Proteomics Analysis of Lactobacillus casei Zhang, a New Probiotic Bacterium Isolated from Traditional Home-made Koumiss in Inner Mongolia of China*

Lactobacillus casei Zhang, isolated from traditional home-made koumiss in Inner Mongolia of China, was considered as a new probiotic bacterium by probiotic selection tests. We carried out a proteomics study to identify and characterize proteins expressed by L. casei Zhang in the exponential phase and stationary phase. Cytosolic proteins of the strain cultivated in de Man, Rogosa, and Sharpe broth were resolved by two-dimensional gel electrophoresis using pH 4–7 linear gradients. The number of protein spots quantified from the gels was 487 ± 21 (exponential phase) and 494 ± 13 (stationary phase) among which a total of 131 spots were identified by MALDI-TOF/MS and/or MALDI-TOF/TOF according to significant growth phase-related differences or high expression intensity proteins. Accompanied by the cluster of orthologous groups (COG), codon adaptation index (CAI), and GRAVY value analysis, the study provided a very first insight into the profile of protein expression as a reference map of L. casei. Forty-seven spots were also found in the study that showed statistically significant differences between exponential phase and stationary phase. Thirty-three of the spots increased at least 2.5-fold in the stationary phase in comparison with the exponential phase, including 19 protein spots (e.g. Hsp20, DnaK, GroEL, LuxS, pyruvate kinase, and GalU) whose intensity up-shifted above 3.0-fold. Transcriptional profiles were conducted to confirm several important differentially expressed proteins by using real time quantitative PCR. The analysis suggests that the differentially expressed proteins were mainly categorized as stress response proteins and key components of central and intermediary metabolism, indicating that these proteins might play a potential important role for the adaptation to the surroundings, especially the accumulation of lactic acid in the course of growth, and the physiological processes in bacteria cell.

stress resistance that is known as an adaptation to tolerance response (9). Previous tests implied that the cells of L. casei Zhang at stationary phase were more tolerant to some stresses than exponential phase cells, including acid (pH 2.5), 1.5% oxgall, or heating at 56°C for 1 h; especially the viability of the strain increased from 50 to 90% after acid shock (data not shown).
Considerable research has been focused on the bacterial reaction in the stationary phase, including Enterobacteriaceae (10), Bacillus subtilis (11), Streptococcus mutans (12), and lactic acid bacteria strains (13), to improve the viability of probiotic cultures. Recently in Lactobacillus acidophilus and Lactobacillus plantarum, studies using two-dimensional gel electrophoresis (2-DE) 2 have provided invaluable information for the log adaptive response (14 -16). Monedero et al. (17) have also verified the connection of the phosphotransferase system of L. casei with the cold shock response through a proteomics approach. Nevertheless little is known about the whole protein expression in L. casei. The molecular mechanism of probiotic bacteria for log phase adaptation to tolerance response remains obscure.
In this study, we carried out proteomics research to identify and characterize proteins expressed by L. casei Zhang in the exponential phase and stationary phase. Accompanied by the COG, CAI, and GRAVY value analysis, the study provided a very first insight into the profile of protein expression as a reference map of L. casei. Furthermore knowledge of the proteomics comparison between the exponential phase and stationary phase of L. casei Zhang is useful to enhance the survival of probiotic bacteria in stressful conditions and to improve their technological properties.

MATERIALS AND METHODS
Strain, Medium, and Growth Condition-L. casei Zhang was isolated from traditional home-made koumiss in Inner Mongolia of China. The culture was grown overnight, then diluted in fresh de Man, Rogosa, and Sharpe (MRS) broth (A 600 nm ϭ 0.03) that was adjusted to pH 6.4, and incubated at 37°C. The growth curve was constructed according to each absorbance measurement at 600-nm wavelength at 1-h intervals. The pH values were measured with a pH meter (Rex pHS-3C). The growth phases were confirmed by calibrating the A 600 nm readings against colony-forming units ml Ϫ1 counts.
Preparation of Whole-cell Extracts-The bacteria collected at exponential phase (A 600 nm ϭ 0.7) and stationary phase (A 600 nm ϭ 1.8) were ground into fine powder in the presence of liquid nitrogen. The powdered samples were suspended in cold acetone containing 10% TCA to remove acetone-soluble materials and to precipitate proteins.
