Proteomic Analysis of the Secretome of Haloarchaeon Natrinema sp

Jan 28, 2014 - In total, 46 predicted Tat substrates, 14 predicted Sec substrates, and 3 class III signal peptide-bearing proteins were detected. Appr...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/jpr

Proteomic Analysis of the Secretome of Haloarchaeon Natrinema sp. J7−2 Jie Feng,† Jian Wang,† Yaoxin Zhang, Xin Du, Zhisheng Xu, Yufeng Wu, Wei Tang, Moran Li, Bing Tang, and Xiao-Feng Tang* State Key Laboratory of Virology, College of Life Sciences, Wuhan University, Wuhan 430072, China S Supporting Information *

ABSTRACT: Although in silico predictions have revealed that haloarchaea can be distinguished from other organisms in that the Tat pathway is used more extensively than the Sec pathway for haloarchaeal protein secretion, only a few haloarchaeal-secreted proteins have been experimentally confirmed. Here, the culture supernatant and membrane fraction of the haloarchaeon Natrinema sp. J7−2 grown at 23% salt concentration were subjected to RPLC−ESI−MS/MS analysis. In total, 46 predicted Tat substrates, 14 predicted Sec substrates, and 3 class III signal peptide-bearing proteins were detected. Approximately 65% of the detected Tat substrates contain lipoboxes, emphasizing the role of the Tat pathway in haloarchaeal lipoprotein secretion. Most of the detected Tat substrates are extracellular substrate (solute)-binding proteins and redox proteins. Despite the small number of Sec substrates, two of them, a cell surface glycoprotein and a putative lipoprotein carrier protein, were identified to be high-abundance secreted proteins. While limited proteins were detected in the culture supernatant, most of the secreted proteins were found in the membrane fraction. The anchoring of secreted proteins to the cell surface via a lipobox or a PGF-CTERM seems to be an adaptation strategy of haloarchaea to handle the harsh extracellular environment. Additionally, ∼15% of the integral membrane proteins (IMPs) detected in the membrane fraction possess putative Sec signal peptides or signal anchors, implying that the Sec pathway is important for membrane insertion of IMPs. This is the first report to describe the experimental secretome of haloarchaea and provide new information for better understanding of haloarchaeal protein secretion patterns. KEYWORDS: haloarchaea, secretion, secretome, Tat pathway, Sec pathway, signal peptide, lipoprotein, mass spectrometry

■

the Sec pathway.5,6 It is unclear, however, why some haloarchaeal proteins are still secreted via the Sec pathway. Probably, there is a selective pressure stronger than protein misfolding in high-salt conditions that directs these substrates for Tat-independent export.6 Although the Tat pathway was predicted to be the dominant export route in haloarchaea, only a few haloarchaeal proteins have been experimentally confirmed to secrete via the Tat pathway by genetic and biochemical approaches, such as α-amylase,1,7 halocyanins,4,8 halolysins SptA9,10 and Nep,11 exoarabinanase, iron-binding protein, DsbA-like thioredoxin domain protein, and maltose-binding protein.8 For haloarchaeal Sec substrates, very few proteins, such as the cell surface glycoproteins from Halobacterium halobium,12 Haloferax volcanii13 and Haloarcula japonica,14 and three lipobox-containing proteins of Hf x. volcanii,15 have been determined experimentally. Additionally, haloarchaea also encode a diverse set of class III signal peptide-bearing proteins, including archaeal flagellins/pilins, substrate binding proteins, and hypothetical proteins.5,16 Although the haloarchaeal

