Extracellular Polysaccharide-Degrading Proteome of Butyrivibrio

ARTICLE pubs.acs.org/jpr

Extracellular Polysaccharide-Degrading Proteome of Butyrivibrio proteoclasticus Jonathan C. Dunne,†,‡,§ Dong Li,† William J. Kelly,† Sinead C. Leahy,† Judy J. Bond,†,‡,§ Graeme T. Attwood,† and T. William Jordan*,‡ †

Rumen Microbial Genomics, Food Metabolism and Microbiology Section, Food and Textiles Group, AgResearch Limited, Grasslands Research Centre, Palmerston North 4442, New Zealand ‡ Centre for Biodiscovery and School of Biological Sciences and §AgResearch Limited/Victoria University of Wellington Proteomics Laboratory, Victoria University of Wellington, New Zealand

bS Supporting Information ABSTRACT: Plant polysaccharide-degrading rumen microbes are fundamental to the health and productivity of ruminant animals. Butyrivibrio proteoclasticus B316T is a Gram-positive, butyrate-producing anaerobic bacterium with a key role in hemicellulose degradation in the rumen. Gel-based proteomics was used to examine the growthphase-dependent abundance patterns of secreted proteins recovered from cells grown in vitro with xylan or xylose provided as the sole supplementary carbon source. Five polysaccharidases and two carbohydrate-binding proteins (CBPs) were among 30 identified secreted proteins. The endo-1,4-β-xylanase Xyn10B was 17.5-fold more abundant in the culture medium of xylan-grown cells, which suggests it plays an important role in hemicellulose degradation. The secretion of three nonxylanolytic enzymes and two CBPs implies they augment hemicellulose degradation by hydrolysis or disruption of associated structural polysaccharides. Sixteen ATP-binding cassette (ABC) transporter substrate-binding proteins were identified, several of which had altered relative abundance levels between growth conditions, which suggests they are important for oligosaccharide uptake. This study demonstrates that B. proteoclasticus modulates the secretion of hemicellulose-degrading enzymes and ATP-dependent sugar uptake systems in response to growth substrate and supports the notion that this organism makes an important contribution to polysaccharide degradation in the rumen. KEYWORDS: Butyrivibrio proteoclasticus, lignocellulose, proteome, rumen, xylanase

’ INTRODUCTION Ruminant animals have a highly evolved symbiotic relationship with a complex rumen microbiota that allows them to utilize insoluble plant polysaccharides as their main energy source.1 The microbial degradation of lignocellulose and fermentation of the released soluble sugars produces short-chain fatty acids that are absorbed across the rumen epithelium, while microbial biomass that leaves the rumen is digested in the small intestine and provides more than half the total protein supply of the animal.2 However, in pasture-based systems the conversion of lignocellulose to utilizable energy is often poor, and several studies have demonstrated that enhanced degradation of high fiber forage diets leads to improved ruminant performance in terms of animal weight gain and milk production.3 7 In particular, the degradation of the hemicellulose component of forage material is believed to be a limiting factor in lignocellulose utilization,8 due to the array of covalent linkages within hemicellulose9 and between hemicellulose and the closely associated polymers such as pectin10 and lignin.9 Hemicellulose of forage plants is a chemically complex and structurally diverse substrate comprising mainly a xylan backbone composed of (1f4)-β-linked D-xylopyranose (xylose) monomers that are substituted to varying degrees with α-L-arabinan and r 2011 American Chemical Society

α-L-arabinofuranosyl, O-acetyl, and α-glucuronic or 4-O-methylacid groups. The α-L-arabinofuranosyl side groups may also be esterified with ferulic and p-coumaric acids11 that cross-link to other xylan chains or to lignin. As a result of this complexity hemicellulose degradation requires the synergistic activity of a consortium of polysaccharide-degrading enzymes that are either glycoside hydrolases (GHs), carbohydrate esterases (CEs), or polysaccharide lyases (PLs). Examples include endo-β-1,4-xylanases (endoxylanases) (EC3.2.1.8) that cleave internal linkages within the xylan backbone, β-xylosidases (EC3.2.1.37) that hydrolyze xylose monomers from the nonreducing ends of xylooligomers, and debranching enzymes such as arabinases (EC3.2.1.99), α-L-arabinofuranosidases (EC3.2.1.55), acetylxylan esterases (EC3.1.1.72), ferulic acid esterases (EC 3.2.1.73), and p-coumaric acid esterases (EC3.2.1.-). In the rumen, bacteria belonging to the genera Butyrivibrio, Pseudobutyrivibrio,12,13 and Prevotella are regarded as the primary hemicellulose-degraders and make an important contribution to D-glucuronic

Special Issue: Microbial and Plant Proteomics Received: August 31, 2011 Published: November 07, 2011 131

dx.doi.org/10.1021/pr200864j | J. Proteome Res. 2012, 11, 131–142

Journal of Proteome Research

ARTICLE

ruminant energy supply.14,15 We recently reported the genome sequence of Butyrivibrio proteoclasticus B316T (subsequently referred to as B316T),16 a low G+C%, Gram-positive polysaccharidedegrading, butyrate-producing bacterium that is present at high numbers in the rumen of animals consuming pasture or grass silage based diets17 20 and is commonly detected in rumen bacterial 16S rDNA libraries.21 On the basis of the near full-length 16S rRNA gene sequence analysis, B. proteoclasticus is taxonomically most closely related to the fibrolytic Butyrivibrio/Pseudobutyrivibrio group of bacteria (family Lachnospiraceae), which belongs to the Clostridial rRNA subcluster XIVa.18 B316T cells growing in vitro can utilize a variety of soluble mono- and disaccharides including arabinose, cellobiose, galactose, glucose, rhamnose, sucrose, and xylose.18,22 In addition, B316T is one of a small number of rumen fibrolytic microbes capable of efficiently degrading and utilizing the insoluble, complex polysaccharide xylan and is also capable of utilizing pectin and starch.18,22 Metabolic footprinting of several Butyrivibrio species demonstrated that B316T cells cultured in glucose-, xylan-, and pectin-containing medium degraded more fiber, liberated more than 10-fold the amount of xylose, and utilized almost 10-fold more glucose than the closely related and fibrolytic B. fibrisolvens AR10.23 The annotated B316T sequence showed that 6% of the genome is devoted to the degradation, uptake, and utilization of plant polysaccharides and includes 129 GHs, CEs, and PLs classified within 44 different CAZy database families (www.cazy. org).24 Many of these are annotated as endo-1,4-β-xylanases, β-xylosidases, or α-L-arabinofuranosidases, which reflects the ability of B316T to grow well on xylan in vitro.22 Almost half of these enzymes are clustered within 34 polysaccharide utilization loci (PULs)25 that also include transporters, transcriptional regulators, environmental sensors, and genes involved in further metabolism.16 We are undertaking functional analysis of the polysaccharide-degrading capability of this rumen bacterium, and in this paper we report the proteomic examination of B316T secreted proteins (secretome) with particular emphasis on the enzymes involved in hemicellulose degradation. This research improves our understanding of the contribution of Butyrivibrio species to ruminant digestion and the complex conversion of plant material to milk and meat for human consumption and also gives an improved choice of enzymes with which to manipulate and enhance fiber degradation in ruminants.

protein pellets were dissolved in 2-DE sample buffer containing 7 M urea, 2 M thiourea, 2% CHAPS (w/v), and 50 mM DTT. For 1-DE, protein pellets were dissolved in 1 SDS sample buffer containing 50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 20 mM DTT. 2-DE and Image Analysis

