Previous Article | Next Article ![]()
Journal of Bacteriology, February 2003, p. 703-713, Vol. 185, No. 3
0021-9193/03/$08.00+0 DOI: 10.1128/JB.185.3.703-713.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Sheila I. McCrae,1 Jennifer C. Martin,1 Vincenzo Aurilia,4 Raphael Lamed,3 Yuval Shoham,5, Edward A. Bayer,2 and Harry J. Flint1*
Gut Microbiology Group, Rowett Research Institute, Aberdeen, United Kingdom,1 Department of Biological Chemistry, The Weizmann Institute of Science, Rehovot,2 Department of Molecular Microbiology and Biotechnology, Tel Aviv University, Ramat Aviv,3 Department of Food Engineering and Biotechnology,5 Institute of Catalysis Science and Technology, TechnionIsrael Institute of Technology, Haifa, Israel,6 Institute of Protein Biochemistry, National Research Council, Naples, Italy4
Received 5 July 2002/ Accepted 30 September 2002
|
|
|---|
110-kDa supernatant protein expressed selectively in cultures grown on xylan. Our findings indicate that in addition to the ScaA dockerin-ScaB cohesin interaction, at least two distinct dockerin-binding specificities are involved in the novel organization of plant cell wall-degrading enzymes in this species and suggest that different scaffoldins and perhaps multiple enzyme complexes may exist in R. flavefaciens. |
|
|---|
In Clostridium thermocellum, which has provided the paradigm for cellulosome organization, the complex is integrated via the binding of dockerin-containing enzyme subunits to nine cohesin domains present in a noncatalytic scaffolding protein (CipA). CipA also carries a cellulose-binding module (CBM) and a C-terminal dockerin that is involved in its attachment to cell wall anchoring proteins (16, 18, 26, 27).
As information on cellulosome structures from other species has increased, however, striking variations on the paradigm have emerged. The C-terminal dockerin is absent from many scaffoldins thus far examined, implying alternative anchoring mechanisms, while the scaffoldin protein of Acetivibrio includes a catalytic family 9 glycoside hydrolase domain. Recent work with R. flavefaciens has identified likely scaffolding and anchoring protein components that carry multiple cohesin domains that diverge very significantly in their sequences from those examined previously. The R. flavefaciens cohesins were considered to constitute a new group, designated type III (11). Furthermore the dockerins identified in R. flavefaciens enzymes also show among themselves a high degree of sequence divergence (1). In view of these findings, the present investigation explores further the organization of plant cell wall-degrading enzyme complexes in R. flavefaciens. In particular, we focused on the interactions between the dockerin of ScaA and different cohesins in ScaB. In addition, we examined the interaction between different enzyme-derived dockerin-containing probes and cohesins present in structural protein components. The investigations reported here firmly establish the role of the ScaA protein in R. flavefaciens as a scaffoldin to which a number of enzyme units can attach. Moreover, our results also demonstrate for the first time that the dockerins of the different enzyme subunits vary in their specificity of attachment, presumably to different types of cohesin-containing scaffoldins. Evidence is thus presented that indicates the participation of different enzyme subunits in the formation of different types of enzyme complexes.
|
|
|---|
Completion of the scaA gene sequence. The ScaA coding sequence was completed by PCR walking from chromosomal DNA (Fig. 1), using the primers listed in Table 1. Size-fractionated R. flavefaciens 17 chromosomal DNA fragments of between 2 and 4 kb from a Sau3AI partial digestion were ligated with BamHI-cut pUC18. Amplification with the M13 forward primer (recognizing the vector sequence) and the reverse primer scacoh5r yielded three bands of 0.6, 0.8, and 1.1 kb. Sequencing revealed cohesin-encoding sequences at both ends of the 0.6- and 1.1-kb fragments, but the 0.8-kb fragment showed a cohesin sequence at one end only. The new primer ScaAcoh7r was designed within the unique region. A further amplification from ligated Sau3AI fragments with M13F and ScaAcoh7r gave a 0.9-kb product, which was sequenced and found to contain the start of the scaA gene. Further PCR amplifications and sequencing confirmed the presence of only three cohesins within the ScaA product, as detailed in Fig. 1.
