Chymotryptic Digestion Research Articles

Identification of Cross-linking Site(s) of Myosin Heavy Chains in Oxidatively Stressed Chicken Myofibrils

ABSTRACT: The cross-linking site(s) of myosin heavy chains (MHC) in chicken myofibrils exposed to non-enzymatic, hydroxyl radical-generation oxidizing systems (HRGS) was investigated by means of chymotryptic digestion and subsequent electrophoresis. Oxidation of the chymotryptic digests resulted in cross-linking of the rod or light meromyosin (LMM) segment of MHC mostly via disulfide bonds, while subfragment-1 (S-1) or heavy meromyosin (HMM) was not affected. A mixture of cross-linked rod or LMM and uncross-linked S-1 or HMM was also produced when myofibrils were 1st oxidized and then digested with chymotrypsin, confirming that cross-linking of myosin in HRGS-oxidized myofibrils occurred initially in the LMM portion of the myosin rod.

Fragmentation of Angiotensin-I Converting Enzyme Inhibitory Peptides from Bonito Meat Under Intestinal Digestion Conditions and their Characterization

Kinetics of fragmentation of angiotensin-I converting enzyme inhibitory peptides obtained by digestion in gastric juice were studied under intestinal digestion conditions and their inhibitory activities were determined. A fragment IKYGD produced by digestion, as well as IKWGD synthesized, showed similar inhibitory activity to the original peptides. These peptides somehow were resistant to tryptic and/or chymotryptic digestion, and IK + aromatic amino acid might be important functional parts in some kinds of ACE inhibitory peptides.

Identification of phosphorylation sites in glycine N‐methyltransferase from rat liver

Previous studies have shown that rat glycine N-methyltransferase (GNMT) is phosphorylated in vivo, and could be phosphorylated in vitro on serine residues with a significant increase of enzyme activity, but no phosphorylation sites were identified. In this work the identification of the specific phosphorylation sites of rat GNMT is reported. Three different preparations of rat GNMT were analyzed: (1) purified from liver by standard methods of protein purification, (2) prepared from isolated hepatocytes and from liver tissue by immunoprecipitation, and (3) recombinant protein expressed in Escherichia coli. We measured the molecular weights of protein isoforms using electrospray mass spectrometry and used liquid chromatography-tandem mass spectrometry (LC-MS/MS) of peptides resulting from tryptic and chymotryptic digests. We also performed chemical analysis of phosphoamino acids and protein sequencing. In all samples, the phosphorylated serine residues 71, 182, and 241 were found. In GNMT prepared from liver tissue and hepatocytes an S9 additional residue was found to be phosphorylated. In hepatocytes and in recombinant GNMT S139 was detected. Serine 9 was also identified as a target for cAMP-dependent protein kinase in vitro. The positions of these phosphorylated residues in the tertiary structure of GNMT indicate their possible effect on enzyme conformation and activity.

Detection of human butyrylcholinesterase-nerve gas adducts by liquid chromatography–mass spectrometric analysis after in gel chymotryptic digestion

To verify the exposure to nerve gas, a method for detecting human butyrylcholinesterase (BuChE)-nerve gas adduct was developed using LC-electrospray mass spectrometry (ESI-MS). Purified human serum BuChE was incubated with sarin, soman or VX, and the adduct was purified by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) and digested in gel by treatment with chymotrypsin. The resulting peptide mixture was subjected to LC-ESI-MS. From the chymotryptic digest of untreated human BuChE, one peak corresponding to the peptide fragment containing the active center serine residue was detected on the extracted ion chromatogram at m/z 948.5, and the sequence was ascertained to be "GESAGAASVSL" by MS/MS analysis. From the chymotryptic digest of the human BuChE-sarin adduct, a singly charged peptide peak was detected on the extracted ion chromatogram at m/z 1,069.5, and the sequence was ascertained to be "GEXAGAASVSL" by MS/MS analysis (X denotes isopropylmethylphosphonylated serine). The difference in molecular weight (120.0 Da) between the active center peptide fragments corresponding to the untreated BuChE and BuChE-sarin adduct was assumed to be derived from the addition of an isopropyl methylphosphonyl moiety to the serine residue. The formation of human BuChE adducts with soman, VX and an aged soman adduct was confirmed by detecting the respective active center peptide fragments using LC-ESI-MS. To apply the established method to an actual biological sample, human serum was incubated with VX, and the adduct was purified by procainamide affinity chromatography followed by SDS-PAGE. After chymotryptic in gel digestion, the ethylphosphonylated active center peptide fragment could be detected, and the structure of the residue was ascertained by LC-ESI-MS analysis.