After being placed at Ϫ20°C overnight and centrifuged at 35,000 ϫ g for 20 min, the pellet was washed three times with cold acetone and lyophilized. The protein powder was suspended in lysis buffer containing 7 M urea, 2 M thiourea, 4% CHAPS, 2% Pharmalyte, pH 3-10, 1 mM PMSF, 1 mM EDTA, and 65 mM DTT. The solution was sonicated for 5 min and centrifuged at 35,000 ϫ g for 30 min. Proteins in the supernatant were treated by DNase (0.02 units l Ϫ1 ; Fermentas) and RNase (25 g l Ϫ1 ; Fermentas) and then stored in Ϫ80°C freezer. The concentration of each protein sample was determined with the Bradford assay. Samples containing 20 g of total proteins were subjected to electrophoresis under denaturing conditions on 12% polyacrylamide gels according to Laemmli (18).
The first dimension of electrophoresis (electrofocusing) was done with IPG strips (18 cm with a linear range of pH 4 -7; Amersham Biosciences) in an IPGphor at 20°C for 65 kV-h. Prior to the second dimensional electrophoresis, IPG strips were equilibrated with two steps: reduction with DTT and carboxymethylation with iodoacetamide. The equilibrated strips were run on 12% SDS-acrylamide gels (26 ϫ 20 cm; Ettan DALT Twelve system with a programmable power controller). The proteins were visualized with silver (19).
A total of two independent cultures for each growth phase were prepared, and proteins subjected to further analysis were visible in at least two gels from each time point. Two parallel gels were consistent duplicates for any one growth phase and were used for image analysis through an Image Scanner (Amersham Biosciences) in transmission mode. The image files were subsequently analyzed with ImageMaster 2D Elite (Amersham Biosciences). Comparison of spot volumes was performed after background subtraction, normalization, and spot match. The relative volume of each spot was obtained from its spot intensity in pixel units and normalized to the sum of the intensities of all the spots of the gel to calculate the volume % (percent by volume). Proteins displaying at least 2.5-fold volume % variations in a different phase were considered in the present work for statistic analysis (Student's t test). Significant differentially expressed protein spots (p Ͻ 0.05) were considered reliable for further analysis.
Peptide Extraction, MS Analysis, and Protein Identification-Spots from 2-DE were excised and dehydrated with acetonitrile. The gels were rehydrated with 10 mM DTT in 25 mM NH 4 HCO 3 for reduction at 56°C for 1 h followed by 55 mM iodoacetamide in 25 mM ammonium carbonate for alkylation in the dark at room temperature for 45 min. Finally the gel slices or pieces were thoroughly washed with 25 mM ammonium bicarbonate in water/acetonitrile (50:50) solution and completely dried in a SpeedVac. Proteins were digested in 25 l of modified trypsin solution (10 ng l Ϫ1 in 25 mM ammonium bicarbonate) by incubation at 37°C for 12 h. The supernatant containing peptides was stopped by adding 1% TFA directly and used for MALDI-TOF/MS analysis.
Gel spot digestions were desalted by Poros R2 and mixed with 0.6 ml of matrix solution consisting of ␣-cyano-4-hydroxycinnamic acid (12 mg ml Ϫ1 ) in 70% ACN with 0.1% TFA. The slurry was applied onto the target well, dried at room temperature, and injected into a Bruker AutoFlex MALDI-TOF/MS or MALDI-TOF/TOF instrument that was controlled by the Flexcontrol 2.4 package using default parameters (Bruker Daltonics, Karlsruhe, Germany). This instrument was used at a maximum accelerating potential of 19 kV and was operated in reflector mode. The m/z range was from 600 to 4000. Six external standards (Peptide Calibration Standard II, Bruker Daltonics) were used to calibrate each spectrum to a mass accuracy within 10 ppm. MALDI-TOF/TOF tandem mass spectrometry was performed using an AutoFlex TOF/TOF instrument (Bruker Daltonics) operated in "LIFT" mode to confirm some spots.