![]() View larger version (9K): [in a new window] |
FIG. 1. ScaA sequencing strategy and domain architecture. The sequence of ScaA was completed by PCR walking, by sequencing extended products produced with ScaAcoh5r ( 0.8 kb) and then with 7r ( 0.9 kb) (Table 1; also see Materials and Methods). Additional primers and amplifications were used to obtain sequences for both strands and to confirm the relationship of the ScaA N terminus with unique ScaA dockerin, and ScaB sequences and their positions are also indicated. Domains within the ScaA protein are shown as follows: X, N terminal domain of unknown function; 1, 2, and 3, cohesin domains; Doc, dockerin. Solid boxes indicates threonine-rich linkers and striped boxes indicate N-terminal signal peptide.
|
|
View this table: [in a new window] |
TABLE 1. Oligonucleotide primers used in this studya
|
Antibodies against ScaA cohesin 2. A rabbit antibody against the cohesin2 from the protein ScaA was raised following standard procedures. Immune serum was collected and subjected to purification using an Affi-Gel protein A column (Bio-Rad Laboratories Ltd., Hemel Hempstead, United Kingdom). The column was equilibrated with 10 bed volumes of binding buffer (10 mM sodium phosphate buffer, 0.15 M NaCl, pH 8.2). After the sample was applied, the column was washed with 10 bed volumes of binding buffer, and elution of the immunoglobulin fraction was performed using 5 bed volumes of 0.1 M sodium citrate, pH 3.0. The eluant was then concentrated using a protein concentrator tube (Vivaspin 4 [molecular mass cutoff, 5,000 Da]; VivaScience, Lincoln, United Kingdom) and stored at 4°C after the addition of 0.1% sodium azide to avoid microbial growth.
Biotinylation of recombinant proteins. Biotinylation was carried out by incubating purified protein with a 20-fold-greater concentration of N-hydroxysuccinimido biotin (Sigma Chemical Co.) in 100 mM HEPES buffer, pH 8.0, for 1 h at room temperature. The excess of unincorporated biotin was washed out using a concentrator (Vivaspin 4 [molecular mass cutoff, 10,000 Da]; VivaScience). The sample was stored at 4°C with 0.05% (wt/vol) sodium azide. Detection was performed using peroxidase-conjugated streptavidin and chemiluminescence.
Western blotting and detection of protein-protein interactions. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto a polyvinylidene difluoride (PVDF) transfer membrane (Immobilon-P; Millipore, Bedford, United Kingdom) in a semidry MultiPhor II (Pharmacia Biotech, Little Chalfont, United Kingdom) or Biometra Dry Blot system as described previously (11). Membranes were blocked for 1 h at ambient temperature in blocking buffer composed of T-TBS buffer (50 mM Tris, 150 mM NaCl, 10 mM CaCl2, 2 mM dithiothreitol, 0.1% Tween [pH 7.5]) containing 2% bovine serum albumin. They were then incubated with the recombinant His-tagged protein or with the biotin-labeled protein (approximately 5 µg of protein · ml-1 in blocking buffer for 1 h at room temperature). The membrane was again incubated for 1 h at room temperature with a peroxidase-conjugated antibody [Anti-His(C-term)-HRP; Invitrogen, San Diego, Calif.] at 0.2 µg · ml-1 in the case of His-tagged proteins, or peroxidase-conjugated streptavidin (0.1 µg ml-1; Sigma) in the case of biotin-labeled proteins, in blocking buffer, after which the membrane was washed three times with T-TBS buffer and twice with TBS buffer (the same buffer, but lacking Tween). The bands were visualized using a chemiluminescence approach (Western blotting with SuperSignal Substrate [Pierce, Rockford, Ill.]) following the manufacturer's instructions.