Epitope mapping factor VIII A2 domain by affinity-directed mass spectrometry: residues 497-510 and 584-593 comprise a discontinuous epitope for the monoclonal antibody R8B12.

The murine monoclonal antibody R8B12 recognizes the C-terminal region (residues 563-740) of the A2 subunit of factor VIIIa [J Biol Chem 266: 1991; p. 20139], as judged by Western blotting. However, the location of the epitope within this region is not known. In the present study, we used affinity-directed mass spectrometry to map the epitope. A2 subunit was digested with trypsin or chymotrypsin and then subjected to immunoprecipitation (IP) using R8B12 IgG. Masses of the affinity-selected peptides were determined directly from the immune complexes by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). Proteolysis of A2 with the two proteases generated a pre-IP peptide fingerprint that covered approximately 70% of the A2 domain sequence. Analysis of the post-IP tryptic peptide fingerprint showed two masses, 1309 and 1653 Da representing residues 584-593 and 497-510, respectively, determined from a theoretical database search and confirmed by direct sequencing. Results using a chymotryptic digest yielded a single, weakly reactive fragment consistent with residues 577-586, suggesting the importance of residues Ser584-Tyr586 in forming the epitope. A synthetic peptide to residues 584-593 was immunoprecipitated by the IgG and blocked R8B12-directed blotting to A2 subunit. The 497-510 and 584-593 segments were observed to be adjacent and surface exposed in the A2 domain model, and together with the above results suggest that A2 domain residues 497-510 and 584-593 represent a discontinuous epitope for R8B12. Furthermore, based upon blotting specificity, we speculate that residues 584-593 make a substantially greater contribution to the binding energy for this interaction.

Differential Inhibition of Arabidopsis Methionine Adenosyltransferases by Protein S-Nitrosylation

In animals, protein S-nitrosylation, the covalent attachment of NO to the thiol group of cysteine residues, is an intensively investigated posttranslational modification, which regulates many different processes. A growing body of evidence suggests that this type of redox-based regulation mechanism plays a pivotal role in plants, too. Here we report the molecular mechanism for S-nitrosylation of methionine adenosyltransferase (MAT) of Arabidopsis thaliana, thereby presenting the first detailed characterization of S-nitrosylation in plants. We cloned three MAT isoforms of Arabidopsis and tested the effect of NO on the activity of the purified, recombinant proteins. Our data showed that incubation with GSNO resulted in blunt, reversible inhibition of MAT1, whereas MAT2 and MAT3 were not significantly affected. Cys-114 of MAT1 was identified as the most promising target of NO-induced inhibition of MAT1, because this residue is absent in MAT2 and MAT3. Structural analysis of MAT1 revealed that Cys-114 is located nearby the putative substrate binding site of this enzyme. Furthermore, Cys-114 is flanked by S-nitrosylation-promoting amino acids. The inhibitory effect of GSNO was drastically reduced when Cys-114 of MAT1 was replaced by arginine, and mass spectrometric analyses of Cys-114-containing peptides obtained after chymotryptic digestion demonstrated that Cys-114 of MAT1 is indeed S-nitrosylated. Because MAT catalyzes the synthesis of the ethylene precursor S-adenosylmethionine and NO is known to influence ethylene production in plants, this enzyme probably mediates the cross-talk between ethylene and NO signaling.

Identification of Immunodominant Epitopes of α-Gliadin in HLA-DQ8 Transgenic Mice following Oral Immunization