INTRODUCTION Protein translocation across or into the cytoplasmic membrane is an essential biological process that exists in all domains of life. Haloarchaea thrive in extremely saline environments such as solar salterns, salt lakes, and deposits. Investigation of protein translocation in haloarchaea will provide insight into the molecular strategies used by these organisms to adapt to hypersaline environments. In eukarya, bacteria, and nonhalophilic archaea, the general secretory (Sec) pathway is used by most secreted proteins to pass across the cytoplasmic membrane in an unfolded state. By contrast, in silico predictions revealed that the vast majority of haloarchaeal-secreted proteins were substrates of the twin-arginine translocation (Tat) pathway for secretion of folded proteins.1,2 In addition, genetic analysis showed that the haloarchaeal Tat system is essential for viability.3,4 It was hypothesized that, in response to extremely high-salt conditions, haloarchaea route most secretory proteins to the Tat pathway, allowing them to fold correctly in the controlled cytoplasmic environment where chaperones are present.1 Nevertheless, haloarchaea also encode components of the Sec system, and there are several haloarchaeal proteins that are likely cotranslationally delivered to and exported through © 2014 American Chemical Society

Received: July 14, 2013 Published: January 28, 2014 1248

dx.doi.org/10.1021/pr400728x | J. Proteome Res. 2014, 13, 1248−1258

Journal of Proteome Research

Article

flagellins have been well characterized to assemble in a manner similar to that of bacterial type IV pili, and are required for motility, the biological roles of class III signal peptides associated with substrate binding proteins and some hypothetical proteins are still unclear.17,18 Proteomic analysis is a powerful approach for identification of proteins in complex mixtures and has proven to be valuable to validate genome-based predictions. Recently, proteomic techniques were employed for global analysis of protein secretion patterns in bacteria19−21 and fungi,22 as well as in the crenarchaeal genus Sulfolobus,23−25 Antarctic archaeon Methanococcoides burtonii,26 and hyperthermophilic archaeon Aeropyrum pernix.27 Regarding haloarchaea, proteomic analyses were performed on Hbt. salinarum NRC-1,28 Hbt. salinarum,29−35 Hfx. volcanii,36 and Natronomonas pharaonis.37 These studies, however, focus on cytosolic and membrane proteomes of haloarchaea.38 To the best of our knowledge, there is no report of proteomic analysis of the haloarchaeal secretome so far. Natrinema sp. J7 was isolated from a salt mine in China. We previously characterized an extracellular halolysin, SptA, from this haloarchaeon, and mutation analysis of its signal peptide suggested that this protease was a Tat-dependent substrate.9,10 Recently, the complete genome of Natrinema sp. J7−2, a subculture of Natrinema sp. J7 lacking the plasmid pHH205,39 has been sequenced.40 In the present study, we report the proteomic analysis of the culture supernatant and membrane fraction of Natrinema sp. J7−2. Our purpose was (i) to experimentally identify the secretome consisting of the soluble proteins secreted into the extracellular milieu and those anchored/attached to the cell surface; (ii) to validate genome-based predictions concerning Sec and Tat pathways, as well as the flagella/pilus assembly system in haloarchaea; and (iii) to probe the biological roles of Sec and Tat pathways in haloarchaeal protein translocation across or into the cytoplasmic membrane.

■

TCA to remove traces of salt, followed by washing with acetone twice to remove residual TCA. Subsequently, the protein precipitate was dried at room temperature, lyophilized, and used as the protein samples of the culture supernatants. The samples of culture supernatant taken at other growth stages were prepared accordingly, and their protein concentrations were measured by the Bradford assay using bovine serum albumin (BSA) as a standard. The protein concentration from medium alone was subtracted. The membrane fractions were prepared according to the methods described previously29,30 except that the sucrose density gradient centrifugation was omitted to detect possible secreted proteins associated with the membrane. Briefly, the midlog-phase cells were washed with the basal salt solution (4.3 M NaCl, 81 mM MgSO4, and 27 mM KCl), resuspended in the same solution, and disrupted by sonication on ice. The solid debris was removed by centrifugation at 8000 × g for 10 min at 4 °C. Membrane vesicles were collected by three cycles of ultracentrifugation (125 000 × g, 1 h, 4 °C). The final pellet was resuspended in 1 mL of deionized H2O, and the membrane were delipidated as described previously.41 The protein pellets were lyophilized and used as the protein samples of the membrane fractions. In-Solution Digestion of Proteins for Mass Spectrometry

The lyophilized protein samples were digested by trypsin and desalted as described previously.42 Briefly, the lyophilized proteins were resuspended in lysis buffer (8 M Urea, 4 mM CaCl2, 0.2 M Tris-HCl [pH 8.0]), reduced with 10 mM DTT at 56 °C for 30 min, and alkylated with 40 mM iodoacetamide in the dark for 30 min. Protein concentrations were measured by the Bradford assay. Samples were diluted with water until the final concentration of urea was below 2 M, and then were incubated with trypsin (Promega) at 37 °C overnight. The digested peptides were desalted using a SepPak C18 cartridge (Waters) and dried under vacuum for subsequent proteomic analysis.