Three biological replicates were analyzed for each growth condition, and each biological replicate was analyzed in triplicate, for a total of nine gels per sample. Prior to IPG strip rehydration the concentration of each protein sample was assayed (2-D Quant Kit, GE Healthcare, Uppsala, Sweden), and fresh 2-DE buffer was added to achieve a final protein concentration of 0.4 μg μL 1. Each sample was supplemented with 0.5% (v/v) IPG Buffer (GE Healthcare, Uppsala, Sweden), and 125 μL was applied to 7 cm IPG strips. All strips were passively rehydrated at room temperature for 16 h, and IEF was conducted for 9 11 kVh using an IPGphor II IEF unit (GE Healthcare, Uppsala, Sweden). Focused IPG strips were equilibrated for 15 min in 1 LDS Buffer (Invitrogen, Carlsbad, CA, USA) containing 1 Reducing Agent (Invitrogen, Carlsbad, CA, USA) and then for 15 min in 1 LDS Sample Buffer containing 100 mM iodoacetamide (Bio-Rad, Hercules, CA, USA). Second dimension electrophoresis was performed using a 1 mm thick NuPAGE Novex 4 12% Bis-Tris ZOOM gel in a NuPAGE MOPS Buffer system (Invitrogen, Carlsbad, CA, USA) at constant 200 V. Protein spots were visualized using colloidal CBB-G250, and gel images were acquired using a Molecular Dynamics Personal Densitometer SI (Sunnyvale, CA, USA). All spot volumes were within the linear range previously defined for colloidal CBBG250 staining (Supporting Information, Figure S1). Image analysis and gel matching and spot set creation was performed using Image Master 2D Platinum software (Version 5.0, GE Healthcare, Uppsala, Sweden), and normalized spot volume values, defined as the spot volume expressed as a percentage of the total detectable spot volumes, were exported and log-ratio transformed.26,27 A spot set was created for each protein species when the spot could be matched in at least seven of the nine replicate gels. Spot sets with more than two missing values were not included in the differential abundance analysis. Missing data points were imputed using the lowest spot volume value from the respective gel that contributed to a matched spot set. Statistical analysis of protein abundance patterns was performed using SPSS (Statistical Package for the Social Sciences) (Ver. 14.0). Two-tailed Student's t tests (p < 0.01) were used to assess statistical significance of protein abundance changes and were calculated as the mean ( SEM using three biological replicate experiments.

’ MATERIALS AND METHODS Culture Conditions

B316T cells were grown anaerobically at 37 °C in modified DSM Medium 704 (see Supporting Information for composition and preparation) containing 0.1% (w/v) oat-spelt xylan or 0.5% (w/v) xylose (Sigma-Aldrich, St. Louis, MO, USA). All cultures were grown in triplicate, and mid-log phase and stationary phase cultures were harvested at approximately OD600 = 0.5 and 0.7, respectively. Culture supernatants were recovered by centrifugation at 3,000 g for 30 min at 4 °C and supplemented immediately with Complete Protease Inhibitor (Roche Applied Science, Mannheim, Germany) before freezing at 80 °C.

Protein Identification

Protein spots were excised and processed for MALDI-TOF MS as described in Beddek et al.,27 using 10 mg mL 1 CHCA in 50% ACN/0.1% TFA. Positive-ion tryptic PMF were acquired using a Voyager DE Pro MALDI-TOF MS (Applied Biosystems, Foster City, CA, USA) across a 700 3500 m/z mass range and were processed manually using Data Explorer software (Version 4.0, Applied Biosystems, Forster City, CA, USA). Peptide masses corresponding to established contaminants such as trypsin and keratin were removed from the peak mass lists. Tryptic digests were also analyzed by 1-D reverse phase-HPLC ESI-MS/MS. Peptide separation was achieved using a Dionex UltiMate 3000 Nano HPLC system (LC Packings, The Netherlands) fitted with an Acclaim PepMapTM C18 nanocolumn (75 μm i.d. 15 cm,

Secreted Protein Preparation

Proteins were precipitated overnight at 4 °C using trichloroacetic acid (10% w/v) and pelleted by centrifugation at 6,000 g for 60 min. The protein pellet was washed three times with icecold 90% ethanol and air-dried at room temperature. For 2-DE, 132

dx.doi.org/10.1021/pr200864j |J. Proteome Res. 2012, 11, 131–142


ARTICLE

100 Å pore size). The mobile phase gradient was constructed from 0.1% formic acid (Solvent A) and 0.1% formic acid in 80% ACN (Solvent B): 0% to 15% linear gradient Solvent B over 25 min; 15% to 35% linear gradient Solvent B 25 to 67 min; 35% to 100% linear gradient Buffer B 67 to 77 min. Online peptide identification was achieved using a Finnigan LTQ Linear Iontrap (Thermo Scientific, Waltham, MA, USA) operating in automated Data Dependent mode and using dynamic exclusion. MS/ MS data files were processed using BioWorks 3.3.1 (Thermo Scientific, Waltham, MA, USA). MALDI-TOF peak mass lists and MS/MS data files were searched in-house using MASCOT (Ver. 2.2.03) (www.matrixscience.com) against the functionally annotated B316T genome sequence database (the complete nucleotide sequence of B316T chromosomes BPc1 and BPc2 and megaplasmids pCY360 and pCY186 have been deposited in Genbank under Accession Numbers CP001810, CP001811, CP001812, and CP001813, respectively),16 allowing for one missed tryptic cleavage site, carbamidomethylation of cysteine residues as a fixed modification, methionine oxidation as a variable modification, maximum 50 ppm peptide mass tolerance for peptide mass fingerprints, and maximum 2 Da peptide mass tolerance, maximum 0.5 Da MS/MS mass tolerance, and 2+ and 3+ monoisotopic peptide charge for MS/MS data. PMF and MS/ MS protein identifications were manually examined to ensure peptide masses corresponding to known contaminants were not falsely matched to the protein sequence. Bioinformatics

Signal peptides were predicted by SignalP 3.0 (www.cbs.dtu. dk/services/SignalP/) (cutoff score p > 0.5), LipoP 1.0 (http:// www.cbs.dtu.dk/services/LipoP/) and pattern searching as described by Sutcliffe and Harrington for Gram-positive bacteria.28 30 Membrane-spanning domains were predicted using the TMHMM 2.0 (www.cbs.dtu.dk/services/TMHMM-2.0/) and SOSUI/G (http://bp.nuap.nagoya-u.ac.jp/sosui/sosuiG/sosuigsubmit.html) utilities.31,32 Proteins containing at least one membrane-spanning domain distinct from the N-terminal signal peptide were defined as integral membrane proteins. Protein isoelectric point and molecular weight values were calculated using Emboss iep (http://emboss.sourceforge.net/index.html) and Protein Molecular Weight (http://www.bioinformatics.org/sms2/protein_mw.html), respectively. Functional domains were identified using Pfam, Tigrfam, and BLASTp analysis.33 35 Protein sequence alignments were performed using ClustalW.36 The codon adaptation index was calculated as follows: A codon usage table for 40 highly expressed B316T reference genes (translation elongation factors tufA, tsf, fusA, and 37 ribosomal protein encoding genes rplA-rplF, rplI-rplT, rpsB-rpsT, as used by Sharp et al.,37 for comparisons among diverse species of bacteria) was created by Emboss-CUSP (http://bioweb.pasteur.fr/docs/EMBOSS/cusp.html), which was then used to calculate a CAI value for each predicted secreted protein by Emboss-CAI (http://emboss.sourceforge.net/apps/ release/5.0/emboss/apps/cai.html).