Celiac disease, triggered by wheat gliadin and related prolamins from barley and rye, is characterized by a strong association with HLA-DQ2 and HLA-DQ8 genes. Gliadin is a mixture of many proteins that makes difficult the identification of major immunodominant epitopes. To address this issue, we expressed in Escherichia coli a recombinant alpha-gliadin (r-alpha-gliadin) showing the most conserved sequence among the fraction of alpha-gliadins. HLA-DQ8 mice, on a gluten-free diet, were intragastrically immunized with a chymotryptic digest of r-alpha-gliadin along with cholera toxin as adjuvant. Spleen and mesenteric lymph node T cell responses were analyzed for in vitro proliferative assay using a panel of synthetic peptides encompassing the entire sequence of r-alpha-gliadin. Two immunodominant epitopes corresponding to peptide p13 (aa 120-139) and p23 (aa 220-239) were identified. The response was restricted to DQ and mediated by CD4+ T cells. In vitro tissue transglutaminase deamidation of both peptides did not increase the response; furthermore, tissue transglutaminase catalyzed extensive deamidation in vitro along the entire r-alpha-gliadin molecule, but failed to elicit new immunogenic determinants. Surprisingly, the analysis of the cytokine profile showed that both deamidated and native peptides induced preferentially IFN-gamma secretion, despite the use of cholera toxin, a mucosal adjuvant that normally induces a Th2 response to bystander Ags. Taken together, these data suggest that, in this model of gluten hypersensitivity, deamidation is not a prerequisite for the initiation of gluten responses.

Analysis and sequencing of the active-site peptide from native and organophosphate-inactivated acetylcholinesterase by electrospray ionization, quadrupole/time-of-flight (QTOF) mass spectrometry

A method to identify and sequence recombinant mouse acetylcholinesterase (rMoAChE) including the native and organophosphate-modified active-site peptides was developed using capillary liquid chromatography with electrospray ionization, quadrupole/time-of-flight mass spectrometry. Addition of 2-propanol to the reversed-phase gradient system and a decreased gradient slope improved the peptide resolution and the signal of the active-site peptide. The highest protein coverage and active-site peptide signal were achieved when the rMoAChE:chymotrypsin ratio of 5:1 was used with digestion at 37 degrees C. rMoAChE and the active-site peptide were identified and sequenced from chymotryptic digests of native, methyl paraoxon-, and ethyl paraoxon-inactivated rMoAChE showing unequivocally that the exact modification site was the active-site serine.

Arginine Methylation of Yeast mRNA-binding Protein Npl3 Directly Affects Its Function, Nuclear Export, and Intranuclear Protein Interactions

Arginine methylation can affect both nucleocytoplasmic transport and protein-protein interactions of RNA-binding proteins. These effects are seen in cells that lack the yeast hnRNP methyltransferase (HMT1), raising the question of whether effects on specific proteins are direct or indirect. The presence of multiple arginines in individual methylated proteins also raises the question of whether overall methylation or methylation of a subset of arginines affects protein function. We have used the yeast mRNA-binding protein Npl3 to address these questions in vivo. Matrix-assisted laser desorption/ionization Fourier transform mass spectrometry was used to identify 17 methylated arginines in Npl3 purified from yeast: whereas 10 Arg-Gly-Gly (RGG) tripeptides were exclusively dimethylated, variable levels of methylation were found for 5 RGG and 2 RG motif arginines. We constructed a set of Npl3 proteins in which subsets of the RGG arginines were mutated to lysine. Expression of these mutant proteins as the sole form of Npl3 specifically affected growth of a strain that requires Hmt1. Although decreased growth generally correlated with increased numbers of Arg-to-Lys mutations, lysine substitutions in the N terminus of the RGG domain showed more severe effects. Npl3 with all 15 RGG arginines mutated to lysine exited the nucleus independent of Hmt1, indicating a direct effect of methylation on Npl3 transport. These mutations also resulted in a decreased, methylation-independent interaction of Npl3 with transcription elongation factor Tho2 and inhibited Npl3 self-association. These results support a model in which arginine methylation facilitates Npl3 export directly by weakening contacts with nuclear proteins.

Characterization of wheat gliadin proteins by combined two-dimensional gel electrophoresis and tandem mass spectrometry

