Fabrication of a cell-recognition/electron-transfer/cross-linker, peptide-immobilized electrode for the sensing of K562 cells.

We designed an electrode that has the ability to sense a target cell. This new electrode is intended for use in cell recognition via electron-transfer and cross-linker peptide immobilization. Myelopeptide-4 (MP-4:FRPRIMTP) is a marrow-origin peptide that interacts with receptors of the human leukemia cell line (K562 cells), and allows their differentiation. The YYYYC electron-transfer peptide improves the electron-transfer accessibility from an electroactive compound to an electrode. Oligoalanine plays the role of a cross-linker that immobilizes a peptide series (Ac-FRPRIMTPYYYYCAAAAA) to collagen, which then allows it to be cast onto an electrode. Use of the electrode with a peptide increased the peak currents of [Fe(CN)6]4-/3- and also improved the reversibility of redox. These improvements are due to the interaction between [Fe(CN)6]4-/3- and the peptide. When electrochemical impedance spectroscopy (EIS) measurements were carried out using a collagen/peptide probe-immobilized electrode, the electron transfer resisitance was lower than that without the peptide. The detection of K562 cells was based on an increase in resistance, because MP-4 was bound to the receptors on the cell surface. The responses were linear and ranged in number from 27 to 2,000 cells/mLwith a detection limit of 8 cells/mL. Recoveries of 50 and 1,000 cells/mL in human serum were accomplished at rates of 98 and 101%, respectively. Consequently, the proposed procedure is a powerful new concept for cytosensing.

Label-free cytosensing of cancer cells based on the interaction between protein and an electron-transfer carbohydrate-mimetic peptide.

We used an electron-transfer carbohydrate-mimetic peptide (YYYYC) to construct an electrochemical cytosensing system. Magnetic beads were modified with either asialofetuin (ASF) or soybean agglutinin (SBA) to evaluate the effect on cell sensing. Because SBA binds to the galactose residue that exists at the terminals of the carbohydrate chains in ASF, the target protein was accumulated on the protein magnetic beads. SBA is an example of N-acetylgalactosamine- and galactose-binding proteins that readily combine with YYYYC. When the peptides and protein-immobilized beads competed for a target protein, the peak current of the peptides changed according to the concentration of the protein at the 10-12 M level. Next, human myeloid leukemia cells (K562 cell) were measured using the peptide and the carbohydrate chains on the cell surface that recognize SBA. The electrode response was linear to the number of K562 cells and ranged from 1.0 × 102 to 5.0 × 103 cells mL-1. In addition, detection of a human liver cancer cell (HepG2 cell) was carried out using interactions with the peptide, the ASF receptors in HepG2 cells, and the carbohydrate chains of ASF. The peak currents were proportional and ranged between 5.0 × 101 and 1.5 × 103 cells mL-1. When the values estimated from an electrochemical process were compared with those obtained by ELISA, the results were within the acceptable range of measurement error.

Magnetic beads modified with an electron-transfer carbohydrate-mimetic peptide for sensing of a galactose-dependent protein.

For use in the voltammetric sensing of galactose-dependent proteins, we modified magnetic beads with a peptide that had both electroactive- and molecular recognition properties. The peptide consisted of a YXY sequence and behaved as an electron-transfer carbohydrate-mimetic peptide that would combine with proteins. With this tool, the protein could be detected via a label-free system. We synthesized several penta- and hexa-peptides with a cysteine residue on the C-terminals to examine the properties of peptides. These peptides contained amino acid residues (X) of alanine, serine, or tyrosine. The peptides were immobilized on magnetic beads via N-(8-maleimidocapryloxy) succinimide. Soybean agglutinin(SBA), the in vivo function of which has been well established in animals, was selected as a model protein. The protein was detected via the changes in electrode response due to the oxidation of tyrosine residues from the phenol group to quinone. As a result, SBA was selectively accumulated on the beads modified with YYYYC. The calibration curve of SBA was linear and ranged from 2.5 × 10-12 to 1.0 × 10-10 M. With this system, SBA was recovered in human serum at values that ranged from 98 to 103%. Furthermore, the beads with peptides were regenerated five times using a protein denaturant. Accordingly, this electrochemical system was simple and could be rapidly applied to the detection of galactose-recognition proteins.