Glycoprotein staining. Proteins were transferred to a PVDF membrane following SDS-PAGE. The membrane was then stained for glycoprotein using a digoxigenin glycan detection kit (Boehringer, Mannheim, Germany) in an enzyme immunoassay following the manufacturer's guidelines. The proteins were labeled on the membrane with digoxigenin to avoid a shift in the molecular weight caused by the attachment of the digoxigenin-spacer to the glycoprotein.
Assay of cellulose and xylan binding. Substrate binding was studied by incubating the protein (approximately 10 µg in 10 µl) with 5 mg of prewetted cellulose (Avicel) or water-insoluble oat spelt xylan and 10 µl of 50 mM sodium phosphate buffer (pH 6.8)-1 mM dithiothreitol either for 1 h at 37°C or for 16 h at 4°C, with shaking. After incubation the substrate was pelleted at 10,000 x g and washed three times with 50 mM sodium phosphate buffer. Any residual bound protein was then eluted from the pellet by heating at 100°C for 5 min with SDS-PAGE loading buffer, and the released material was analyzed by SDS-PAGE (22).
Transmission immunogold labeling EM. R. flavefaciens was grown anaerobically in M2CGS medium for 48 h and then spun down at 2,500 x g for 10 min and resuspend in sterilized phosphate-buffered saline (PBS) three times for washing. The cells were resuspended in PBS containing rabbit anti-ScaA/Coh2 antibody (5 µg · ml-1) and incubated for 1 h at 37°C in a shaker incubator (200 rpm). Cells were centrifuged again and resuspended in PBS three times to wash out the excess of unbound proteins. Gold-labeled goat anti-rabbit antibody (British BioCell International, Cardiff, United Kingdom) was used as a second antibody at 5 µg · ml-1 in PBS for 1 h at 37°C in a shaker incubator. After the second incubation cells were washed with PBS three times to remove the excess of unbound material and then embedded in the fixative solution (4% paraformaldehyde and 1% glutaraldehyde) overnight and further processed to be visualized by transmission electron microscopy (transmission EM) (Tillydrone Electron Microscopy Unit, Department of Zoology, University of Aberdeen, Aberdeen, United Kingdom). A negative control was carried out using cells with the second incubation reagent only.
Phylogenetic analysis.
Phylogenetic trees were generated using the ClustalW program (http://www2.ebi.ac.uk/clustalw/). Dockerin sequences were obtained from the GenBank website (http://www.ncbi.nlm.nih.gov/) or via the carbohydrate-active enzymes server (CAZy website [http://afmb.cnrs-mrs.fr/
pedro/CAZY/db.html]), designed by Coutinho and Henrissat (6, 8). The following enzyme-borne dockerin sequences were used in this work (accession numbers in parentheses): XynB (Z35226), XynD (S61204), XynE (AJ272430), EndA (Z83304), EndB (AJ298117), and CesA (AJ238716) from R. flavefaciens; EgV (AB028320), EgVI (AB028321), and EgVII (AB028321) from R. albus; Xyn1 (Z49970) from Ruminococcus sp.; CbhA (X80993), CelS (L06942), CelD (X04584), CelF (X60545), CelH (M31903), CelP (AJ275974), and XynA (AF047761) from C. thermocellum; CelF (M87018), CelC (M87018), CelG (M87018), CelE (M87018), CelH (AF316823), and CelM (AF316823) from Clostridium cellulolyticum; PelA (AF105330) and ExgS (U34793) from Clostridium cellulovorans; and Clostridium josui CelD (AB004845). Sources for the scaffoldin-borne dockerin sequences in this work were R. flavefaciens ScaA (AJ278969), C. thermocellum CipA (L08665), B. cellulosolvens CipBc (AF224509), and A. cellulolyticus CipV (AF155197).
