this circumstance, ribosomes that are more active under cold stress might be very important for S. putrefaciens under cold stress.

3.9.
Translation. In the current study, several factors are upregulated under low temperature stress, such as proteins involved in translation (such as initiation factors IF-2 and extension factors), chaperones involved in protein folding, and proteins involved in transcription (such as DNA-directed RNA polymerase) found in Streptococcus putrefaciens. is shows that S. putrefaciens entered a stabilized phase, which is distinct from the state upon cold shock [15].
In the case of elongation factors, GreA, EF-P, and EF-Ts were upregulated. All were produced at a high level in S. putrefaciens under cold stress, indicating that protein synthesis was maintained to live under cold stress. e transcript cleavage factor, GreA, interacts with the RNAP secondary channel and stimulates the intrinsic transcript cleavage activity of RNAP for the removal of the aberrant RNA 3′ ends. erefore, polymerization activity can be restarted from the end of a cleaved RNA allowing transcription to resume [43]. GreA is essential for the survival of bacteria under stress [44] and it can facilitate RecBCDmediated resection and inhibits RecA, which plays an important role in impeding DNA break repair in E. coli [45], indicating the role of GreA in adapting to the stressful environments. e eukaryotic and archaea extension factor 5A (E/AEF-5A) and its prokaryotic bacterial translation extension factor P (EF-P) would slow the ribosome stagnation. e L-type EF-P is also composed of three bucket domains. Ef-p binds to the polyproline-stalled ribosomes between the peptide-TRNA binding site (P site) and the exiting tRNA (E site) and stimulates the formation of peptide bonds by stabilizing the CCA end of prolyl-TRNA at P site. During translation elongation, the ribosome, with the help of translation elongation factors EF-G, EF-TU, and EF-TS, binds the corresponding amino acid of each codon to the growing polypeptide chain and drives along the coding sequence of the mRNA. Ef-ts are involved in protein synthesis and translation extension, and their expression is increased [46][47][48][49].

Lipid Transport and Metabolism.
e effect of temperature on bacterial membrane lipids has been extensively studied [50,51]. Bacteria adapt their membranes to lower the phase-transition temperature below which their membrane  changes from a "fluid" (liquid-crystalline) to a "rigid" phase to maintain sufficient membrane fluidity under cold stress [52,53]. In the current research, 13 proteins (5 downregulated and 8 upregulated) were associated with lipid transport and metabolism in S. putrefaciens under cold stress. PlsY, lipB, fadR, fadI, and lpxD related to the lipid transport and metabolism were downregulated in S. putrefaciens under cold stress. FadR regulatory protein acts as a regulator controlling bacterial lipid metabolism by inhibiting the fatty acid degradation (fad) system and activating the synthesis of unsaturated fatty acids [54]. In E. coli, FadR is a key gene for the synthesis of unsaturated fatty acids and positively regulates fabA and fabB. However, the FadR in S. putrefaciens under cold stress only regulates

Fatty acid degradation
Gluconeogenseis    fabA fatty acid synthesis gene and the fabB protein did not change significantly, which was similar to the result of Yang et al. [54]. Fatty acid degradation occurs through the wellcharacterized β-oxidation cycle and produces acetyl-CoA, which is further metabolized for energy and precursors of cellular biosynthesis [55]. Acyl-CoA dehydrogenases could catalyze the initial steps in fatty acid β-oxidation. e longchain fatty acyl-CoA ligase could activate free fatty acids to acyl-CoA thioesters and DE proteins related to lipid synthesis including ketoacyl-ACP synthase III (FabH), which was upregulated under cold stress, which shows the protein expression in S. putrefaciens involved in fatty acid elongation and plays a critical role in maintaining the fluidity of membrane [56]. In some research studies, the FabH enzyme of L. monocytogenes prefers 2-methylbutyryl-CoA as the precursor of odd-numbered anteiso fatty acids under cold stress and increases the synthesis of anteiso fatty acids [57].

Conclusions
e ability to canvass a high proportion of the expressed proteome and define quantitatively large or small changes in protein abundance with strict statistical rigour has provided a strong view of S. putrefaciens under cold stress. e proteomics provided functional evidence supporting the importance of specific functional classes of genes that were identified as genomic markers of cold adaptation (e.g., energy metabolism). e quantitative analyses particularly showed the factor in S. putrefaciens under cold stress and the study identified specific pathways that were linked to the cold stress environment. e study provides a platform for comparative analyses of cold adaptation of other bacteria.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.