Integrin ß1 is bound to galectin-1 in human trophoblast.

Interaction of sugar binding proteins-galectins, with glycoconjugates is considered relevant for various reproductive processes. Galectin-1 (gal-1) is a molecule involved in trophoblast cell invasion, which is accomplished through interaction with cell surface and/or extracellular matrix glycoproteins. A possibility of interaction of endogenous gal-1 and trophoblast ß1 integrins, both previously shown relevant for trophoblast invasion, was investigated. Confocal microscopy showed overlap in gal-1 and ß1 integrin localization at the plasma membrane of isolated cytotrophoblast, HTR-8/SVneo extravillous trophoblast cell line and JAr choriocarcinoma cells. Immunoprecipitation confirmed an interaction of gal-1 with integrin ß1, but not with α1 or α5 integrin subunits. Nondenaturing electrophoresis and subcellular fractionation suggested association of gal-1 with ß1 integrin in intracellular and plasma membrane compartments of HTR-8/SVneo cells. Gal-1/ß1 integrin complex was sensitive to chemical and enzyme treatments, indicating carbohydrate dependent interaction. Down-regulation of gal-1 by siRNA, however, had no effect on level or distribution of ß1 integrin, as determined by qPCR and flow cytometry. These results suggest complex lectin type interaction of gal-1 with ß1 integrin at the trophoblast cell membrane, which could influence trophoblast cell adhesion, migration and invasion.

Design of carbohydrate/electron-transfer peptides for human histocytic lymphoma cell sensing.

A carbohydrate/electro-transfer peptide probe was fabricated to perform cell sensing. The probe consisted of a cello-oligosaccharide that was created by the conjugation of an electron-transfer peptide (Y5C) and a carbohydrate via a Schiff base. An oxidation wave due to a phenolic hydroxyl group was obtained by scanning with a glassy carbon electrode. This cell-sensing system was based on a competitive reaction between carbohydrates on a cell surface and the probe as each reacted to a protein that recognized the carbohydrate. When amounts of the protein and probe were constant, the peak current of the probe was changed as the number of cells increased. A human histocytic lymphoma cell (U937 cell) with carbohydrates such as glucose and N-acetylglucosamine on its surface was selected as the target cell. Wheat germ agglutinin (WGA) binded to both the probe and the carbohydrates on U937 cells, which resulted in a linear peak current of the cellobiose/electron-transfer peptide at concentrations that ranged from 100 to 3500 cells/ml. The values of the cell sensing using this electrochemical method were consistent with those established via ELSIA. The sensitivity of this procedure, however, was two-fold superior to that of ELISA. Consequently, this carbohydrate/electron-transfer peptide could be a powerful tool for cell sensing and searching for carbohydrate chains on a cell surface.

Galectin-1 enhances the generation of neural crest cells.

Neural crest (NC) cells are multipotent cells that emerge from the dorsal region of the neural tube. After delaminating from the neural tube, NC cells migrate throughout the developing embryo and differentiate into various cells: neurons and glial cells of the peripheral nervous system, melanocytes of skin, and skeletal elements of the face and head. We previously analyzed the gene expression profile of a NC subpopulation isolated from Sox10-IRES-Venus mice and found that the carbohydrate-binding protein, Galectin-1 (Gal-1) was strongly expressed in generating NC cells. In the present study, we identified GAL-1 as a factor that promotes NC cell generation. Gal-1 was significantly expressed in NC cells generated in explanted neural tubes. The presence of GAL-1 enhanced the generation of NC-like cells from mouse embryonic stem (ES) cells. In the differentiation of ES cells into NC-like cells, GAL-1 enhanced neurogenesis in the early stages and facilitated NC-like cell generation in the later stages. GAL-1 also enhanced the generation of NC cells from explanted neural tubes. These results suggest that GAL-1 plays a facilitative role in NC cell generation.

Stress-induced galectin-1 influences immune tolerance in the spleen and thymus by modulating CD45 immunoreactive lymphocytes.