Updated nucleotide sequence accession number. The sequence under accession number AJ278969 has been updated to include the complete sequence of ScaA.
|
|
|---|
G, -11.8 kCal [42] when matched with the Bacillus subtilis consensus). A recombinant form of the second cohesin domain of the ScaA protein (ScaA-Coh2) had previously been purified after cloning into pET28a and overexpression in E. coli (11). In order to detect expression of the ScaA product in R. flavefaciens 17, polyclonal antibodies were raised against the purified cloned ScaA-Coh2 domain, as described in Materials and Methods. This antibody preparation was found to bind specifically to a polypeptide of 130 kDa present among extracellular proteins from R. flavefaciens 17, which was also present among supernatant proteins eluted with SDS after binding to microcrystalline cellulose (Fig. 2). This protein corresponds to an isolated band of 130 kDa in Coomassie-stained gels that showed evidence of glycosylation. EM studies using immunogold labeling revealed that the ScaA protein is distributed both on the cell surface and presumably detached cell surface material (Fig. 3). The extensive labeling of cell surfaces observed in the figure is indicative of most of the cells observed in the sample. In contrast, some of the apparently detached material was heavily labeled whereas some was unlabeled. At least a portion of the former fraction would represent tangentially sectioned cell surfaces.
![]() View larger version (80K): [in a new window] |
FIG. 2. Detection of native ScaA with antibodies raised against purified, recombinant ScaA-Coh2. Proteins from supernatants of cellobiose-grown R. flavefaciens 17 cultures were separated by SDS-PAGE and analyzed by Coomassie blue staining (lane 1) or glycostaining (lane 2). A duplicate sample was transferred to PVDF membranes and probed with anti-ScaA-Coh2 antibodies (lane 3). Numbers at left are molecular masses in kilodaltons.
|
![]() View larger version (77K): [in a new window] |
FIG. 3. Electron micrograph showing the distribution of ScaA, detected by immunogold labeling, in cellulose-grown R. flavefaciens 17 at late stationary phase.
|
Interactions of different ScaB cohesins with ScaA. We established previously that two adjacent cohesin domains (4 and 5) of the putative anchoring protein ScaB recognize the dockerin of ScaA, and interact specifically with a polypeptide of 130 kDa (now confirmed as ScaA) among native R. flavefaciens 17 proteins (11). The cohesin domains of ScaB show considerable sequence divergence, however, apart from Coh5 and Coh6, which resemble each other closely (11). In order to examine the interactions of the ScaB cohesins in more detail, His-tagged constructs were prepared for six of the seven individual cohesins of ScaB, and the purified products were probed against native R. flavefaciens extracellular proteins. As shown in Fig. 4, all six of the ScaB cohesins, when tested individually, interacted with a band of approximately 130 kDa. Although we were unable to recover stable clones for ScaB-Coh5, the close resemblance of its sequence to that of ScaB-Coh6 (11) and the behavior of the combined ScaB-Coh4-Coh5 construct suggest that it fails to recognize additional proteins. Therefore, it appears that all seven ScaB cohesins recognize predominantly a single protein present in cellulose-grown R. flavefaciens 17. Additional bands smaller than 130 kDa were detectable with longer exposures, suggesting that some binding occurs to additional proteins, but this could be the result of lower affinity binding to enzyme dockerins. Alternatively, multiple-modular proteins like ScaA are known to undergo fragmentation (25, 31, 32, 35), which could account for the observed interaction of dockerin-containing probes with smaller bands.
![]() View larger version (67K): [in a new window] |
FIG. 4. Recognition of a common 130-kDa protein band by different cohesin domains from ScaB. Total R. flavefaciens 17 supernatant proteins (approximately 7 µg) from a cellulose-grown culture were loaded onto successive lanes of an SDS-PAGE gel, separated, and subjected to Western blotting. Membrane strips containing the electrophoretically transferred material were then incubated with each of the designated cohesin probes (standardized by His tag detection, corresponding to approximately 10 to 50 µg of protein). Binding was detected by chemiluminescence using peroxidase-conjugated anti-His tag antibody. Lanes are marked 1 through 7 according to the cohesin probe used. Lane Rf, profile of Coomassie-stained R. flavefaciens supernatant proteins. Numbers at left are molecular masses in kilodaltons.