Figure 1. Theoretical 2-DE (A) and protein function summary (B) of the 337 predicted B316T secreted proteins. Solid and open symbols in (A) represent predicted secreted polysaccharide-degrading enzymes and all other secreted proteins, respectively.

showed a strong skew toward the acidic pI range (Figure 1A), where 294 proteins (87%) had a pI value between 3 and 5.6. In silico protein function prediction showed that more than 25% of the secretome is devoted to lignocellulose degradation and the uptake of released soluble sugars (Figure 1B). The B316T secretome included 38 polysaccharide-degrading enzymes that contained catalytic domains collectively representing 14 GH, four CE, and three PL families. All but three of the secreted polysaccharidedegrading enzymes had a theoretical pI value below 5.5. (Figure 1A). More than 16% of the secretome were components of one or more ABC transport systems and included 34 sugar-binding proteins (SBPs), some of which are likely to mediate the uptake of lignocellulose-derived soluble sugars. Strikingly, all 34 SBPs had a theoretical pI value below 4.7. Almost 5% of the secretome was involved in protein processing, but surprisingly for a highly proteolytic bacterium, only one extracellular protease and one peptidase were predicted. More than half the B316T secretome consisted of hypothetical proteins or proteins of unknown function. Relative protein abundance estimation using the codon adaptation index demonstrated that 20% of the secretome were the products of predicted highly expressed (PHX) genes, and almost half of these were either polysaccharide-degrading enzymes or ABC transporter SBPs. The PHX polysaccharidases included enzymes targeting the xylan backbone, xylooligomers, and backbone substituents, as well as cellulose, starch, and pectin. Secreted proteins that were the products of predicted non-highly expressed genes were dominated by hypothetical proteins, proteins of unknown function, and proteins involved in cell envelope biogenesis.

’ RESULTS Theoretical 2-DE

An N-terminal secretory signal peptide was detected in 537 B316T proteins, 200 of which were predicted to be transmembrane proteins due to the presence of at least one membranespanning domain distinct from the signal peptide. The theoretical pI distribution of the remaining 337 secreted proteins 133



ARTICLE

Figure 2. Narrow-range (pI 3 5.6) 2-DE of B316T secreted proteins harvested at mid-log phase (A and B) and stationary phase (C and D), from cells grown in the presence of 0.1% xylan (A and C) or 0.5% xylose (B and D). Annotations correspond to spot numbers in Tables 1 and 2.

2-DE of Proteins Recovered from the B316T Culture Medium

16 ABC transporter system substrate-binding proteins, the only secreted subtilisin family serine protease, one cell-surface protein, and three hypothetical proteins were identified (Table 2). To examine whether unidentified protein spots that were visible in the 2-DE separations originated from yeast proteins that may have been present in the culture medium prior to use, each unidentified PMF was searched against the eukaryotic division of the NCBI non-redundant database. There were no significant matches to any yeast or other proteins contained in the eukaryotic database. Functional domains characteristic of secreted polysaccharidases or CBPs were identified in all seven gene products (Figure 3). Pme8B and Xsa43J were distinct in that they each contained two different GH domains, while the remainder each possessed a single catalytic module. With the exception of Cel5C, each enzyme possessed Type-I cell wall binding repeat regions (CWBD1, PF01473) at their C-termini. Single copies of Family 2 and Family 9 carbohydrate-binding modules (CBMs) were detected in Cel5C and Xyn10B, respectively, and pairs of CBMs were identified in both CBPs (Bpr_I0736 and Bpr_I1599). Summary of the predicted hydrolytic function demonstrated the potential for this group of enzymes to extensively degrade

The pI 3 10 2-DE spot pattern of proteins recovered from the culture medium of xylan-grown, mid-log phase harvested cells was similar to the theoretical secretome 2-DE map, with approximately 70% of the detectable proteins focused between pI 3.0 and 5.6. Consequently, B316T secreted proteins recovered from the culture medium of mid-log phase (Figure 2A and B) and stationary phase (Figure 2C and D) harvested cells grown in the presence of xylan (Figure 2A and C) and xylose (Figure 2B and D), were separated by 2-DE using pI 3.0 5.6 IPG strips. In each pI 3.0 5.6 gel set an average of 195 individual protein spots were detected, 111 of which were identified as the products of 74 B316T non-redundant open reading frames. Thirty of these gene products contained a Type-I or Type-II N-terminal signal peptide, which included five polysaccharidases and two carbohydrate-binding proteins (CBPs) (Table 1). The five secreted polysaccharide-degrading enzymes were identified by matching at least 13 peptide masses to the full-length protein sequences with a maximum mass error of 50 ppm and achieving a minimum 20% sequence coverage. The B316T genome encodes two secreted CBPs at ORFs Bpr_I0736 and Bpr_I1599, and both were detected with 15% and 16% coverage, respectively. In addition, 134



ARTICLE

Table 1. Predicted Secreted Polysaccharidases and CBPs Identified in the Mid-log and Stationary Phase Harvested B316T Culture Mediuma spot 1

protein Endo-1,4-β-glucanase, Cel5C

locus

EC

Bpr_I1710 3.2.1.4

scoreb

PHX

pI

kDa

peptc cov (%)

Y

1.2 10

9

4.6

61.1 16/54

36

4.6 61.1 13/46 4.3 136.9 24/45

28 20

MLd n/c 17.2 ( 4.8 (0.008) xylan

statd 5.7 ( 1.7 (0.002) 12.1 ( 5.3 (0.008) n/d

2 3

Endo-1,4-β-glucanase, Cel5C Endo-1,4-β-xylanase, Xyn10B

Bpr_I1710 3.2.1.4 Bpr_I0026 3.2.1.8

Y Y

7.4 10 1.2 10

7

4

Endo-1,4-β-xylanase, Xyn10B

Bpr_I0026 3.2.1.8

Y

9.6 10

14

4.3 136.9 24/55

25

xylan

5


Bpr_I0026 3.2.1.8

Y

1.5 10

9

4.3 136.9 20/53

20

xylan

n/d

6

Pectate lyase, Pel1A

Bpr_I2372 4.2.2.2

N

3.0 10

13

4.4 115.5 19/56

28

n/d

xylose

7

Pectin methyl-esterase, Pme8B

Bpr_I2473 3.1.1.11

Y

2.4 10

10

4.2 294.1 24/58

14

n/d

n/c

8

Xylosidase/arabinofuranosidase, Xsa43J Bpr_I2935

-

Y

3.0 10

18

4.2 251.9 31/61

17

xylan

n/c

9

Carbohydrate-binding protein

Bpr_I0736

-

N

1.0 10

5

4.1

57.5

6/20

15

n/c

n/c

10 Carbohydrate-binding protein

Bpr_I1599

-

Y

4.8 10

8

4.1

76.9 11/21

16

n/c

n/c

14

17.5 ( 2.5 (0.000)

a

Secretory signal-peptides were predicted using SignalP (Ver. 3.0) and LipoP (Ver. 1.0) and pattern searching as described by Sutcliffe and Harrington.29 31 b Score value is the statistical expectation that the top ranked protein match is a false positive identification. c Values denote number of matched peptides masses/searched peptides masses. d ML, culture medium harvested at mid-log phase (OD600 = 0.5); stat, culture medium harvested at stationary phase (OD600 = 0.7). Mean fold-change ( SEM calculated from three biological replicate experiments is shown. p-Values are shown in brackets. Xylan/xylose denotes uniquely detected in culture medium harvested from xylan/xylose-grown cells respectively; n/d, not detected in either growth condition; n/c, no protein abundance change between growth conditions.

hemicellulose and the surrounding structural polysaccharides (Table 3). The majority of the 44 predicted cytosolic proteins identified in the B316T culture medium (Supporting Information, Table S1) were involved in carbohydrate utilization, energy metabolism, nucleotide and nucleic acid metabolism, and protein synthesis. Relative abundance analysis of the complete set of identified proteins demonstrated that the 20 most abundant 2-DE protein spots were all secreted proteins. In addition, theoretical protein abundance estimation using the codon adaptation index predicted that 75% of the identified cytosolic proteins were the products of highly expressed genes. The relative abundance of cytosolic proteins was not significantly different between growth substrates. Therefore, although there may have been some leakage of cytosolic proteins into the culture medium due to cell-death, only the relative abundance changes detected for the predicted secreted proteins were due to growth substrate.