A proteomics-based approach was used for characterizing wheat gliadins from an Italian common wheat (Triticum aestivum) cultivar. A two-dimensional gel electrophoresis (2-DE) map of roughly 40 spots was obtained by submitting the 70% alcohol-soluble crude protein extract to isoelectric focusing on immobilized pH gradient strips across two pH gradient ranges, i.e., 3-10 or pH 6-11, and to sodium dodecyl sulfate-polyacrylamide electrophoresis in the second dimension. The chymotryptic digest of each spot was characterized by matrix-assisted laser desorption/ionization-time of flight mass spectrometry and nano electrospray ionization-tandem mass spectrometry (MS/MS) analysis, providing a "peptide map" for each digest. The measured masses were subsequently sought in databases for sequences. For accurate identification of the parent protein, it was necessary to determine de novo sequences by MS/MS experiments on the peptides. By partial mass fingerprinting, we identified protein molecules such as alpha/beta-, gamma-, omega-gliadin, and high molecular weight-glutenin. The single spots along the 2-DE map were discriminated on the basis of their amino acid sequence traits. alpha-Gliadin, the most represented wheat protein in databases, was highly conserved as the relative N-terminal sequence of the components from the 2-DE map contained only a few silent amino acid substitutions. The other closely related gliadins were identified by sequencing internal peptide chains. The results gave insight into the complex nature of gliadin heterogeneity. This approach has provided us with sound reference data for differentiating gliadins amongst wheat varieties.

Comparative study on thermal denaturation modes of myosin in walleye pollack and carp myofibrils as affected by salt concentration

The effect of salt concentration on the thermal denaturation profile of myosin in walleye pollack and carp myofibrils was compared by studying the subfragment-1 (S-1) and rod denaturation rates upon heating. Species-specific denaturation mode observed at 0.1 M KCl was no longer detected when samples were heated above 0.5 M KCl, where S-1 and rod denaturation rates were identical to each other. As the heating of the chymotryptic digest of myofibril formed practically no rod aggregates, S-1 de naturation in a form of myosin was the rate limiting step for rod aggregate formation. As the aggregate formation by rod was remarkably suppressed by lowering the temperature, the free movement of myosin tail upon heating was suggested to play an important role in the rod aggregate formation in a high salt medium.

Species-specific thermal denaturation pattern of fish myosin when heated as myofibrils as studied by myosin subfragment-1 and rod denaturation rates

Thermal denaturation of myofibrils from various species of fish was investigated by measuring ATPase inactivation, myosin aggregation, myosin subfragment-1 (S-1) and rod denaturation rates as studied by chymotryptic digestion. Decrease in monomeric myosin (myosin aggregation) was always faster than the ATPase inactivation for all myofibrils tested. The relative denaturation rate of rod to that of S-1 differed from species to species. Preceded denaturation of rod was observed with some species, and the opposite was true with other species. The denaturation pattern was explained by the different magnitude of S-1 stabilization by F-actin in myofibrils at low salt medium. Myofibrils which receive a great stabilization by F-actin as studied by ATPase inactivation showed the preceded rod denaturation pattern, and vice versa. S-1 portion, not F-actin, determined the different stabilization of S-1 by F-actin in myofibrils.

Affinity Labelling of Enzymes with Triazine Dyes

Horse liver alcohol dehydrogenase is irreversibly inactivated by Procion blue MX-R, a dichlorotriazinyl structural analogue of Cibacron blue F3G-A, with over 90% loss of activity within 30 min at pH 8.5 and 37 degrees C at a reactive dye concentration of 1 mM and enzyme subunit concentration of 5 microM. Methoxylated Procion blue MX-R does not inactivate the enzyme. The inactivation of horse liver alcohol dehydrogenase by Procion blue MX-R is competitively inhibited by the pyridine nucleotides NAD+ and NADH. Quantitatively inhibited horse liver alcohol dehydrogenase contains 1 mol dye/mol subunit of Mr 40 000. Chymotryptic digestion and resolution of the peptides by reverse-phase high-performance liquid chromatography yields a single blue peptide which on sequencing and analysis yields an amino acid sequence of: (formula; see text) with the affinity label, Procion blue MX-R, unambiguously identified as being attached to the thiol side chain of Cys-174 in the catalytic domain of the enzyme. The specific active-site-directed reaction of Procion blue MX-R with horse liver alcohol dehydrogenase is interpreted in terms of the known crystallographic structure of the enzyme.