EXPERIMENTAL PROCEDURES

In-Gel Digestion of Proteins for Mass Spectrometry

Strains and Growth Conditions

The protein samples were dissolved in the loading buffer (50 mM Tris-HCl [pH 6.8], 2% SDS, 10% glycerol, 1% βmercaptoethanol, and 0.1% bromophenol blue), boiled at 100 °C for 5 min, and then were subjected to SDS-PAGE analysis.43 After staining with Coomassie Brilliant Blue R-250, the target bands on the gel were excised and subjected to in-gel digestion.44,45 Briefly, the excised pieces were destained by washing three times with 50% acetonitrile (ACN)/100 mM NH4HCO3, and dried. The dried gel pieces were incubated with a mixture of 10 mM DTT/50 mM NH4HCO3 (pH 8.0) at 56 °C for 30 min and with 40 mM iodoacetamide in the dark for 30 min subsequently. The gel pieces were then washed with 10 mM NH4HCO3 for 10 min and ACN for 10 min, followed by incubation overnight with trypsin solution diluted with 10 mM NH4HCO3. Thereafter, the peptides were extracted from gels using an equal volume of 60% ACN/5% formic acid for 10 min and dried for mass spectrometry.

The Natrinema sp. J7−2 (CCTCC AB91141) strain was grown aerobically at 37 °C in light with shaking at 180 rpm in liquid modified growth medium with 23% total salts (23% MGM) containing 1 g of yeast extract, 5 g of peptone, 184 g of NaCl, 23 g of MgCl2•6H2O, 13 g of MgSO4, 5.4 g of KCl, 0.5 g of CaCl2, and 0.15 g of NaHCO3 per liter. For preparation of the samples for proteomic analysis, the start seed culture from a single colony of Natrinema sp. J7−2 was grown in 5 mL of 23% MGM until the early stationary phase. Next, 100 mL of fresh 23% MGM was inoculated with 1 mL of the start seed culture. The cells were grown under the conditions mentioned above, and the absorbance of the culture at 600 nm (OD600) was monitored to construct the growth curve. The midlog (OD600 ≈ 0.4) and midlog-phase (OD600 ≈ 0.7) cultures were recovered for proteomic analysis. Preparation of Protein Samples

RPLC−ESI−MS/MS and Data Processing

The culture supernatants were collected by centrifugation of the midlog or early stationary-phase cultures at 9000 × g for 10 min at 4 °C. After being passed through a 0.2-μm cellulose acetate filter, the clear supernatant was mixed with an equal volume of cold 40% trichloroacetic acid (TCA) and placed on ice for 1 h. The precipitate was harvested by centrifugation at 11000 × g for 15 min at 4 °C, and then was washed with 20%

Peptide mixtures were subjected to RPLC−ESI−MS/MS analysis using a QSTAR ELITE mass spectrometer (Applied Biosystems) coupled online with a nano flow multiple dimensional HPLC system (TempoTM nanoMDLC system, Applied Biosystems) as described previously.42,44 To maximize proteome coverage, the peptide mixtures from three 1249

dx.doi.org/10.1021/pr400728x | J. Proteome Res. 2014, 13, 1248−1258

Journal of Proteome Research

Article

Table 1. Number of Predicted and Experimentally Detected Secreted Proteins and IMPs in Natrinema sp. J7-2 IMPs with predicted signal peptidesd

secreted proteins with predicted signal peptides b

Sec predicted detected detection rate (%)

without signal peptide

b

Sec

Tata

totalc

A

B

C

D

class III

Tata

totalc

A

B

C

D

142 (87) 46 (30) 32 (34)