4 and 5 and smaller than the position of spot 3 (Figure 2A and C). Examination of the peptide mass fingerprints of spots 3, 4, and 5 (Supporting Information, Figure S2) showed that six peptides detected in spot 3 but not detected in spots 4 or 5 matched to the 110-residue C-terminal region of the Xyn10B (Supporting Information, Figures S3 S5). Furthermore, MS/MS analysis of protein extracted from spot 4 identified 21 peptides that matched to Xyn10B but also failed to detect peptides matching to the C-terminal region (Supporting Information, Figure S6). These results suggest that spot 3 is the full-length form of Xyn10B and spots 4 and 5 are C-terminal truncated protein forms. During mid-log phase, the 252 kDa xylosidase/arabinofuranosidase Xsa43J was also detected in the secreted proteins recovered from xylan-grown cells only (Figure 2A, spot 8). In contrast, at stationary phase Xsa43J was detected at equal intensity in the culture medium harvested from cells grown on both substrates (comparison of Figure 2C and D). The 2-DE position of Xsa43J corresponded well with the theoretical pI and size of the enzyme, and peptides were matched to the N- and C-terminal regions of the mature protein sequence (Supporting Information, Figure S7), which together indicate the excised spot represented the fulllength protein species. Three polysaccharides with predicted hydrolytic activities toward cellulose or pectin were also identified. Cel5C is a 62 kDa endocellulase containing a GH5 endoglucanase catalytic domain that is 99% homologous to that of B. fibrisolvens H17c(SA) end1 and a C-terminal CBM2 domain (Figure 3). At both stages of growth Cel5C was detected in two protein spots located at approximately 60 and 52 kDa (Figure 2A D, spots 1 and 2, respectively). The higher mass protein spot was detected at equal abundance during mid-log phase growth on xylan or xylose but during stationary phase growth was significantly less abundant in the xylan-grown cells (Table 1). Furthermore, at both growth stages the lower mass form was more than 12-fold less abundant in xylan-grown cells. Peptide sequence coverage of the two Cel5C protein forms confirmed the presence of several contiguous peptides spanning the catalytic domain in both cases (Supporting Information, Figure S8). One peptide derived from the higher mass protein spot matched to the N-terminal end of the CBM2 domain but was not detected in the lower mass

Abundance Patterns of Secreted Polysaccharidases and CBPs

Xyn10B was the most abundant polysaccharide-degrading enzyme detected in the B316T secretome. Xyn10B was identified in the 2-DE separations of proteins extracted from xylan-grown, mid-log phase harvested cells as three distinct protein spots that together comprised 5.6% of the total detectable protein (Figure 2A). The most abundant of these contributed more than 80% of the total Xyn10B abundance at mid-log phase (spot 4). During stationary phase, the enzyme was identified in a single highly abundant spot at the same 2-DE position as spot 4 detected in the mid-log phase separations and represented 8.7% of the total protein (Figure 2C). In contrast, Xyn10B was undetectable at mid-log phase when cells were grown in the presence of xylose as a sole carbon source (Figure 2B) and was only weakly detectable during stationary phase growth (Figure 2D). Differential expression analysis showed Xyn10B was 17.5-fold more abundant in the secreted proteome of xylan-grown cells during stationary phase growth compared to growth on xylose (Table 1). The xyn10B gene is predicted to encode a 136.9 kDa mature protein, which is similar to the observed 2-DE positions of spots 135



ARTICLE

Table 2. Predicted Non-polysaccharide-degrading Secreted Proteins Identified in the Mid-log and Stationary Phase Harvested B316T Culture Medium spot

protein

locus

funca PHX

scoreb

SigP pI kDa peptc cov (%)

MLd

statd

n/c

n/c

11

3-Hydroxybutyryl-CoA

Bpr_I2486

E

Y

1.2 10

8

SpI 5.4 31.5 11/47

12

dehydrogenase, Hbd ABC transporter SBP

Bpr_I1600

T

N

4.8 10

5

SpII 4.2 93.5 13/46

16

n/c

n/c

13

Amino acid ABC transporter SBP

Bpr_I2466

T

Y

4.9 10

6

SpII 4.1 32.4 6/15

29

xylose

xylose

14

Bmp family protein

Bpr_I1560

U

Y

1782

SpI 3.9 40.7

4

10

n/c

n/c

Sugar ABC transporter SBP

Bpr_I1667

T

Y

1094

ISpI 4.1 47.8

5

14

15

Cell surface protein

Bpr_I2508

I

N

1.1 10

6

SpI 4.6 141.4 12/31

13

xylose

n/c

16

Hypothetical protein

Bpr_I0139

H

N

3.0 10

8

SpI 4.7 26.7 8/34

33

n/c

n/c

Hypothetical protein

Bpr_I0188

H

N

3.9 10

5

SpI 5.1 27.0 9/34

39

14

SpI 7.8 31.8 12/34 SpII 4.3 83.3 10

39 15

7

12 11

Hypothetical protein Bpr_I2628 Oligopeptide ABC transporter SBP, OppA1 Bpr_I1276

H T

Y Y

1.5 10 3195

Peptide/nickel ABC transporter

Bpr_I2750

T

Y

1303

SpII 4.1 58.2


Bpr_I0313

T

Y

5047

SpII 4.0 55.3

9


Bpr_I1720

T

Y

3195

SpII 4.2 50.0

10

19

Oligopeptide ABC transporter SBP, OppA1 Bpr_I1276

T

Y

7.6 10

20

Oligopeptide ABC transporter SBP, OppA1 Bpr_I1276

T

Y

8794

21

Oligopeptide ABC transporter SBP, OppA2 Bpr_III023 Oligopeptide ABC transporter SBP, OppA1 Bpr_I1276

T T

Y Y

1252 1.5 10

27

22

Serine protease subtilisin family

P

Y

6.1 10

8

17 18

41

8.2 ( 2.8 (0.004) xylan

n/c xylan

periplasmic protein

23

Bpr_I2629

20

15

SpII 4.3 83.3 24/40

36

n/c

n/c

SpII 4.3 83.3

17

n/c

n/c

12

SpII 4.3 77.5 7 SpII 4.3 83.3 28/38

10 42

n/c

n/c

SpI 4.0 153.2 26/89

24

n/c

n/c

n/c

3.7 ( 1.7 (0.034)


Bpr_I0313

T

Y

4180

SpII 4.0 55.3

9

10


Bpr_I2010

T

Y

2395

SpII 4.0 47.6

6

29

SpII 4.1 52.4

5


Bpr_I2344

T

Y

1152

24


Bpr_I0182

T

Y

3.8 10

10

SpII 4.4 63.5 15/54

39

14 Xylan

Xylan

25


Bpr_I0182

T

Y

1.2 10

9

SpII 4.4 63.5 14/47

35

7.6 ( 3.0 (0.010)

5.3 ( 0.2 (0.000)

26

Sugar ABC transporter SBP Sugar ABC transporter SBP

Bpr_I0237 Bpr_I0313

T T

Y Y

1087 1682

SpII 4.1 65.1 SpII 4.0 55.3

5 5

10 5

n/c

n/c

SpII 4.2 61.2

10


Bpr_I1589

T

Y

2260

27


Bpr_I0937

T

Y

1.0 10

5

SpII 4.4 61.5 12/41

24

18 n/c

n/c

28


Bpr_I0937

T

Y

3.8 10

9

SpII 4.4 61.5 16/53

32

n/c

n/c

29


Bpr_I1720

T

Y

4737

SpII 4.2 50.0

9

20

n/c

n/c

30


Bpr_I1720

T

Y

3706

SpII 4.2 50.0

7

18

n/c

31


Bpr_III244

T

Y

2.4 10

10

SpII 4.3 35.3 11/34

48

4.5 ( 0.6 (0.005)

3.6 ( 1.5 (0.010)

32

Xylose ABC transporter SBP

Bpr_I1173

T

Y

3.0 10

8

SpII 4.4 38.6 10/34

43

4.8 ( 1.1 (0.009)