Database searches for MS or MS/MS spectra were conducted using Mascot software 1.9 (Matrix Science) against the complete proteome NCBInr database for known proteins from firmicutes (NCBInr 2008.11.22; 7,308,559 sequences; 2,531,675,135 residues). The algorithm was set to use trypsin as the enzyme, allowing for one missed cleavage site and assuming carbamidomethyl as a fixed modification of cysteine and oxidized methionine as a variable modification. Peptide mass tolerance and fragment mass tolerance were set to 100 ppm and Ϯ0.5 Da, respectively. Protein hits were considered identified if the Mascot score was greater than 69 for peptide mass fingerprinting and 39 for MS/MS analysis (significance level, p Ͻ 0.05). Other criteria for confident identification were that the protein match should have at least 14% sequence coverage and match at least four peptides.
In Silico Analysis-The CAI of all ORFs of L. casei Zhang 1 was generated by Codon W software in two steps according to Guillot et al. (20). The same software was used for the calculation of the GRAVY value of each protein as described previously (21). Experimentally identified proteins and all ORFs were grouped into cellular roles according to COGs as well.

RESULTS
Growth of L. casei Zhang Strain-To understand the change of growth situation, L. casei Zhang strain was cultured in 100-ml volumes of nutrient-rich MRS medium with a pH of 6.4 for 24 h. The growth curve is shown in Fig. 1. After 3 h of fermentation, the bacteria grew from the lag phase into the exponential phase. The A 600 nm was 0.7 and 1.8 during the mid-exponential phase and early stationary phase after 6 and 10 h of fermentation, respectively. The strain grew up to 10 8 -10 9 colony-forming units ml Ϫ1 at the late stationary phase. During its growth, the bacterium produces mainly Llactic acid with facultative heterofermentation, reducing the pH to 5.6 and 4.2 at the mid-exponential phase and the beginning of the stationary phase, respectively, with an end pH of 3.7. The growth curve was very similar to those of L. plantarum (17) and other lactobacilli strains in skim milk (23).
Reference Map of L. casei Zhang-To understand the profile of protein expression in the growth phases in L. casei Zhang, we analyzed the proteins of bacteria in the exponential phase and stationary phase. The representative 2-DE gels are shown in Fig. 2, and the numbers corresponding to the identified proteins are listed in Table I.
The results showed that the number of protein spots quantified from the gels was about 487 Ϯ 21 (exponential) and 494 Ϯ 13 (stationary). A total of 131 spots were identified among 146 spots, including significant growth phase-related difference spots and some additional spots with high intensity. Although we used NCBInr protein sequences as the database, all the identified proteins were confirmed by the genome information of L. casei Zhang. Ten percent of identification failures might be associated with weakly stained spots. The molecular mass of the identified protein ranged from 16 (spot 3) to 70 kDa (spot 40). Besides, there were 14 proteins, each of them showed in 2-5 spots, namely fructose/ tagatose-bisphosphate aldolase (FbaA), glyceraldehyde-3phosphate dehydrogenase (GapA), pyruvate kinase, tagatose 1,6-diphosphate aldolase, phosphoenolpyruvate-protein kinase, phosphotransacetylase (Pta), L-lactate dehydrogenase (Ldh), GroEL, Hsp20, UspA, B subunit of Asp-tRNA-Asn/Glu-tRNA-Gln amidotransferase (GatB), cysteine  Table I  and supplemental Table 2. Identification of GroEL based on results from MALDI-TOF/MS (C) and MALDI-TOF/TOF spectra of parent ions of 1168.630, 2118.006, and 2149.224 (D, E, and F, respectively) is shown as an example of data quality. Intens., intensity; a.u., arbitrary units. synthase (CysK), S-ribosylhomocysteinase (LuxS), and NAD(FAD)-dependent dehydrogenase.
The cellular localizations of all identified proteins predicted by PSORT Version 2.0 are also listed in Table I. Ninety-eight identified protein spots were in the cytoplasm, eight proteins were predicted to be in the cytoplasmic membrane, and six of them were in ABC-type transport system. Twenty-five spots had unknown cellular localization. Experimentally identified proteins grouped into cellular roles according to COGs are summarized in Table I. From Fig. 3, those related to carbohydrate metabolism and energy production comprise a great part of the identified proteins. L. casei Zhang can utilize many carbohydrates, but in this study we supplied glucose as the major resource.
