Correspondence Yasuo Hitsumoto, Department of Life Science, Okayama University of Science, 1-1 Ridai-cho, Kita-ku, Okayama, Japan. Tel: +81 86 256 9692; fax: +81 86 256 9559; email: firstname.lastname@example.org
The Clostridium perfringens strain 13 genome contains two genes (fbpA, fbpB) that encode putative Fbp. Both rFbpA and rFbpB were purified and their reactivity with human serum Fn was analyzed. To determine the region of the Fn molecule recognized by rFbp, a plate binding assay using N-terminal 70-kDa peptide, III1-C peptide, and 110-kDa peptide containing III2–10 of Fn was performed. Both rFbp bound to the III1-C peptide of Fn but not to the other peptides. However, the III1-C fragment of Fn is known to be cryptic in serum Fn. Then, rFbp-BP from Fn were purified by rFbp-affinity chromatography. The yield of purified proteins was approximately 1% of the applied Fn on a protein basis. Western blotting analysis of the rFbp-BP, using four different anti-Fn monoclonal antibodies, revealed that the rFbp-BP carried partial Fn antigenicity. Bindings of rFbp to rFbp-BP were inhibited by the presence of the III1-C peptide, suggesting that rFbp-BP express the III1-C fragment. The binding of Fn to III1-C was inhibited by the presence of either rFbpA or rFbpB. This result that suggests C. perfringens Fbps may inhibit the formation of Fn-matrix in vivo.
C. perfringens, a Gram-positive, sporulating pathogen of humans and animals, causes gas gangrene and food poisoning (1). Following invasion of the host tissue, the bacterium encounters many host components, including Fn. Fn is a 450-kDa dimeric glycoprotein found in plasma, on cell surfaces and in extracellular matrices. The Fn polypeptide comprises a number of repeats, of which there are three kinds of modules, types I, II, and III (2). Fn is known to interact with various extracellular matrix molecules including collagen, fibrin, heparin and gelatin, as well as with membrane proteins of the integrin family (3). Fn is known to be involved in the process of wound-healing and to function in promotion of cell attachment, phagocytosis, and activation of CD4+ T cells and macrophages (4, 5). Many bacteria are thought to utilize Fn for proliferation in host tissue and to escape from their hosts’ defense systems (6).
Indeed, the bacteria Staphylococcus (7–9), Streptococcus (10–13), Listeria (14–16), and Clostridium difficile (17) have been shown to have Fbp. C. perfringens is also thought to have Fbps since Fn has been observed to specifically bind to this bacterium (18). Genomic analysis of C. perfringens strain 13 has revealed that this bacterium contains two genes (CPE0703: fbpA and CPE1847: fbpB) that encode putative Fbp (19). These genes were found to be constitutively expressed in three strains of C. perfringens that were isolated from cases of gas gangrene in humans. Both recombinant proteins expressed from these genes, rFbpA and rFbpB, have been shown to bind to Fn in a ligand blotting assay when rFbp are immobilized on either a PVDF membrane or a plastic microplate (20).
In the present study, the Fn epitope recognized by rFbp was determined. Further, the characteristics of serum Fn which has been bound by rFbp were analyzed.
MATERIALS AND METHODS
Cloning, expression, and preparation of rFbpA and rFbpB
To generate His-tagged rFbpA and rFbpB proteins the C. perfringens strain 13 genes fbpA and fbpB were first amplified by PCR as described previously (20). The resultant DNA fragments were cloned into the pET16-b vector (Merck KGaA, Darmstadt, Germany) and transformed into the E. coli BL21-CodonPlus (DE3) RIL strain. The transformants were grown at 37°C in Luria-Bertani broth (Invitrogen, Carlsbad, CA, USA) containing 100 μg/ml ampicillin and 34 μg/ml chloramphenicol to an optical density of 0.6 at 600 nm. Induction of gene expression was accomplished with 1 mM IPTG for 3 hr at 37°C. After incubation, the cells were harvested, and were lysed in a French press (10 000 pounds per square inch). His-tagged proteins were purified on a Ni2+-Sepharose column.