3.3 ( 1.7 (0.010)

n/c

a

E, energy metabolism; H, hypothetical; I, cell envelope biogenesis; P, protein fate; T, transporters; U, unknown function. b Statistical expectation values (e.g., 1.2 10 8) are given for MALDI-TOF identifications. MOWSE scores (four digit integers) are given for LC MS/MS identifications. c Values denote number of matched peptides masses/searched peptides masses. d ML, culture medium harvested at mid-log phase (OD600 = 0.5); stat, culture medium harvested at stationary phase (OD600 = 0.7). Mean fold-change ( SEM is shown. p-Values are shown in brackets. Xylan/xylose denotes uniquely detected in culture medium harvested from xylan/xylose-grown cells. respectively; n/d, not detected in either growth condition; n/c, no protein abundance change between growth conditions.

protein spot. Nonetheless, as no tryptic peptides were matched to either the N- or C-terminus, it is difficult to predict where a protein truncation might occur in order to account for the presence of the lower molecular weight spot. During stationary phase only, two pectinolytic enzymes were detected at low abundance. The 294 kDa pectin methylesterase/ pectate lyase Pme8B appeared to be constitutively expressed (Figure 2C and D, spot 7) as the full-length form of the enzyme (Supporting Information, Figure S9), while the pectate lyase Pel1A (EC4.2.2.2) was detected only in xylose-grown, stationary phase cells (Figure 2D, spot 6). The 2-DE spot position of Pel1A was approximately 45 kDa smaller than the predicted full-length size. PMF sequence coverage showed that all 20 peptide masses contributing to the MASCOT identification matched to the

N-terminal half of the protein sequence (Supporting Information, Figure S10), which suggests the 2-DE detected protein was an N-terminal fragment. The B316T genome encodes two secreted CBPs that lack any identifiable catalytic domains. Both were identified in the secreted proteome of xylan- and xylose-grown cells at equivalent levels. Protein spots 9 and 10 (Figure 2A D) were matched to the ORFs Bpr_I0736 and Bpr_I1599, respectively, and the abundance of both proteins was unaffected by growth substrate. Abundance Patterns of Non-carbohydrate-active Secreted Proteins

Several proteins with predicted functions other than polysaccharide degradation exhibited differential abundance patterns 136



ARTICLE

Figure 3. Schematic view of the functional domains within the secreted polysaccharidases and CBPs identified in the B316T culture medium. SP, signal peptide; GH5, GH Family 5 (PF00150); GH10, GH Family 10 (PF00331); GH43, GH Family 43 (PF04616); CE8, CE Family 8 (PF01095); PL1, PL Family 1 (PF00544); PL9, PL Family 9 (IPR011050); CWB, Type-I cell wall binding domain (PF01473); Big4, bacterial Ig-like domain-group 4 (PF07532); SBD, uncharacterized sugar-binding domain (PF07554); CBM2, CBM6, and CBM9, Family 2 (PF00553), Family 6 (PF03422), and Family 9 (IPR010502) carbohydrate binding modules, respectively. Predicted GH domain active sites are shown where possible. Residue length is indicated to the right of each enzyme.

Table 3. Predicted Catalytic Function of the Polysaccharidases Identified in the B316T Secretome protein Endo-1,4-β-glucanase, Cel5C

reaction catalyzeda

locus

substrate

Bpr_I1710

cellulose, lichenin and cereal β-D-glucans

endohydrolysis of (1f4)-β-D-glucosidic linkages; will also hydrolyze (1f4)-linkages in β-D-glucans containing (1f3)-linkages


Bpr_I0026

xylan

endohydrolysis of (1f4)-β-D-xylosidic linkages

Pectate lyase, Pel1A

Bpr_I2372

pectate

eliminative cleavage of (1f4)-α-D-galacturonan to give oligosaccharides with 4-deoxy-α-D-galact-4-enuronosyl groups at their nonreducing ends; favors pectate

Pectin methylesterase, Pme8B

Bpr_I2473

over the methyl ester pectin hydrolysis of pectin yielding pectate and methanol/cleavage

pectin

of (1f4)-α-D-galacturonan to give oligosaccharides with 4-deoxy-α-D-galact-4-enuronosyl groups at their nonreducing ends Xylosidase/arabinofuranosidase, Xsa43J

Bpr_I2935

α-L-arabinofuranose containing xylans

hydrolysis of (1f4)-β-D-linkages, releasing D-xylose residues from the nonreducing termini/hydrolysis of terminal nonreducing α-L-arabinofuranoside residues; also hydrolyzes xylobiose, arabinoxylans, and arabinogalactans containing (1f3) and/or (1f5)-linkages

a

Swiss Institute of Bioinformatics Enzyme nomenclature (http://us.expasy.org/enzyme/).

between the culture supernatants of xylan- and xylose-grown cells. An ABC transporter substrate-binding protein (Bpr_I0182) was the most abundant protein present in both the mid-log and stationary phase harvested xylan-grown culture medium where it constituted 8.1% and 17% of the total detectable protein, respectively (Figure 2A and C, spot 25). At mid-log and

stationary phase growth, it was 7.2- and 5.4-fold more abundant in xylan-grown cells, respectively. Bpr_I0182 is most similar to a Family 1, extracellular solute-binding protein (SBPbac1) secreted by Geobacillus sp. Y412MC10 (Supporting Information, Table S2) and also shows similarity to several other SBPbac1 proteins produced by Gram-positive bacteria. Protein spot 24 also 137



ARTICLE

oat-spelt xylan by B. fibrisolvens H17c led to the accumulation of an extracellular pool of soluble low molecular weight xylooligosaccharides. The relative abundance pattern of Xyn10B in response to growth on xylan and xylose suggests that it is an important xylanolytic enzyme in the extracellular polysaccharide-degrading system of B316T. The GH10 and CBM9 catalytic domains in the B316T Xyn10B display homology to corresponding domains in XynA of Eubacterium ruminantium (Q47871_9FIRM, 42% and 57% BLASTp identity, respectively) and in the XynA precursor of Thermoanaerobacterium thermosulfurigenes (Q60046_THETU, 32% and 32% BLASTp identity, respectively) (Supporting Information, Figure S11). The E. ruminantium XynA exhibited high endoxylanase activity against oat-spelt xylan and xylooligosaccharides when expressed in E. coli, and the primary degradation products were xylobiose and xylotriose.39 On the basis of the BLASTp identity between the B316T Xyn10B catalytic domains and domains within experimentally characterized enzymes produced by other microbes, it is likely that Xyn10B is an endo-1,4β-xylanase capable of liberating variable length xylooligomers from hemicellulose (Figure 4A). The constitutively expressed xylosidase Xsa43J contains a C-terminal GH43 catalytic domain and an N-terminal GH30 domain. GH43 Family enzymes include β-xylosidases, α-Larabinofuranosidases, and endoxylanases and are necessary for the efficient degradation of hemicellulose. In particular, α-Larabinofuranosidases are important for their ability to cleave arabinose side chains that participate in inter- and intrapolymer cross-linking within the plant cell wall.9

contained the product of ORF Bpr_I0182 and was detected only in xylan-grown cells harvested at both mid-log and stationary phases of growth. Protein spot 23 was more abundant in xylangrown cells during stationary phase and contained three ABC transporter SBPs, two of which (Bpr_I0313 and Bpr_I2344) also contained SBPbac1 domains (Table 2). Despite this, the dissimilar BLASTp homology matches for each SBPbac1 domaincontaining protein (Supporting Information, Table S2) suggests they each have different substrate-binding specificities. Several SPBbac1 sugar binding proteins were significantly more abundant in xylose-grown cells at both mid-log and stationary phases of growth. Consistent with the culture conditions used, protein extracted from spots 31 and 32 (Figure 2B and D) was identified as ABC transporter sugar binding proteins specific for simple sugars (RbsB) (Bpr_III244) and xylose (Bpr_I1173) respectively (Table 2). Low homology between their substrate-binding domains (5.0 10 6) indicates different substrate specificities. During stationary phase, B316T cells grown in the presence of xylose also secreted an abundant 32 kDa amino acid ABC transporter substrate-binding protein (Bpr_I2466) that was undetectable in the secreted protein recovered from cells grown on xylan (Figure 2D, spot 13). This protein was similar to polar amino acid substrate-binding proteins produced by clostridial species, and may be involved in glutamine transport. The single subtilisin-family serine protease (EC3.4.1.62) encoded by the B316T genome was identified from protein spot 22 (Figure 2 and Table 2). It was 33% identical to a peptidase synthesized by Thermoanaerobacter sp. X514 (Supporting Information, Table S2). No detectable peptide masses were matched to the first 250 N-terminal residues, which contains 12 potential tryptic peptides larger than 800 Da, indicating a portion of the N-terminal region may be missing from the protein. The protease was detected at a slightly greater abundance in xylose-grown cells at both time points and was considerably more abundant during stationary phase growth.