The comparison of CAI distribution and GRAVY index of the genes coding for the proteins identified on the pH 4 -7 gels with those of the whole set of genes encoding proteins of L. casei Zhang is presented in Figs. 4 and 5, respectively. The CAI calculation is not only used to calculate the codon usage bias for a gene but also demonstrates the high tendency of the corresponding protein synthesis. All genes of identified proteins have a CAI value Ͼ0.5, and over 90% of them have a CAI higher than 0.6. Glyceraldehyde-3-phosphate dehydrogenase (GapA) (spots 89, 90, and 91), belonging to the identified proteins, is one of the top abundant proteins among them all. The result of CAI values implies that proteins were easily identified with relatively high CAI values. The parameter GRAVY index was also calculated for hydrophobicity of all identified and predicted proteins by using Codon W software. This comparison revealed that the GRAVY values of all identified proteins ranged from Ϫ0.7 to 0.1. Only 29 of all identified proteins had a GRAVY index value above zero, indicating that strongly hydrophobic proteins and very hydrophilic proteins might be lost during the preparation of proteins or 2-DE gel electrophoresis and were consequently not detected. These results are similar to those found by 2-DE for other lactic acid bacteria (20,24).
Differentially Expressed Proteins of L. casei Zhang between Exponential Phase and Stationary Phase-We compared 2-DE gel images between exponential phase and stationary phase pairwise to identify differentially expressed proteins (Fig. 2, A and B). Forty-seven protein spots showed statistically significant differences between exponential phase and stationary phase (Table II). Further analysis of the gels indicated that two proteins were produced exclusively in the exponential phase (spots 135 and 144; Fig. 2A), and eight proteins were produced only in the stationary phase (spots 9, 55, 67, 93, 130, 131, 137, and 138; Fig. 2B); one of them was identified as an ABC-type sugar transport system (spot 93).   16), and two proteins involved in lipid metabolism (spots 10 and 50). There was also one hypothetical protein identified (spot 80) as well as three proteins of unclearly known function (spots 31, 75, and 129).    mass fingerprints with the available L. lactis IL1403 genome information made it possible to describe a significant number of cellular pathways related to important physiological processes first at the proteome level. Then Yuan et al. (24) constructed the reference map of Bifidobacterium longum. All of these 2-DE reference maps and most of the main identified proteins might serve as a reference guide to study certain features of lactic acid bacteria.

Verification of Gene Transcriptional Expression-We chose eight proteins (genes) that showed clearly differences in in-
Recently, however, Cohen et al. (16) and Koistinen et al. (15), respectively, generated reference maps of L. plantarum strains in physiological changes occurring between the growth phases. The proteome map of the cytosolic fraction of different growth conditions is able to give a global picture of the relative abundance of proteins during growth phases in the selective medium for lactobacilli. Therefore, we built the reference map of L. casei Zhang upon the 2-DE gels of the exponential phase and stationary phase.
We constructed the proteomic map of probiotic strain L. casei Zhang with a pI at pH 4 -7. The theoretical 2-DE map of pI 3-14 analysis and a preliminary 2-DE map with a pH gradient from 3 to 10 non-linear indicated that most L. casei Zhang proteins were distributed within pI 4 -7 (data not shown). Previous studies have discovered that most proteins of lactic acid bacteria are located within this narrow pH range. To obtain reproducible protein patterns, we chose MRS medium, which is widely used for lactic acid bacteria, and always collected the L. casei Zhang culture at a given absorbance according to the growth curve. The protein extraction method has also been successfully used to establish the 2-DE map of another L. casei strain, XM2-1, from same origin (data not shown).
From ϳ490 spots on the gels in this study, a total of 131 spots were identified (Table I). Among these identified proteins, proteins with housekeeping functions in carbohydrate metabolism were the major spots on the acidic map, and CAI values of corresponding genes were all higher than 0.7. Thus glycolytic enzymes associated with fermentative metabolism appear to be playing an important role in L. casei Zhang. Similar results were found in the studies mentioned above. The results also showed that there were 47 differentially expressed proteins between the exponential and stationary phase among which 70% up-shifted in the stationary phase. This indicated more strongly up-regulated proteins than down-regulated changes in the stationary phase in L. casei Zhang. From another point, the reference map of L. casei Zhang made upon different growth phases might be able to provide more detailed insight into metabolic pathways and their regulation in time during growth. In addition, we observed a number of reproducible vertically bunched spots of low molecular proteins in each phase gel that may be modified and that need further explanation (Fig. 2).