101 14 14

45 6 13

46 10 22

26 5 19

72 14 18

18 3 17

8 0 0

119 17 14

46 6 13

44 6 14

15 4 27

78 11 14

674 96 14

a

The Tat signal peptides were predicted by TATFIND, and the value in parentheses represents the number of putative lipoproteins. bThe Sec signal peptides were predicted by eukaryotic SignalP (A), Gram-positive (B) and Gram-negative bacterial SignalP (C), and Pred-signal (D), respectively. c The number of proteins that was positive for at least one of the four predictive options for Sec signal peptides. dSome of the predicted signal peptides of IMPs may function as signal anchors.

1) using the TATFIND program.1 Of these, 142 are predicted secreted Tat substrates, and eight are membrane proteins with multiple transmembrane segments (TMSs). Like the cases of Hbt. salinarum NRC-12 and Nmn. pharaonis,5 most (61%, 87 of 142 proteins) of the predicted Tat substrates of Natrinema sp. J7−2 are putative lipoproteins (Table 1). The annotated Tat substrates are mainly substrate (solute)-binding proteins, redox proteins, and extracellular hydrolases, and nearly all the binding proteins are putative lipoproteins (Supporting Information Table S2). A total of 101 proteins were predicted to be secreted Sec substrates with a signal peptide that possesses the common tripartite structural organization of Sec signal peptides (Table 1) by using the eukaryotic, bacterial Gram-positive and Gramnegative SignalP program,46 as well as the Pred-signal program.47 Seven of these proteins have a predicted C-terminal PGF-CTERM (Supporting Information Table S2), which is a prokaryotic C-terminal protein-sorting signal to anchor the extracellular protein to the membrane lipid.50 Additionally, about 15% (119 of 801 proteins) of the possible integral membrane proteins (IMPs), for example, transporters, membrane peptidases, electron carriers, and the protein secretion apparatus (Table 1 and Supporting Information Table S2), were predicted to bear a Sec signal peptide. Although the putative Sec signal peptides of these IMPs have a potential signal peptidase I (SpI) cleavage site, the possibility that they might actually not be cleavable but act as a transmembrane segment or a signal anchor46,51 cannot be excluded. Using the FlaFind program,16 30 FlaFind positives were predicted in the Natrinema sp. J7−2 genome. Of these, 12 were also predicted to be secreted Sec substrates with a SpI cleavage site and were subtracted from this set. Among the remaining 18 class III signal peptide-bearing proteins, two were annotated as flagellin domain protein and SCP-like extracellular protein, respectively, and the other 16 are hypothetical proteins (Supporting Information Table S2). Nevertheless, at least 7 of the hypothetical proteins are possible components of archaeal flagella/pilus based on the findings that they were predicted to be coregulated with additional FlaFind positives and/or genes coding for FlaI and FlaJ (Supporting Information Table S2), as suggested by Szabo et al.16 The Natrinema sp. J7−2 genome encodes key components of the secretory apparatus, including the signal recognition particle (SRP) complex, SRP-receptor (SR), and Sec and Tat systems (Supporting Information Table S3). In addition, the genome also harbors the genes encoding the VirB11-like ATPase FlaI/ GspE, archaeal flagella/pilus assembly platform protein FlaJ/

independent cultures were subjected to mass spectrometry by RPLC−ESI−MS/MS. Raw data from QSTAR ELITE were analyzed with Mascot Daemon software (version 2.4.0) (Matrix Science, London, UK) using an in-house MASCOT server (version 2.4.0) (Matrix Science, London, UK) as described by Chen et al.42 Briefly, data were searched against the Natrinema sp. J7−2 genome database (4302 sequences) using the following parameters: fixed modifications were set to carbamidomethylation on cysteine, variable modifications were set to oxidation of methionines, as well as phosphorylation at serine, threonine or tyrosine, and the taxonomy was set to “All”. Peptide and MS/MS tolerances were set at 200 ppm and 0.4 Da, respectively. The peptide charge was set to 2+, 3+, 4+, or 5+, allowing for up to two missed cleavages, and the significance threshold was set at p < 0.05. The data of identified proteins were imported into a local MySQL database by a custom-made script for further analysis. Sequences of all peptides assigned are listed in the Supporting Information Table S1, wherein only the peptide gets the highest score is listed if it has been detected more than one time. Mass spectra of each protein annotated by a single peptide are shown in the Supporting Information Figure S1. Motif Predictions