|
Interaction of ScaA-Coh2 with dockerin-containing enzymes from R. flavefaciens. ScaA-Coh2 was shown previously to interact with a dockerin-containing fragment of the enzyme XynD (11). Owing to the distinctive variation in dockerin sequences between different enzymes from R. flavefaciens 17 (1), it was of interest to assess the ability of other dockerin-containing enzyme subunits from R. flavefaciens to bind to ScaA. Appropriate catalytically active fragments from the enzymes EndB, XynB, XynE, and CesA (1, 15, 38) were all overexpressed and purified as His-tagged products after cloning in pET28a vectors. The interaction of the resultant dockerin-containing fragments to the purified ScaA-Coh2 domain was examined in a series of Western blotting experiments, using biotinylated ScaA-Coh2 as the probe in each case. ScaA-Coh2 was able to recognize the dockerin of EndB, but no binding was detected to XynE or to CesA (Fig. 5). This observation is of interest since, as noted previously (1, 38), the dockerin sequences of XynE and CesA show significant divergence from those of XynB, XynD and EndB, and this divergence includes residues implicated in binding specificity in cellulosomal enzymes from Clostridium spp. (35).
![]() View larger version (101K): [in a new window] |
FIG. 5. Affinity blotting of dockerin-containing constructs from CesA, XynE, and EndB using recombinant ScaA-Coh2. Dockerin-containing fragments of the three enzymes and the ScaA-Coh2 cohesin domain, were constructed as six-His-tagged fusions (Table 1), and the products were purified. The enzyme constructs were separated by SDS PAGE, transferred to a PVDF membrane, and probed with biotinylated ScaA-Coh2 (right-hand panel). Coomassie-stained lanes from the same gel are shown in the left-hand panel. Significant fragmentation of the recombinant proteins was observed in all preparations. The estimated molecular masses for the CesA, XynE and EndB constructs are 45,427, 41,625, and 87,040 Da, respectively. EndB fragmented into two major fractions, of which the full-length construct and the smaller ( 50-kDa) fragment were labeled by the cohesin-containing probe. No labeling of XynE or CesA was observed. Numbers at left are molecular weights (MW) in thousands.
|
![]() View larger version (38K): [in a new window] |
FIG. 6. Affinity blotting of cell-associated proteins from R. flavefaciens 17 by dockerin-containing constructs from CesA, XynE, and EndB (a) and XynB (b). Cultures were grown on either microcrystalline cellulose (C) or birch wood xylan (X) as the energy source. Proteins were separated in successive lanes of the same SDS-PAGE gel before transfer to a PVDF membrane. The molecular size of the XynB product used as probe was 44,782 Da. The blots were probed with the indicated biotinylated protein construct, and labeled bands were detected by chemiluminescence using peroxidase-conjugated streptavidin. A band of around 45 kDa (not shown in the figure), representing a native biotinylated protein present in R. flavefaciens cells, was detected in all samples and was used as an internal standard. The EndB-induced labeling of the 130-kDa band derived from both cellulose- and xylan-grown cultures was confirmed by immunochemical staining using a nonbiotinylated construct and anti-His-Tag antibodies (not shown). Numbers at left and right are molecular weights (MW) in thousands.
|
![]() View larger version (100K): [in a new window] |
FIG. 7. Binding of biotinylated CesA, XynE, and EndB probes to extracellular material of R. flavefaciens 17 cultures grown on either cellulose (C) or birchwood xylan (X). Concentrated culture supernatants were subjected to SDS-PAGE (gel). Duplicate samples were transferred to membranes and probed using the designated biotinylated construct. Numbers at left are molecular weights (MW) in thousands.
|
|
|
|---|
![]() View larger version (37K): [in a new window] |
FIG. 8. Schematic representation of dockerin-cohesin interactions involved in cellulosome organization in R. flavefaciens 17. Three different specificities of cohesin-dockerin interaction are shown: between the ScaA dockerin and ScaB cohesins (a), between the ScaA cohesins and enzymes with EndB-type dockerins (b), and between the cohesins of a putative 110-kDa scaffoldin (ScaX) and the CesA and XynE dockerins (c). Although ScaB is a cell-associated protein, the proposed interaction of the C-terminal X domain with the cell surface awaits experimental verification.