Differentially Expressed Proteins of L. casei Zhang between Exponential Phase and Stationary Phase-We compared the exponential phase and stationary phase of L. casei Zhang growing in MRS medium, and the results showed that there were 33 spots up-shifted significantly over 2.5-fold in the stationary phase among which we focused our attention on the 93% identified proteins. Lactic acid is constantly produced during sugar fermentation; this implies a frequent confrontation of the lactic acid bacteria cells with acid stress. Therefore, the overproduction of these proteins might be linked to the adaptation to the acidic environment according to analysis of an up-shifted 16-kDa protein in the stationary phase responding to the acidic environment of strain S. thermophilus PB18 (29).
The differentially expressed proteins included general stress response proteins, such as DnaK, GroEL, Hsp20, and UspA-related proteins. The induction of DnaK (spot 41, LCAZH_1552) and GroEL (spots 20, 21, and 22; LCAZH_2207) we found might be favored by the acidic conditions in which lactic acid interacts with the cells of stationary phase. The chaperones play a key role in the maturation of synthesized proteins and are pivotal in the degradation or refolding of denatured proteins (30), which may interact with the glycolytic enzymes at low pH and increase their stability in the presence of a gradually changed acid challenge. These heat-shock proteins can be induced by multiple stresses, such as acid, heat, bile salts, high pressure stress, and so on (31)(32)(33)(34). Enormous progress has been made in the elucidation of the major chaperones belonging to the Hsp60 (GroEL) and Hsp70 (DnaK) families because protein folding has been recognized as one of the central problems in biology (35). The study of GroESL-overproducing L. lactis and Lactobacillus paracasei NFBC 338 demonstrated that technologically sensitive cultures can be potentially manipulated to become more robust for survival under harsh conditions, such as food product development and gastrointestinal transit (36). Genes encoding the transcriptional repressor HrcA (LCAZH_1554) and the accessory proteins GrpE (LCAZH_1553) and DnaJ (LCAZH_1550), which comprise the DnaK operon, were not identified on the current gels. The GroES protein (LCAZH_2208) was not found in the current 2-DE; this might be related to the low molecular mass (near 10 kDa). Nevertheless there are corresponding ORFs in the genome of L. casei Zhang.
We further observed up-regulation of the low molecular weight heat-shock protein (small HSP (sHSP)) Hsp20 (spot 1, LCAZH_0619; spot 3, LCAZH_2811) in response to low pH stress. Our understanding of other chaperones except for DnaK and GroEL is comparatively limited. The main function of these sHSPs appears to be the prevention of the accumulation of unfolded protein intermediates during periods of stress (37). Although the sHSPs are always constitutively expressed, the rate of synthesis is significantly enhanced under stress (38).
Except for the HSPs, universal stress protein UspA-related nucleotide-binding protein was induced strongly in the stationary phase (spot 13, LCAZH_1180). The levels of UspA have been evaluated in the response to cell stasis of E. coli (39,40) under a large variety of stress conditions (41,42). The UspA protein of E. coli is required for resistance to DNA damage. Based on the conclusion, the induced expression of nucleotide-binding protein UspA in L. casei Zhang may be associated with DNA protection of cells during stationary growth phage. Yet, the exact physiological role of UspA in L. casei Zhang requires further study.
In the study, there were eight proteins belonging to carbohydrate and energy metabolism (spots 44, 45, 85, 86, 54, 12, 14, 82, 35, and 73) that were significantly up-regulated over 2.5-fold. In many microorganisms the control of glycolytic flux depends on the activity of phosphofructokinase (Pfk) and pyruvate kinase (43). Based on our results, the expression of pyruvate kinase protein (spots 44 and 45, LCAZH_1352) increased above 3.0-fold, whereas the intensity of GapA (spots 89, 90, and 91; LCAZH_0910) and Pfk (spot 37, LCAZH_1351) were unchanged. However, these proteins are overexpressed in lactic acid bacteria and other microorganisms under environmental stress conditions (44,45). The results from L. lactis showed that pyruvate kinase is an important bottleneck to carbon flux only when glucose becomes limited (46). When the resources of a medium start to become limited, the cell division rate of bacteria slows down, and the bacteria enter into the stationary phase. The stationary phase was chosen to represent this transition in our study. Therefore, pyruvate kinase is playing a pivotal regulatory role in the glycolytic metabolism of L. casei Zhang when the strain entered the stationary phase. Moreover the enhanced expression of glucosamine-6-phosphate isomerase (spot 35, LCAZH_2896), which catalyzes the conversion of the amino sugar glucosamine into fructose 6-phosphate, might suggest that L. casei Zhang was switching from using glucose as a carbon source to alternative sources and pathways with the decreasing glucose content of the medium at the stationary phase.