Potential Role of Non-xylanolytic Secreted Enzymes

An endoglucanase, a pectin methylesterase/pectate lyase, and a pectate lyase were also identified in the B316T culture medium. The catalytic domain of Cel5C is homologous to that of B. fibrisolvens H17c(SA) end1, which displayed high endoglucanase activity when expressed in E. coli.40 Cel5C was detected at lower abundance in the culture medium of xylan-grown cells during mid-log and stationary phase growth, which suggests that the products of xylan breakdown repress cel5C expression. Pectin in forage plant cell walls restricts the access of xylanases and cellulases to their substrates and hinders lignocellulose degradation.41 It is likely that removal of esterified methyl groups from pectin by Pme8B increases the susceptibility of the pectin backbone to degradation by itself and by Pel1A and improves the access of other polysaccharidases to their substrates (Figure 4B). Although we were unable to detect PelA in the culture medium of xylan-grown cells during stationary phase, the low abundance in xylose-grown cells means we cannot discount the possibility that B316T PelA is constitutively expressed, as shown for the C. cellulovorans secreted PelA.42 In view of the fact that B316T is incapable of growth on either cellulose or galacturonic acid alone in vitro,22 the role of the Cel5C endoglucanase and the Pme8B and PelA pectinases could be to assist better access of the xylanolytic enzymes to their substrates to enhance the overall rate and extent of plant cell wall degradation.43 The role of the two carbohydrate-binding proteins (Bpr_I0736 and Bpr_1599) (Table 1) in the secreted proteome is unclear. Both proteins appear to be constitutively expressed and do not contain identifiable cell wall binding domains, implying they are not cell-associated and do not mediate bacterial cell attachment to plant polysaccharides. Both proteins have tandem CBM2a domains similar to those present in Cel5C (Figure 3), and

’ DISCUSSION Xylanolytic Activity in the B316T Secretome

This study verifies that B316T secretes an assortment of polysaccharide-degrading enzymes that are likely to target hemicellulose, cellulose, and pectin. B316T clearly modulates the secretion of these polysaccharidases, as well as a range of substrate-binding components of ATP-dependent soluble sugar uptake systems, in response to the extracellular growth substrates encountered. On the basis of the examination of the genome sequence, Kelly et al.16 hypothesized a mechanism of extracellular polysaccharide breakdown by B316T, where a group of nine large cell-associated proteins that are able to degrade xylan, pectin, and starch form the core of the extracellular catalytic potential.16 The current work provides the first evidence in support of this hypothesis, and a proposed model of hemicellulose degradation and carbohydrate transport by B316T based on our proteomic results is presented in Figure 4. The predicted function of the secreted enzymes and transporter associated proteins discovered in this study implies that B316T attacks primarily the xylan backbone of hemicellulose and actively transports substituted or unsubstituted xylooligosaccharides into the cell, where it will perform the final stages of degradation. This scenario is supported by the observations of Cotta and Zeltwagner,38 who demonstrated that the rapid hydrolysis of 138



ARTICLE

Figure 4. Diagram of extracellular plant polysaccharide metabolism by B316T. Simplified structure of xylan and cellulose (A) and pectin (B) showing the predicted sites of hydrolysis by the identified B316T secreted polysaccharide-degrading enzymes. Dashed lines denote intra- and interpolymer hydrogen bonds. (C) Schematic representation of the ATP-driven transport systems likely to mediate the uptake of hemicellulose-derived soluble sugars. Ac, O-acetyl group; Af, α-L-arabinofuranose; Fe, ferulic acid; G, glucose; GalA, D-galacturonic acid; Gu, 4-O-methyl-D-glucuronic acid; X, xylopyranose (xylose).

Bpr_I1599 also contains tandem CBM6 domains (PF03422) located toward the N-terminus. The fact that neither CBM contains an identifiable catalytic domain suggests a nonhydrolytic function for these proteins that may facilitate subsequent enzymatic attack. The CBM2 of the C. fimi endoglucanase A, for example, was shown to disrupt the structure of cellulose fibers

causing the release of small particles, despite having no detectable fibrolytic activity.44 B316T was originally isolated in a screen for proteolytic rumen bacteria17 and was found to possess high cell-associated serinetype proteolytic activity,22 which correlates well with the protein abundance pattern of the subtilisin family serine protease seen in 139


Journal of Proteome Research this study. An intriguing role for extracellular proteases in plant polysaccharide degradation has recently been proposed by Eun et al.,45 who observed a significantly increased in vivo fiber digestibility in the rumen after treatment of ruminant feed with a commercially prepared protease cocktail. The mixture contained high subtilisin-like serine protease activity but negligible fibrolytic activity and accelerated the rumen fluid induced disappearance of alfalfa hay hemicellulose.46 Cell wall structural proteins are implicated in mediating polysaccharide/lignin cross links,47 and it was hypothesized that cell wall proteolysis allowed rumen fibrolytic microbes better access to plant polysaccharides. The secreted B316T serine protease may play a similar role in targeting cell wall structural proteins and act synergistically with the secreted polysaccharidases to enhance the rate and extent of polysaccharide degradation in vivo. ABC-Mediated Transport of Lignocellulose-Derived Soluble Sugars

The ABC transporter SBPs detected in this study are comparable to those identified in proteomic analyses of other highly fibrolytic Gram-positive bacteria,48 which signifies the importance that ATP-driven active transport of sugars plays in B316T growth. Furthermore, the expression of several of these proteins is regulated depending on the growth substrate encountered, which is likely to provide an advantage within the complex rumen microbial ecosystem allowing B316T to utilize the products of its polysaccharide-degrading activity, rather than having them lost to competing microbes. The differential abundance pattern of the Bpr_I0182 protein product indicates that this SBP may be important for the uptake of soluble lignocellulose-derived sugars. Furthermore, the SBP is clustered in PUL3, which is the largest of the 34 PUL and contains a cytosolic β-xylosidase and a xylulokinase,16 both of which are almost 5-fold more abundant in the cytosol of xylan-grown cells (unpublished data). Together, these findings imply that PUL3 may be particularly important for hemicellulose degradation and utilization by B316T. Enzymes and SBPs May Be Localized to the External Cell Surface in Vivo

A feature of Pme8B, Xsa43J, Xyn10B, and Pel1A is that they all contain multiple CWBD1 modules at the C-terminus. CWBD1s are found in a variety of surface-associated proteins secreted by Lactobacillales and Clostridiales, where they mediate the noncovalent attachment to the peptidoglycan layer.49 It is likely that Pme8B, Xsa43J, Xyn10B, and Pel1A are noncovalently attached to the external surface of the cell wall of B316T in vivo. The presence of these enzymes in the culture media may be explained by the activity of extracellular proteases, as demonstrated in a number of Bacillus strains,48,50 or the ionic composition of the culture medium.51 An α-amylase associated with the external cell surface of B316T (Dunne et al., unpublished data), four unidentified polysaccharide-degrading proteins (lic16A, agn53A, est12B, and Bpr_I0264), and two predicted secreted GH25 lysozymes16 also possess a C-terminal CWBD1 module. Together these results imply that CWBD1 modules may play a major role in the attachment of secreted proteins to the external cell surface of B316T. Each of the 16 SBPs identified in the B316T culture media contained a Type-II N-terminal signal peptide, which suggests they are lipoproteins that are localized to the outer surface of the plasma membrane.52 This scenario is consistent with the fact that in Gram-positive bacteria, most SBPs of ABC transporter systems are lipoproteins,53 and the role that SBPs play in the