Preparation of Fn and Fn fragments
Fn was purified from pooled human serum using a gelatin-Sepharose column. Fn was obtained by elution with 4 M urea in 5 mM VBS, pH 7.4. Human Fn proteolytic N-terminal 70-kDa and human Fn proteolytic N-terminal 30-kDa fibrin/heparin binding, human Fn proteolytic 45-kDa gelatin binding and recombinant human III1-C (7 kDa) fragments were purchased from Sigma (St. Louis, MO, USA). The 110-kDa Fn fragment (type III2–10) was obtained by digestion of Fn with thermolysin, followed by gel-filtration on a HiLoad 16/60 Superdex 200 column (GE Healthcare, Little Chalfont, UK) as described by Borsi et al. (21).
The anti-Fn mAbs HB91 and HB39, obtained from their respective mAb-producing hybridomas, were purchased from ATCC (Manassas, VA, USA). The anti-Fn mAbs ZET1 and ZET2 were obtained from hybridomas established by us as follows: SP-2/0 myeloma cells were hybridized with spleen cells from BALB/c mice immunized with Fn (ZET1), an 80-kDa Fn fragment containing Fn type III3–11 (ZET2). Each mAb (IgG1) was purified from the hybridoma culture supernatant using a protein G column.
Plate binding assays
All plate binding assays were carried out by individually coating the wells of an EIA/RIA plate (Corning, NY, USA) with 50 μl protein solution at a concentration of 0.02 mg/ml in 10 mM BB, (pH 8.5), for 30 min at room temperature. The wells were then blocked by incubation for 1 hr at room temperature with 250 μl of 1% (w/v) BSA in BB. Following three washes with 20 mM PBST (pH 7.4), the binding of biotinylated proteins or specific antibodies was tested by addition of 100 μl of a 0.01 mg/ml solution of the proteins or antibodies in BVBS to each well, and the microplate was incubated for 1 hr at room temperature. After washing with PBST, HRPO-streptavidin (1:5000; Vector Laboratories, Burlingame, CA, USA) or HRPO-conjugated goat anti-mouse IgG (1:5000; Biosource, Camarillo, CA, USA) in 10 mM TBS (pH 7.2) was then added and reacted for 30 min at room temperature. After washing with PBST, the wells were subjected to color development by the addition of 0.1 ml of 0.91 mM 2,2′azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid) in 0.1 M citrate (pH 4.1) containing 0.04% (v/v) H2O2. The reaction was stopped by the addition of 0.1 ml of 0.1 M citric acid containing 0.01% (w/v) NaN3. The absorbance at 405 nm was then measured in a microplate reader (SpectraMax 340 C, Molecular Devices, Sunnyvale, CA, USA).
Biotinylation of Fn and rFbp
Fn or rFbp (each at 1 mg/ml) were incubated with 0.1 mM biotinamidohexanoic acid 3-sulfo-N-hydroxysuccinimide ester sodium salt (Sigma) in VBS for 1 hr at room temperature. After incubation, a one-fifth volume of 0.5 M Tris-glycine buffer (pH 7.5) was added and the mixture was then further incubated for 1 hr at room temperature. Unattached biotin was removed using a desalting column (GE Healthcare).
Reactivity of rFbp with Fn fragments
A plate binding assay was carried out by coating the wells with Fn fragments (70 kDa, 30 kDa, 45 kDa, 110 kDa or III1-C) and by assay of the binding of biotinylated rFbpA or biotinylated rFbpB in BVBS containing 0.02% (v/v) Tween 20.
Affinity purification of rFbp-BP
Both rFbpA-Sepharose and rFbpB-Sepharose were prepared by coupling NHS-activated Sepharose (GE Healthcare) with rFbpA and rFbpB respectively, according to the instruction manual. Both rFbpA-Sepharose and rFbpB-Sepharose were applied with 25 mg and 30 mg Fn respectively. Bound proteins were then eluted with 4 M urea in VBS. The resulting eluates were designated as rFbpA-BP and rFbpB-BP, respectively.
Reactivity of anti-Fn mAbs with various Fn fragments
A plate binding assay was carried out by coating the wells with Fn fragments (70 kDa, 30 kDa, 45 kDa, 110 kDa, or III1-C) or with Fn and by assay of binding of the anti-Fn mAbs HB91, HB39, ZET1, or ZET2.