The intensity of TDP aldolase (spots 85 and 86, LCAZH_ 0609) increased by 2.5-fold in the stationary phase. TDP aldolase is a key enzyme in galactose metabolism, which catalyzes the conversion of D-tagatose 1,6-diphoshate to Dglyceraldehyde 3-phosphate that subsequently participate in glycolysis. Based on the analysis of the genome of L. casei Zhang, the whole genes for the tagatose 6-phosphate pathway were found (LCAZH_0608 -LCAZH_0613), including one phosphotransferase system (PTS) galactose operon. 3 The enhanced expression of TDP aldolase may result in the increase in ATP production to support the increased H ϩ extrusion under acidic conditions. Just like the significant overexpressions of some glycolytic proteins of the probiotic bacterium Lactobacillus reuteri in response to acid stress, Lee et al. (44) suggested that under low pH conditions the cellular metabolism shifted toward production of more energy-rich intermediates, such as ATP, and NADH for survival at low pH. Galactose mutarotase-related enzyme (spot 12, LCAZH_2563), which might be involved in galactose metabolism, was upregulated as well.
Moreover GalU (spot 54, LCAZH_1073), a protein that participates in the conversion of glucose 1-phosphate into UDPglucose, was observed to be up-regulated over 2.5-fold. UDP-glucose is one essential substrate for the biosynthesis of cell wall components. This suggested that the proteins predicted to be involved in cell wall structures such as UDPsugars were increased in relative abundance at the beginning of the early stationary phase at the cost of the glucose 1-phosphate pool, which would result in less availability of glucose 1-phosphate for glycolysis (16). Subsequently UDPglucose was converted into UDP-galactose by the Leloir enzyme GalE (spot 92, LCAZH_0597) encoded by the galE gene, which is located in the galKETRM operon of L. casei Zhang. 3 In addition, the increased expression of N-acetylglucosamine-6-phosphate deacetylase (NagA) (spot 73, LCAZH_1799) might be playing a role in strengthening the biosynthetic pathway required for cell wall peptidoglycan biosynthesis (47). The biosynthesis of these sugar nucleotides also serve as precursors for EPS biosynthesis (48). By comparative analysis with L. casei ATCC334, we found two insertions encoding two eps clusters in L. casei Zhang (EPSC1 and EPSC2). 3 PDH E1 is one of the three subunits (E1, pyruvate dehydrogenase, LCAZH_1299 and LCAZH_1300; E2, dihydrolipoamide dehydrogenase transacetylase, LCAZH_1301; and E3, lipoamide dehydrogenase, LCAZH_1302) of the pyruvate dehydrogenase multienzyme complex, which is an assemblage that plays a pivotal role in cellular carbohydrate metabolism, catalyzing the oxidative decarboxylation of pyruvate and the subsequent acetylation of CoA to form acetyl-CoA (49). During the process of pyruvate dehydrogenase enzyme complex reactions, PDH E1 is responsible for the first step of the process and catalyzes pyruvate decarboxylation followed by transfer of the hydroxyethyl group to thiamine diphosphate, which together with Mg 2ϩ acts as the reaction cofactor (50). The ␤ subunit of pyruvate dehydrogenase complex E1 component (spot 14, LCAZH_1300) showed enhanced expression in the stationary phase.
We also identified acid stress-associated changes in the expression levels of proteins involved in regulating translation and the transport system, the quorum sensing system, and lipid metabolism. The synthesis rate of elongation factor (EF)-Tu (spot 69, LCAZH_1321) increased in L. casei Zhang in the stationary phase; EF-Tu is involved in the delivery of aminoacyl-tRNA to the ribosome (51). However, it has been suggested that in E. coli, in addition to its function in translation elongation, EF-Tu might be involved in protein folding and/or protection from stress (52). The overexpression of EF-Tu has also been reported during acid adaptation in Propionibacterium freudenreichii and S. mutans (53) as well as in Listeria monocytogenes in salts (54). Thus, the overexpression of EF-Tu, which has a protective influence on newly produced proteins, in this study is probably important in protein folding and protein renaturation and may be due to the stress tolerance in the stationary phase of L. casei Zhang.