Characterization of the Oligomer Structure of Recombinant Human Mannan-binding Lectin

Mannan-binding lectin (MBL) belongs to a family of proteins called the collectins, which show large differences in their ultrastructures. These differences are believed to be determined by different N-terminal disulfide-bonding patterns. So far only the bonding pattern of two of the simple forms (recombinant rat MBL-C and bovine CL-43) have been determined. Recombinant MBL expressed in human cells was purified, and the structure of the N-terminal region was determined. Preliminary results on human plasma-derived MBL revealed high similarity to the recombinant protein. Here we report the structure of the N-terminal part of recombinant human MBL and present a model to explain the oligomerization pattern. Using a strategy of consecutive enzymatic digestions and matrix-assisted laser desorption ionization mass spectrometry, we succeeded in identifying a number of disulfide-linked peptides from the N-terminal cysteine-rich region. Based on these building blocks, we propose a model that can explain the various oligomeric forms found in purified MBL preparations. Furthermore, the model was challenged by the production of cysteine to serine mutants of the three N-terminally situated cysteines. The oligomerization patterns of these mutants support the proposed model. The model indicates that the polypeptide dimer is the basic unit in the oligomerization.

Gelsolin binds to polyphosphoinositide-free lipid vesicles and simultaneously to actin microfilaments

Gelsolin is a calcium-, pH- and lipid-dependent actin filament severing/capping protein whose main function is to regulate the assembly state of the actin cytoskeleton. Gelsolin is associated with membranes in cells, and it is generally assumed that this interaction is mediated by PPIs (polyphosphoinositides), since an interaction with these lipids has been characterized in vitro. We demonstrate that non-PPI lipids also bind gelsolin, especially at low pH. The data suggest further that gelsolin becomes partially buried in the lipid bilayer under mildly acidic conditions, in a manner that is not dependent of the presence of PPIs. Our data also suggest that lipid binding involves a number of sites that are spread throughout the gelsolin molecule. Linker regions between gelsolin domains have been implicated by other work, notably the linker between G1 and G2 (gelsolin domains 1 and 2 respectively), and we postulate that the linker region between the N-terminal and C-terminal halves of gelsolin (between G3 and G4) is also involved in the interaction with lipids. This region is compatible with other studies in which additional binding sites have been located within G4-6. The lipid-gelsolin interactions reported in the present paper are not calcium-dependent, and are likely to involve significant conformational changes to the gelsolin molecule, as the chymotryptic digest pattern is altered by the presence of lipids under our conditions. We also report that vesicle-bound gelsolin is capable of binding to actin filaments, presumably through barbed end capping. Gelsolin bound to vesicles can nucleate actin assembly, but is less active in severing microfilaments.

Preparation of heavy meromyosin from the autolyzed squid mantle muscle homogenate

Incubation of squid mantle muscle homogenate caused a selective cleavage of myosin into heavy meromyosin (HMM) and light meromyosin (LMM). HMM was isolated from the incubated homogenate by using ammonium sulfate fractionation. The purified HMM retained two types of light chain components. Its Mg2+-ATPase activity with or without F-actin showed a Ca=-sensitivity. HMM was cleaved into subfragment-1 and subfragment-2 upon chymotryptic digestion with or without Ca2+, possessing different light chain composition. Two types of light chain component were kept intact when digested in the presence of Ca2+, Ca2+ stabilized HMM especially in a bound form to F-actin.

Titania as a chemo-affinity support for the column-switching HPLC analysis of phosphopeptides: application to the characterization of phosphorylation sites in proteins by combination with protease digestion and electrospray ionization mass spectrometry.

A method for the determination of phosphorylation sites in phosphoproteins based on column-switching high-performance liquid chromatography (HPLC) has been developed. The HPLC system consisted of a titania precolumn for the selective adsorption of phosphopeptides, an anion-exchange analytical column and a UV detector (215 nm). Rabbit muscle phosphorylase a (RPa) and porcine stomach pepsin (PSP) were tested as model phosphoproteins. After protease digestion, the resulting phosphopeptides were successfully isolated by column-switching HPLC. The phosphopeptide fractions were analyzed by electrospray ionization mass spectrometry with a positive or negative ion mode after purification by reversed-phase HPLC. Pseudo-molecular ion peaks corresponding to Gln-Ile-Ser(p)-Val-Arg (MW 681.7) and Glu-Ala-Thr-Ser(p)-Gln-Glu-Leu (MW 856.8) were detected from the tryptic digest of RPa and chymotryptic digest of PSP, respectively, which agreed with the theoretically expected phosphopeptide fragments.