SDS-PAGE and Western blotting
Samples containing rFbpA-BP, rFbpB-BP or Fn were mixed with an equal volume of Laemmli sample buffer. Proteins were separated on a 7% SDS-PAGE gel under non-reducing conditions. The electrophoresed components were then either subjected to silver staining or transferred from the gel to a PVDF membrane (Millipore, Billerico, MA, USA) using a transblot unit (Atto, Tokyo Japan). The transblotted PVDF membrane was blocked with casein blocking buffer (Sigma) for 2 hr at room temperature and then incubated with 20 ml of anti-Fn mAbs (0.01 mg/ml) in VBS containing 10% casein blocking buffer for 1 hr at room temperature. After washing with PBST, the membrane was reacted HRPO-conjugated goat anti-mouse IgG (1:5000) in TBS for 30 min at room temperature. After washing with PBST, the membrane was subjected to color development with 0.25 mg/ml 3,3′-diaminobenzidine (Sigma) in 50 mM Tris-HCl, pH 8.0, containing 0.01% (v/v) H2O2.
Inhibition assay of the binding of rFbp to rFbp-BP by III1-C peptide
The binding of biotinylated rFbp to rFbp-BP in the presence of 1 μg or 5 μg III1-C peptide was measured. The biotinylated rFbp were dissolved in BVBS containing 0.02% (v/v) Tween 20.
Inhibition assay of the binding of Fn to III1-C peptide by rFbp
The binding of biotinylated Fn to III1-C in the presence of 1 μg or 10 μg rFbp was measured.
Determination of protein concentration
Absorbance at 280 nm was used to calculate the protein concentration of Fn and Fn fragments using ɛpercent= 10. The concentration of rFbp was measured by the Bradford method (Bio-Rad, Hercules, CA, USA) using BSA as a standard.
All experiments were performed in triplicate. Statistical significance (P < 0.05, P < 0.01) was determined by comparison with controls using Student's two-tailed t-test.
Reactivity of rFbp with Fn fragments
To determine which Fn fragments are recognized by the rFbp, a plate binding assay was performed in which binding of biotinylated-rFbpA or -rFbpB to immobilized Fn fragments (70 kDa, 30 kDa, 45 kDa, 110 kDa or III1-C) was assayed. The Fn fragments were mapped according to their position within the Fn polypeptide (Fig. 1a). Of the Fn fragments tested, both rFbpA and rFbpB bound only to the III1-C fragment of Fn (Fig. 1b).
Reactivity of anti-Fn mAbs with Fn fragments
Both rFbpA and rFbpB were found to bind to the III1-C fragment. However, the III1-C fragment of serum Fn is known to be cryptic. Therefore, rFbp-binding proteins from Fn were purified by affinity chromatography on rFbpA- and rFbpB-Sepharose columns. Following elution of bound proteins with 4 M urea, the yield of affinity purified binding protein from rFbpA-Sepharose and rFbpB-Sepharose chromatography was 0.96% and 1.08% of the applied Fn protein, respectively. In order to characterize the purified rFbp-BP, epitope mapping with various anti-Fn mAbs using immobilized Fn fragments in a plate binding assay was first carried out. The mAb HB91 reacted strongly with both the N-terminal 70-kDa and 30-kDa fragments of Fn, but reacted weakly with the 45-kDa fragment. The other three mAbs tested, HB39, ZET1, and ZET2, reacted with the 110-kDa Fn fragment. The HB39 mAb was the only mAb that also reacted weakly with both the N-terminal 70-kDa and 30-kDa fragments (Fig. 2a). No mAb tested here reacted with III1-C (Fig. 2b).
Reactivity of anti-Fn mAbs with rFbpA-BP and rFbpB-BP
To determine if the rFbp-BP might contain Fn-epitopes, whether the rFbp-BP were recognized by the anti-Fn mAbs, using SDS-PAGE and Western blotting analysis was checked. Silver staining of SDS-gels showed that both rFbpA-BP and rFbpB-BP consisted of a major, slightly broad protein band with a size of 450 kDa, and minor bands with the sizes of 180, 160 and 84 kDa (Fig. 3a and b). When binding of the anti-Fn mAbs was tested by Western blotting, the 450-kDa protein band of the rFbp-BP reacted with both HB91 and HB39, but not with ZET1 or ZET2.
Inhibition of rFbp binding to rFbp-BP by III1-C peptide
To determine whether rFbp-BP expressed III1-C, a rFbp-binding assay to rFbp-BP in the presence of III1-C peptides was performed. Binding of both rFbpA and rFbpB to rFbpA-BP and rFbpB-BP, respectively, was significantly inhibited by the presence of III1-C peptide in a dose-dependent manner (Fig. 4).