Under conditions of low pH, membrane transport systems are required to counteract diffusion and transport of solutes (ions and nutrients) across the membrane and to transport compatible solutes into or out of the cell (55). ABC transporters form a superfamily of diverse membrane proteins that utilize the energy derived from ATP hydrolysis to fuel the transport of solutes across the cell membrane (56). About 8% of the total ORFs of L. casei Zhang encoding ABC transport proteins were identified. In our study, one of the ABC transport proteins was only expressed in the stationary phase; it was identified as ABC-type sugar transport system (spot 93, LCAZH_0939). Another ABC-type multidrug transport system (spot 113, LCAZH_0533), belonging to the defense mechanism, was found to be up-regulated over 3.0-fold in the stationary phase. Although the precise roles of these proteins in the response to acid stress in streptococci has yet to be elucidated, an ABC transporter with homology to genes found in both Bacillus licheniformis and Staphylococcus epidermidis, which both function to confer resistance to antibiotics, made a significant contribution to the ability of S. mutans to grow at low pH (57). Therefore, the induced expression of ABC transports might be associated with the low pH of the stationary phase.
Quorum sensing is a mechanism for regulating gene expression in response to changes in cell density of a bacterial population (58). The production of autoinducer-2 (AI-2), mediated by the activity of the LuxS enzyme, has been demonstrated in recent years. The LuxS-deficient strain of S. mutans presented increased acid killing sensitivity and a higher survival rate under hydrogen peroxide condition (59,60). Previous studies have shown that the L. reuteri strain and probiotic strain Lactobacillus rhamnosus GG were able to produce AI-2-like molecules, and LuxS or AI-2 was involved in multicellular behaviors like biofilm formation (61). However, the work by Winzer et al. (62) indicated that the AI-2 molecule might not act as a signal molecule for the majority of bacteria but was the by-product of an important metabolic pathway. The role of the increasing expression of LuxS (spot 6, LCAZH_0709) in L. casei Zhang still remains to be revealed.
The synthesis of two esterases, EstC (spot 10, LCAZH_0386) and EstA (spot 50, LCAZH_1846), was observed to be up-shifted above 2.5-fold in the stationary growth phase compared with the exponential phase. Yebra et al. (63) have demonstrated that EstC from L. casei BL23 was co-transcribed with genes encoding a mannose class sugar PTS. However, no PTS-related proteins were detected to be overproduced in the same growth phase. Therefore, more detailed study of EstC in L. casei Zhang is needed.
Further we detected one ribosomal protein (spot 16, LCAZH_1573) that was up-regulated in the stationary phase. Ribosomal proteins are thought to act as sensors of heat and cold shock (64), whereas the RNA/protein ratio of the cells increased with growth rate in Vibrio natriegens and E. coli (65).
We also found two ribosomal proteins that were up-regulated when comparing the exponential and stationary phases by 2-DE with pI ranging from 3 to 10 non-linear (data not shown). PepC (spot 72, LCAZH_2303) and PepP (spot 19, LCAZH_1633) were detected as having enhanced expression in the stationary phase. We localized two other peptidases, PepN (LCAZH_0499) and PepT (LCAZH_1062), in the genome of L. casei Zhang. The function of peptidases and transport systems is to obtain a free amino acid supply from peptides for bacteria growth. PepO and PepC have been shown recently to belong to the CodY regulon of L. lactis that is sensitive to the intracellular branched chain amino acid content (66).