ARTICLE

ATP-driven translocation of sugars across the bacterial cell wall.54

’ CONCLUSION Bacterial species belonging to the Butyrivibrio/Pseudobutyrivibrio genera are metabolically versatile rumen bacteria, and their primary role is thought to be the degradation of plant hemicellulose. A clear pattern of the extracellular polysaccharidedegrading system of B316T has emerged during this analysis. The identified secreted polysaccharidases possess one or a combination of GH5, GH10, and GH43, CE8, PL1, and PL9 catalytic activities. A large number of polysaccharidases secreted by prevalent rumen bacteria fall into these enzyme families, which supports the hypothesis that B316T is an important contributor to rumen lignocellulose degradation. It is accepted that the predicted catalytic activity of each of the identified B316T enzymes and transporter-associated proteins, and therefore their likely functional role(s), is drawn from BLASTp sequence homologies to experimentally characterized proteins produced by other organisms. It is also noted that changes in relative protein abundance are not always accompanied by a concomitant change in protein function. Nonetheless, this research has revealed important insights into the extracellular polysaccharide-degrading enzyme system of B316T and provides a framework for the subsequent molecular and biochemical characterization of the secreted proteins. Functional characterization of the B316T fibrolytic enzymes and transporter proteins identified in this study will be vital to advancing our understanding of the role that B316T plays in the conversion of low quality, high fiber forages to premium quality animal products. ’ ASSOCIATED CONTENT

bS

Supporting Information This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Ph: +64 6 4636092. Fax: +64 6 4635331. E-mail: bill.jordan@ vuw.ac.nz.

’ ACKNOWLEDGMENT The authors thank Dr. Lifeng Peng (Centre for Biodiscovery and School of Biological Sciences, VUW) for her invaluable assistance with ESI-MS/MS and data analysis, Dr. Pisana Rawson and Dr. Clifford Young for their helpful technical discussions, Dr. John Koolaard for statistical advice, and Alan McCulloch and Danyl McLauchlan for bioinformatics support. This work was funded by grant C10X0314 from the New Economy Research Fund, administered by the New Zealand Foundation for Research, Science, and Technology. ’ REFERENCES (1) Hungate, R. E. The Rumen and its Microbes; Academic Press Inc: New York, 1966. (2) Ulyatt, M. J.; Beever, D. E.; Thomson, D. J.; Evans, R. T.; Haines, M. J. Measurement of nutrient supply at pasture. Proc. Nutr. Soc. 1980, 39 (3), A67–A67. 140



ARTICLE

assess fiber degradation in complex media. Anal. Biochem. 2006, 349 (2), 297–305. (24) Coutinho, P. M.; Henrissat, B., Carbohydrate-active enzymes: an integrated database approach. In Recent Advances in Carbohydrate Bioengineering; Gilbert, H. J.; Davies, G.; Henrissat, B.; Svensson, B., Eds. The Royal Society of Chemistry: Cambridge, 1999; pp 3 12. (25) Bjursell, M. K.; Martens, E. C.; Gordon, J. I. Functional genomic and metabolic studies of the adaptations of a prominent adult human gut symbiont, Bacteroides thetaiotaomicron, to the suckling period. J. Biol. Chem. 2006, 281 (47), 36269–36279. (26) Aitchison, J. The statistical-analysis of compositional data. J. R. Stat. Soc., Ser. B 1982, 44 (2), 139–177. (27) Beddek, A. J.; Rawson, P.; Peng, L.; Snell, R.; Lehnert, K.; Ward, H. E.; Jordan, T. W. Profiling the metabolic proteome of bovine mammary tissue. Proteomics 2008, 8 (7), 1502–1515. (28) Juncker, A. S.; Willenbrock, H.; Von Heijne, G.; Brunak, S.; Nielsen, H.; Krogh, A. Prediction of lipoprotein signal peptides in Gramnegative bacteria. Protein Sci. 2003, 12 (8), 1652–1662. (29) Bendtsen, J. D.; Nielsen, H.; von Heijne, G.; Brunak, S. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 2004, 340 (4), 783–795. (30) Sutcliffe, I. C.; Harrington, D. J. Pattern searches for the identification of putative lipoprotein genes in Gram-positive bacterial genomes. Microbiology 2002, 148 (7), 2065–2077. (31) Hirokawa, T.; Boon-Chieng, S.; Mitaku, S. SOSUI: classification and secondary structure prediction system for membrane proteins. Bioinformatics 1998, 14 (4), 378–379. (32) Krogh, A.; Larsson, B.; von Heijne, G.; Sonnhammer, E. L. L. Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes. J. Mol. Biol. 2001, 305 (3), 567–580. (33) Altschul, S. F.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J. Basic Local Alignment Search Tool. J. Mol. Biol. 1990, 215 (3), 403–410. (34) Finn, R. D.; Tate, J.; Mistry, J.; Coggill, P. C.; Sammut, S. J.; Hotz, H. R.; Ceric, G.; Forslund, K.; Eddy, S. R.; Sonnhammer, E. L. L.; Bateman, A. The Pfam protein families database. Nucleic Acids Res. 2008, 36, D281–D288. (35) Haft, D. H.; Selengut, J. D.; White, O. The TIGRFAMs database of protein families. Nucleic Acids Res. 2003, 31 (1), 371–373. (36) Larkin, M. A.; Blackshields, G.; Brown, N. P.; Chenna, R.; McGettigan, P. A.; McWilliam, H.; Valentin, F.; Wallace, I. M.; Wilm, A.; Lopez, R.; Thompson, J. D.; Gibson, T. J.; Higgins, D. G., Clustal W and Clustal X version 2.0. Bioinformatics 2007, 23, (21), 2947-2948. (37) Sharp, P. M.; Bailes, E.; Grocock, R. J.; Peden, J. F.; Sockett, R. E. Variation in the strength of selected codon usage bias among bacteria. Nucleic Acids Res. 2005, 33 (4), 1141–1153. (38) Cotta, M. A.; Zeltwanger, R. L. Degradation and utilization of xylan by the ruminal bacteria Butyrivibrio fibrisolvens and Selenomonas ruminantium. Appl. Environ. Microbiol. 1995, 61 (12), 4396–4402. (39) Taguchi, H.; Koike, S.; Kobayashi, Y.; Cann, I. K. O.; Karita, S. Partial characterization of structure and function of a xylanase gene from the rumen hemicellulolytic bacterium Eubacterium ruminantium. Anim. Sci. J. 2004, 75 (4), 325–332. (40) Berger, E.; Jones, W. A.; Jones, D. T.; Woods, D. R. Cloning and sequencing of an endoglucanase (end1) gene from Butyrivibrio fibrisolvens H17c. Mol. Gen. Genet. 1989, 219 (1 2), 193–198. (41) Benshalom, N. Hindrance of hemicellulose and cellulose hydrolysis by pectic substances. J. Food Sci. 1986, 51 (3), 720–725. (42) Han, S. O.; Cho, H. Y.; Yukawa, H.; Inui, M.; Doi, R. H. Regulation of expression of cellulosomes and noncellulosomal (hemi)cellulolytic enzymes in Clostridium cellulovorans during growth on different carbon sources. J. Bacteriol. 2004, 186 (13), 4218–4227. (43) Yu, P. Q.; McKinnon, J. J.; Maenz, D. D.; Olkowski, A. A.; Racz, V. J.; Christensen, D. A. Enzymic release of reducing sugars from oat hulls by cellulase, as influenced by Aspergillus ferulic acid esterase and Trichoderma xylanase. J. Agric. Food Chem. 2003, 51 (1), 218–223. (44) Din, N.; Gilkes, N. R.; Tekant, B.; Miller, R. C.; Warren, A. J.; Kilburn, D. G. Non-hydrolytic disruption of cellulose fibers by the