Multiple Domains of the Co-chaperone Hop Are Important for Hsp70 Binding

The Hop/Sti1 co-chaperone binds to both Hsp70 and Hsp90. Biochemical and co-crystallographic studies have suggested that the EEVD-containing C terminus of Hsp70 or Hsp90 binds specifically to one of the Hop tetratricopeptide repeat domains, TPR1 or TPR2a, respectively. Mutational analyses of Hsp70 and Hop were undertaken to better characterize interactions between the C terminus of Hsp70 and Hop domains. Surprisingly, truncation of EEVD plus as many as 34 additional amino acids from the Hsp70 C terminus did not reduce the ability of Hsp70 mutants to co-immunoprecipitate with Hop, although further truncation eliminated Hop binding. Hop point mutations targeting a carboxylate clamp position in TPR1 disrupted Hsp70 binding, as was expected; however, similar point mutations in TPR2a or TPR2b also inhibited Hsp70 binding in some settings. Using a yeast-based in vivo assay for Hop function, wild type Hop and TPR2b mutants could fully complement deletion of Sti1p; TPR1 and TPR2a point mutants could partially restore activity. Conformations of Hop and Hop mutants were probed by limited proteolysis. The TPR1 mutant digested in a similar manner to wild type; however, TPR2a and TPR2b mutants each displayed greater resistance to chymotryptic digestion. All point mutants retained an ability to dimerize, and none appeared to be grossly misfolded. These results raise questions about current models for Hop/Hsp70 interaction.

Limited proteolysis reveals conformational changes in uncoupling protein-1 from brown adipose tissue mitochondria

Limited proteolytic digestion of uncoupling protein-1 (UCP1) from hamster brown adipose tissue mitochondria was studied. Under optimal conditions, trypsin and chymotrypsin cleave at Lys-292 and at Phe-102, yielding major products 31-kDa T1 and 22-kDa Ch1. Both T1 and Ch1 remained dimers, as in UCP1. Using fluorescent nucleotide derivative 2 ′- O-dansyl GTP, it is shown that T1 retains the nucleotide binding affinity ( K D =1 μ M for dansyl GTP) while Ch1 does not bind nucleotide. Previously kinetic binding and H + transport studies [Biochemistry 35 (1996) 7846] have shown that UCP1 forms tight complexes to varying degrees with nucleotides and their derivatives. Nucleotides strongly protect against tryptic digestion but less against chymotryptic digestion, because the chymotryptic product Ch1 does not bind nucleotide. The nucleotides and derivatives show the same potency profile in protecting against both trypsinolysis and chymotryptic digestion, suggesting that UCP1 undergoes a major conformational change upon nucleotide binding from an initial loose complex into a tight complex, in which the cleavage sites become masked from proteolysis.

Characterization of functional domains of the hemolytic lectin CEL-III from the marine invertebrate Cucumaria echinata.

CEL-III is a Ca(2+)-dependent, galactose/N-acetylgalactosamine (GalNAc)-specific lectin isolated from the marine invertebrate Cucumaria echinata. This lectin exhibits strong hemolytic activity and cytotoxicity through pore formation in target cell membranes. The amino acid sequence of CEL-III revealed the N-terminal two-thirds to have homology to the B-chains of ricin and abrin, which are galactose-specific plant toxic lectins; the C-terminal one-third shows no homology to any known proteins. To examine the carbohydrate-binding ability of the N-terminal region of CEL-III, the protein comprising Pyr1-Phe283 was expressed in Escherichia coli cells. The expressed protein showed both the ability to bind to a GalNAc-immobilized column as well as hemagglutinating activity for rabbit erythrocytes, confirming that the N-terminal region has binding activity for specific carbohydrates. Since the C-terminal region could not be expressed in E. coli cells, a fragment containing this region was produced by limited proteolysis of the native protein by trypsin. The resulting C-terminal 15 kDa fragment of CEL-III exhibited a tendency to self-associate, forming an oligomer. When mixed with erythrocytes, the oligomer of the C-terminal fragment caused hemagglutination, probably due to hydrophobic interaction with cell membranes, while the monomeric fragment did not. Chymotryptic digestion of the preformed CEL-III oligomer induced upon lactose binding also yielded an oligomer of the C-terminal fragment comprising six molecules of the 16 kDa fragment. These results suggest that after binding to cell surface carbohydrate chains, CEL-III oligomerizes through C-terminal domains, leading to the formation of ion-permeable pores by hydrophobic interaction with the cell membrane.