Inhibition of Fn binding to III1-C by rFbp
A Fn-binding assay to III1-C was performed in the presence of rFbp. Binding of biotin-Fn to III1-C was significantly inhibited by the presence of either rFbpA or rFbpB in a dose-dependent manner (Fig. 5).
The present study demonstrates that C. perfringens-derived rFbp (rFbpA and rFbpB) recognize the III1-C fragment of serum Fn. The III1-C fragment of Fn is known to be cryptic in serum Fn and is a site involved in fibril formation of Fn (22). Serum Fn expresses the III1-C fragment only when it binds to a particular cell surface by virtue of specific receptors including integrins (23–25). However, in the present study, affinity chromatography of Fn on rFbp-Sepharose columns yielded a small amount of bound Fn that represented about 1% of the applied Fn protein. Further, the binding of rFbp to rFbp-BP was inhibited by III1-C peptide (Fig. 4). These results suggest that a small proportion of serum Fn expresses the III1-C fragment. The biological significance of the III1-C expressing Fn is, however, unclear as this moment.
HB91 strongly reacted with both the 70-kDa and 30-kDa fragments, indicating that the HB91 epitope is located in the 30-kDa peptide. However, HB91 also reacted with the 45-kDa fragment (Fig. 2a). Because both the 30-kDa and 45-kDa fragments have Type I module repeats, HB91 reactivity with the 45-kDa fragment is thought to represent cross-reactivity towards the Type I module. HB39 strongly reacted with the 110-kDa fragment, while it weakly reacted with both the 30-kDa and 70-kDa fragments (Fig. 2a). Therefore, the HB39 epitope is thought to be located primarily in the 110-kDa peptide. Although the reason for HB39 also reacting with the 30-kDa peptide is unclear, this may be attributable to non-specific reactivity of HB39 between the 110-kDa and 30-kDa peptides. The epitopes recognized by the other mAbs, ZET1 and ZET2, are thought to be located in the 110-kDa peptide.
The 450-kDa protein bands of the rFbp-BP were identified as Fn because they reacted with the two different anti-Fn mAbs, HB91 and HB39, when tested by Western blot. These bands are indistinguishable from intact Fn on the basis of size. However, they were not recognized by the other anti-Fn mAbs, ZET1 or ZET2. Fn isolated from plasma/serum is known to consist of different polypeptides generated by alternative splicing (26, 27). Therefore, rFbp-BP are thought to be splicing variants which may lack or veil the epitopes which are located in the 110-kDa fragment and are recognized by ZET1 and ZET2.
None of the 84-kDa, 160-kDa, and 180-kDa protein bands of either rFbpA-BP or rFbpB-BP reacted with the four different anti-Fn mAbs used here. After storing rFbp-BP for several days at 4°C, the 450-kDa protein bands disappeared while the amount of the 160-kDa and 180-kDa protein bands increased (data not shown). The latter bands reacted with anti-Fn mAbs in a Western blot. Thus, protein bands with a molecular size less than 220 kDa may be Fn fragments which have been degraded from 450-kDa rFbp-BP. The failure of the 84-kDa, 160-kDa, and 180-kDa protein bands to react with anti-Fn mAbs might be due to insufficient doses of protein.
Fbp from other bacteria, FnBPA and FnBPB from Staphylococcus aureus and Sfb1 from S Staphylococcus pyogenes, are known to contain a common motif that bind to the N-terminal type I module of Fn (28, 29). Another Fbp, BBK32 from Borrelia burgdorferi, is reported to bind to III1–3 as well as to I1–5 of Fn (30, 31). BBK32, however, has the capacity to make an aggregation of Fn by virtue of binding to III1–3 of Fn. Unlike BBK32, neither FbpA nor FbpB from C. perfringens has such an Fn aggregating capacity (data not shown). It is known that Fn aggregates when Fn is incubated with III1-C peptide (32). This means that Fn binds to III1-C peptide. In fact, in the present study, Fn reacted with immobilized III1-C peptide. The binding of Fn to III1-C was inhibited by the presence of either rFbpA or rFbpB (Fig. 5). This result suggests that C. perfringens Fbps may inhibit Fn-matrix formation in vivo.
We thank Takahiro Hiraiwa, Tatsuma Tsuchiya and Masaya Okuda for generating the monoclonal antibodies. We also thank Kana Harutsumi for technical support.