(3) Beauchemin, K. A.; Rode, L. M.; Sewalt, V. J. H. Fibrolytic enzymes increase fiber digestibility and growth rate of steers fed dry forages. Can. J. Anim. Sci. 1995, 75 (4), 641–644. (4) Titi, H.; Lubbadeh, W. F. Effect of feeding cellulase enzyme on productive responses of pregnant and lactating ewes and goats. Small Ruminant Res. 2004, 52 (1 2), 137–143. (5) Schingoethe, D. J.; Stegeman, G. A.; Treacher, R. J. Response of lactating dairy cows to a cellulase and xylanase enzyme mixture applied to forages at the time of feeding. J. Dairy Sci. 1999, 82 (5), 996–1003. (6) Kung, L.; Treacher, R. J.; Nauman, G. A.; Smagala, A. M.; Endres, K. M.; Cohen, M. A. The effect of treating forages with fibrolytic enzymes on its nutritive value and lactation performance of dairy cows. J. Dairy Sci. 2000, 83 (1), 115–122. (7) Cruywagen, C. W.; van Zyl, W. H. Effects of a fungal enzyme cocktail treatment of high and low forage diets on lamb growth. Anim. Feed Sci. Technol. 2008, 145 (1 4), 151–158. (8) van Soest, P. J. Nutritional Ecology of the Ruminant, 2 ed.; Cornell University Press: Ithaca, NY, 1994. (9) Grabber, J. H.; Ralph, J.; Hatfield, R. D. Cross-linking of maize walls by ferulate dimerization and incorporation into lignin. J. Agric. Food Chem. 2000, 48 (12), 6106–6113. (10) Nakamura, A.; Furuta, H.; Maeda, H.; Takao, T.; Nagamatsu, Y. Analysis of the molecular construction of xylogalacturonan isolated from soluble soybean polysaccharides. Biosci., Biotechnol., Biochem. 2002, 66 (5), 1155–1158. (11) Kulkarni, N.; Shendye, A.; Rao, M. Molecular and biotechnological aspects of xylanases. FEMS Microbiol. Rev. 1999, 23 (4), 411–456. (12) Bryant, M. P.; Small, N. The anaerobic monotrichous butyric acid-producing curved rod-shaped bacteria of the rumen. J. Bacteriol. 1956, 72 (1), 16–21. (13) van Gylswyk, N. O.; Hippe, H.; Rainey, F. A. Pseudobutyrivibrio ruminis gen nov, sp nov, a butyrate-producing bacterium from the rumen that closely resembles Butyrivibrio fibrisolvens in phenotype. Int. J. Syst. Bacteriol. 1996, 46 (2), 559–563. (14) Krause, D. O.; Denman, S. E.; Mackie, R. I.; Morrison, M.; Rae, A. L.; Attwood, G. T.; McSweeney, C. S. Opportunities to improve fiber degradation in the rumen: microbiology, ecology, and genomics. FEMS Microbiol. Rev. 2003, 27 (5), 663–693. (15) Stewart, C. S.; Flint, H. J.; Bryant, M. P. The rumen bacteria. In The Rumen Microbial Ecosystem; Hobson, P. N., Stewart, C. S., Eds.; Blackie Academic and Professional Publishers: London, 1997; pp 10 72. (16) Kelly, W. J.; Leahy, S.; Altermann, E.; Yeoman, C. J.; Dunne, J. C.; Kong, Z.; Pacheco, D. M.; Li, D.; Noel, S.; Moon, C. D.; Cookson, A.; Attwood, G. T. The glycobiome of the rumen bacterium Butyrivibrio proteoclasticus B316T highlights adaptation to a polysaccharide-rich environment. PLoS One 2010, 8 (9), e11942. (17) Attwood, G. T.; Reilly, K. Identification of proteolytic rumen bacteria isolated from New Zealand cattle. J. Appl. Bacteriol. 1995, 79 (1), 22–29. (18) Moon, C. D.; Pacheco, D. M.; Kelly, W. J.; Leahy, S. C.; Li, D.; Kopecny, J.; Attwood, G. T. Reclassification of Clostridium proteoclasticum as Butyrivibrio proteoclasticus comb. nov., a butyrate producing ruminal bacterium. Int. J. Syst. Evol. Microbiol. 2008, 58 (9), 2041–2045. (19) Reilly, K.; Attwood, G. T. Detection of Clostridium proteoclasticum and closely related strains in the rumen by competitive PCR. Appl. Environ. Microbiol. 1998, 64 (3), 907–913. (20) Paillard, D.; McKain, N.; Rincon, M. T.; Shingfield, K. J.; Givens, D. I.; Wallace, R. J. Quantification of ruminal Clostridium proteoclasticum by real-time PCR using a molecular beacon approach. J. Appl. Microbiol. 2007, 103 (4), 1251–1261. (21) Edwards, J. E.; McEwan, N. R.; Travis, A. J.; Wallace, R. J. 16S rDNA library-based analysis of ruminal bacterial diversity. Antonie Van Leeuwenhoek Int. J. Gen. Mol. Microbiol. 2004, 86 (3), 263–281. (22) Attwood, G. T.; Reilly, K.; Patel, B. K. C. Clostridium proteoclasticum sp nov, a novel proteolytic bacterium from the bovine rumen. Int. J. Syst. Bacteriol. 1996, 46 (3), 753–758. (23) Villas-Boas, S. G.; Noel, S.; Lane, G. A.; Attwood, G.; Cookson, A. Extracellular metabolomics: A metabolic footprinting approach to 141



ARTICLE

binding domain of a bacterial cellulase. Biotechnology 1991, 9 (11), 1096–1099. (45) Eun, J. S.; Beauchemin, K. A. Effects of a proteolytic feed enzyme on intake, digestion, ruminal fermentation, and milk production. J. Dairy Sci. 2005, 88 (6), 2140–2153. (46) Colombatto, D.; Beauchemin, K. A. A protease additive increases fermentation of alfalfa diets by mixed ruminal microorganisms in vitro. J. Anim. Sci. 2008, 87 (3), 1097–1105. (47) Showalter, A. M. Structure and function of plant-cell wall proteins. Plant Cell 1993, 5 (1), 9–23. (48) Voigt, B.; Schweder, T.; Sibbald, M.; Albrecht, D.; Ehrenreich, A.; Bernhardt, J.; Feesche, J.; Maurer, K. H.; Gottschalk, G.; van Dijl, J. M.; Hecker, M. The extracellular proteome of Bacillus licheniformis grown in different media and under different nutrient starvation conditions. Proteomics 2006, 6 (1), 268–281. (49) Fernandez-Tornero, C.; Lopez, R.; Garcia, E.; GimenezGallego, G.; Romero, A. A novel solenoid fold in the cell-wall anchoring domain of the Pneumococcal virulence factor LytA. Nat. Struct. Biol. 2001, 8 (12), 1020–1024. (50) Antelmann, H.; Tjalsma, H.; Voigt, B.; Ohlmeier, S.; Bron, S.; van Dijl, J. M.; Hecker, M. A proteomic view on genome-based signal peptide predictions. Genome Res. 2001, 11 (9), 1484–1502. (51) Yother, J.; White, J. M. Novel surface attachment mechanism of the Streptococcus pneumoniae protein Pspa. J. Bacteriol. 1994, 176 (10), 2976–2985. (52) Braun, V.; Wu, H. C. Lipoproteins, structure, function, biosynthesis and model for protein export. In The Bacterial Cell Wall; Elsevier Science: Amsterdam, 1994; Vol. 27. (53) Sutcliffe, I. C.; Russell, R. R. B. Lipoproteins of Gram-positive bacteria. J. Bacteriol. 1995, 177 (5), 1123–1128. (54) Davidson, A. L.; Dassa, E.; Orelle, C.; Chen, J. Structure, function, and evolution of bacterial ATP-binding cassette systems. Microbiol. Mol. Biol. Rev. 2008, 72 (2), 317–364.

142


Extracellular Polysaccharide-Degrading Proteome of Butyrivibrio

Recommend Documents