|
Phylogenetic analysis of known dockerin sequences derived from enzymes and scaffoldins of various bacteria (Fig. 9) revealed that those of R. flavefaciens EndA, EndB, XynB, and XynD form a relatively tight group which is significantly different from the dockerins of XynE and CesA. All of the R. flavefaciens dockerins are, in fact, markedly distinct from the known enzyme-borne dockerins of the clostridia. The divergent specificities of the R. flavefaciens dockerins observed in this work are further supported by their predicted specificity determinants (Table 2). Earlier work on the interspecies specificities of the clostridial cohesin-dockerin interaction has suggested that four amino acid residues (positions 10 and 11 in the characteristic duplicated segment) of the
70-residue dockerin domain may serve as recognition codes for binding to the cohesin domain (29, 35). In this context, the relevant positions of the EndA-type dockerins are characterized by LA(S)-XD, whereas those of the XynE-type dockerinsSF(L)-VAare entirely different. The latter specificity residues are also consistent with those of Ruminococcus sp. strain Xyn1, which is indeed phylogenetically related to XynE (Fig. 9), and it is thus likely that the former enzyme would also show a similar binding specificity. It is interesting that the predicted specificity residues of the R. flavefaciens dockerins are all distinct from those of other speciesboth for enzyme- and scaffoldin-bearing dockerins. The present results thus appear to support further the role of these residues as specificity determinants.
![]() View larger version (30K): [in a new window] |
FIG. 9. Phylogenetic relationship of R. flavefaciens dockerin domains. Dockerins from the indicated R. flavefaciens enzymes (designated Rumfl-EndA, etc.) are shown as filled circles. Scaffoldin-borne dockerins from R. flavefaciens (Rumfl-ScaA), C. thermocellum (Clotm-CipA), A. cellulolyticus (Acece-CipV), and B. cellulosolvens (Bacce-CipBc) are shown as squares. Other dockerin-borne enzymes are from R. albus (Rumal-EgV, etc.), Ruminococcus sp. (Rumsp-Xyn1), and a selection of enzymes from C. thermocellum and mesophilic clostridia (C. cellulolyticum, C. cellulovorans, and C. josui). For a precise list of the proteins and their accession numbers, consult Materials and Methods. The scale bar indicates the percentage (0.1) of amino acid substitutions.
|
|
View this table: [in a new window] |
TABLE 2. Predicted specificity residues of dockerins derived from the indicated proteins
|
ScaA represents the first case of a cellulosomal scaffolding protein that does not possess an identifiable cellulose-binding module. Nevertheless ScaA corresponds to a 130-kDa native protein that was previously detected in supernatant material that binds to cellulose with a His-tagged EndB probe (38). This suggests that substrate binding, considered to be one of the major functions of the cellulosome, must be mediated through a somewhat-different mechanism in this species. One possibility is that a subunit(s) that binds the scaffolding protein, rather than the scaffolding protein itself, would contain a substrate-binding module(s). Indeed, the EndB cellulase, shown here to bind to ScaA Coh2, was recently found to contain a novel CBM that might play a role in helping the complex attach to cellulose (38). Other possibilities are that additional, as-yet-unidentified, enzymatic or nonenzymatic subunits of the cellulosome may play a role in binding. Alternatively, it has been suggested for R. albus (37) that cellulose-binding pilus-like structures might be involved in delivering the bacterial cell and its enzyme complexes into contact with the substrate.
This work has provided further insights into the complexity and diversity of cellulosome organization within cellulose-digesting microorganisms. Ruminal strains of R. flavefaciens have evolved in an ecosystem that provides a well-regulated environment, but one that is also highly competitive and subject to rapid turnover. The organization of the enzymes that provide this bacterium with its main source of energy presumably reflects its adaptation to survival in this ecosystem.
For a commentary on this article, see page 701 in this issue. ![]()
Present address: National Renewable Energy Laboratory, Biotechnology Center for Fuels and Chemicals, Golden, CO 80401. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»