Finally we observed that S-adenosylmethionine synthetase (spot 66, LCAZH_0820), which is an important substrate for the modification of rRNA nucleotides, polyamine synthesis, and methylation processes, was upshifted during the stationary phase. Aspartate racemase (spot 128, LCAZH_0248), which converts D-aspartate to L-aspartate and phosphoribosyl aminoimidazolesuccinocarboxamide synthase (spot 145, LCAZH_1745), which is important for purine metabolism, were both overexpressed during stationary phase. We also observed expression changes in a set of proteins lacking appropriate functional annotations, including metal-dependent hydrolase of the ␤-lactamase superfamily III (spot 31, LCAZH_1728), aldo/keto reductaserelated enzyme (spot 75, LCAZH_0702), saccharopine dehydrogenase-related protein (spot 129, LCAZH_2188), Llactate dehydrogenase (FMN-dependent)-related ␣-hydroxy acid dehydrogenase (spot 82, LCAZH_2316), and a hypothetical protein (spot 80, LCAZH_0632); all were induced expressions in the stationary phase. The different expression patterns of function-unknown proteins may also imply that they have important roles in this biological process, and further functional studies would bring a better understanding of the molecular mechanism. Future work will be required to clarify the roles of the unknown proteins in the acid stress response.
Verification of Gene Expression by qRT-PCR-We did further investigations to address what actually happens in the 12 h at the molecular level during the growth of L. casei Zhang. Using qRT-PCR, we investigated the transcriptional profiles of a few important proteins, and they have different expression patterns. Among general stress proteins, LCAZH_1552 (DnaK) and LCAZH_2811 (Hsp20-2), LCAZH_2207 (GroEL), and LCAZH_0619 (Hsp20-1), respectively, had similar expression patterns, indicating a similar transcription level. Compared with 2-h expression, the sharp increase of GroEL and Hsp20-1 at 12 h of cells might be related to stress of the stationary phase. The expression of LCAZH_1552 (DnaK), LCAZH_2811 (Hsp20-2), LCAZH_2896 (GalU), LCAZH_1299 (␤ subunit of pyruvate dehydrogenase complex E1 component), LCAZH_2563 (galactose mutarotase-related enzyme), and LCAZH_0386 (EstC) peaked at 8 or 10 h and then came down. The mRNA expression of these proteins at 10 h was still higher than at 6 h by over 1.5-4.2fold except for LCAZH_2563 (galactose mutarotase-related enzyme), LCAZH_1299 (␤ subunit of pyruvate dehydrogenase complex E1 component), and LCAZH_0386 (EstC). That is to say, the expression of LCAZH_1552 (DnaK), LCAZH_2811 (Hsp20-2), LCAZH_2207 (GroEL), LCAZH_0619 (Hsp20-1), and LCAZH_2896 (GalU) exhibited a continuously increasing trend, which matched with the protein expression level. However, the level of LCAZH_0386 declined to lower than 1.0 at 10 h, and similar expression of LCAZH_1299 and LCAZH_ 2563 was found between 6 and 10 h, showing differences in expression levels between proteins (translation) and mRNA (transcription). This might be associated with early translation and trafficking of stored mRNAs during the exponential phase. After all, the results of qRT-PCR demonstrate continuous synthesis of the genes of interest at the transcription level.
Conclusion-A reference proteome map of intracellular proteins of L. casei Zhang has been established using a proteomics approach. For the first time we have described the proteome of L. casei based on the growth phase and analyzed the COG, CAI, and GRAVY values of each identified protein compared with the whole theoretical genes coding proteins of L. casei Zhang. Our proteome analysis extends previous studies in regard to the physiological characteristics and more importantly confirms the expression of a large number of proteins related to their habitat. By comparing the proteomic profile of L. casei Zhang grown in the exponential phase with that in the stationary phase, we further demonstrated that the differentially expressed proteins were mainly categorized as stress response proteins and key components of central and intermediary metabolism that allow the strain in the stationary phase to withstand harsh conditions and sudden environmental changes and for most of the population to develop crossprotection against multiple stresses. The results indicated that L. casei Zhang evolves stress-sensing systems and defenses against stress according to differentially expressed proteins at different phase stages. We did further investigations to address what actually happens in the 12 h at the molecular level during the growth of L. casei Zhang. Using qRT-PCR, we investigated the transcriptional profiles of a few important proteins that showed different expression patterns.
The 2-DE reference maps based on growth phase can facilitate further studies and provide information about the activity and metabolic processes of the cells under various conditions for industrial applications. The big challenge for future studies is to provide an experimental protocol covering the fraction of intrinsic membrane proteins and alkaline proteins that almost totally escaped detection by the experimental procedure used in this study and to explain the reasons for the exchanges of proteins during growth.