Tat signal peptides were predicted by TATFIND.1 The SignalP 4.0 program46 and the PRED-SIGNAL program47 were employed to identify putative signal peptides from the first 70 residues sequences of all predicted CDSs of the Natrinema sp. J7−2 genome. TMHMM 2.0 software48 was used to predict putative transmembrane segments. The SignalP-TM method of the SignalP 4.0 program was used to predict signal peptides of the proteins containing an N-terminal transmembrane segment. If the predicted signal peptide I (SpI) cleavage sites of the PRED-SIGNAL positives are located in N-terminal transmembrane segments (predicted by TMHMM 2.0), these proteins were regarded as false positives. After excluding the Tat signal peptide-bearing proteins from the SignalP 4.0 and PRED-SIGNAL outputs, the remaining proteins were regarded as the Sec signal peptide-bearing proteins. The archaeal class III signal peptides were predicted by FlaFind.16 Lipoproteins were predicted by LipoP1.0a.49

■

RESULTS

Genome-wide Survey of the Secretory Apparatus and Signal Peptide-Bearing Proteins

Among the 4302 putative proteins of Natrinema sp. J7−2,40 150 were predicted to be Tat signal peptide-bearing proteins (Table 1250

dx.doi.org/10.1021/pr400728x | J. Proteome Res. 2014, 13, 1248−1258

Journal of Proteome Research

Article

of proteins were secreted into the culture medium by the haloarchaeon at this growth stage. In agreement with this, only 14 proteins were detected in the midlog-phase culture supernatants from three independent cultures by RPLC− ESI−MS/MS analysis, including three predicted Tat substrates (Table 2), three predicted Sec substrates (Table 3), and eight signal peptide-lacking proteins (Supporting Information Table S4). All three Tat substrates (NJ7G_2327, 3217 and 4073) are hypothetical proteins lacking a lipobox and are most likely secreted directly into the supernatant via the Tat pathway. The detected Sec substrates include a putative lipoprotein carrier protein (NJ7G_0991), a cell surface glycoprotein (NJ7G_1525) and a hypothetical protein (NJ7G_2252). SDS-PAGE analysis of the midlog-phase samples of Natrinema sp. J7−2 showed that, in comparison with the soluble cytosolic fraction, the culture supernatant contained much less protein, with bands 1, 2, and 3 representing the major components (Figure 1B). The three major bands were excised, and then subjected to in-gel digestion and RPLC− ESI−MS/MS analysis. The results showed that bands 1 and 2 correspond to the cell surface glycoprotein (NJ7G_1525) and the putative lipoprotein carrier protein (NJ7G_0991), respectively, and band 3 is a mixture of hypothetical proteins NJ7G_3217 and NJ7G_2252 (Figure 1B). Notably, all the identified high-abundance proteins are either a predicted Tat substrate (NJ7G_3217) or predicted Sec substrates (NJ7G_0991, 1525, and 2252). The evidence that signal peptide-bearing proteins occupy most of the culture supernatant proteins indicates that: (i) cell lysis of Natrinema sp. J7− 2 is limited at the midlog-phase, and (ii) both Tat and Sec pathways were employed by this haloarchaeon to secrete proteins into the extracellular milieu.

GspF, and type IV pilin-like signal peptidase FlaK/SpIII (Supporting Information Table S3). These proteins are homologues of the components of the bacterial type II and IV secretion systems, as well as the type IV pilus-biogenesis apparatus.52,53 Proteins Detected in the Midlog-Phase Culture Supernatant

Samples from the midlog-phase culture of Natrinema sp. J7−2 were prepared to minimize interference from cell lysis (Figure 1A). It was found that the protein concentration of the culture supernatant remained at a low level (