Galectin-1 (Gal-1) is differentially expressed in normal and pathological tissues and regulates immune cell homeostasis. Restraint stress increases serum Gal-1 in rats. However, the function of stress-induced Gal-1 in serum is unknown. We determined if stress-induced Gal-1 in serum accumulates in immunocompetent organs as protection from physiological and/or psychological stress. Western blotting showed that the intensity of Gal-1 bands in stressed groups was significantly higher than that in controls. RT-PCR analysis indicated that the Gal-1 mRNA level did not increase after restraint stress. The numbers of Gal-1 immunoreactive cells in the splenic periarterial lymphatic sheath (PLS) and the thymus medulla of the stressed group were increased compared with those in controls. Furthermore, stress-induced Gal-1 immunoreactive cells corresponded to CD45 immunoreactive lymphocytes (CD45+) in the PLS of the spleen and the medulla of the thymus. Thus, stress-induced Gal-1 immediately accumulates in the spleen and thymus, and may modulate the immune response through apoptosis by binding to CD45+ lymphocytes in immune organs following physiological and/or psychological stress.

Sensing lymphoma cells based on a cell-penetrating/apoptosis-inducing/electron-transfer peptide probe.

To electrochemically sense lymphoma cells (U937), we fabricated a multifunctional peptide probe that consists of cell-penetrating/apoptosis-inducing/electron-transfer peptides. Electron-transfer peptides derive from cysteine residue combined with the C-terminals of four tyrosine residues (Y4). A peptide whereby Y4C is bound to the C-terminals of protegrin 1 (RGGRLCYCRRRFCVCVGR-NH2) is known to be an apoptosis-inducing agent against U937 cells, and is referred to as a peptide-1 probe. An oxidation response of the peptide-1 probe has been observed due to a phenolic hydroxyl group, and this response is decreased by the uptake of the peptide probe into the cells. To improve the cell membrane permeability against U937 cells, the RGGR at the N-terminals of the peptide-1 probe was replaced by RRRR (peptide-2 probe). In contrast, RNRCKGTDVQAWY4C (peptide-3 probe), which recognizes ovalbumin, was constructed as a control. Compared with the other probes, the change in the peak current of the peptide-2 probe was the greatest at low concentrations and occurred in a short amount of time. Therefore, the cell membrane permeability of the peptide-2 probe was increased based on the arginine residues and the apoptosis-inducing peptides. The peak current was linear and ranged from 100 to 1000 cells/ml. The relative standard deviation of 600 cells/ml was 5.0% (n = 5). Furthermore, the membrane permeability of the peptide probes was confirmed using fluorescent dye.

Interaction of extravillous trophoblast galectin-1 and mucin(s)-Is there a functional relevance?

In the course of embryo implantation extensive interaction of the trophoblast with uterine tissue is crucial for adequate trophoblast invasion. This interaction is highly controlled, and it has been pointed out that a specific glycocode and changes in glycosylation may be important for successful implantation and maintenance of pregnancy. Both uterine and trophoblast cells have been shown to express cell surface glycoconjugates and sugar binding proteins, such as mucins (MUC) and galectins (gals). An increasing number of studies have investigated potential candidates interacting in this process. However, knowledge about the biochemical nature of the interactions and their importance for trophoblast cell function, and, consequently, for pregnancy outcome are still lacking. This review is aimed at deliberating the possibility that mucins, as heavily glycosylated proteins, might be among the functionally relevant galectin ligands in human trophoblast, based on both published data and our original research.

Galectin signature of the choriocarcinoma JAr cells: Galectin-1 as a modulator of invasiveness in vitro.

Our previous findings showed that galectin-1 (LGALS1) plays an important role in the in vitro invasion of normal human trophoblast cells. In the present study, choriocarcinoma JAr cells were found to express LGALS1, -2, -3, -8, -10, and -13 mRNA and at least LGALS1, -3, and -8 protein, as determined by reverse-transcriptase PCR and Western blot, respectively. The galectin mRNA signature of JAr cells thus differed from that of normal first-trimester extravillous trophoblasts. A Matrigel migration assay was also used to investigate and confirm the relevance and effect of LGALS1 on the invasive potential of JAr cells, as observed in other trophoblast models. This modulation in behavior was achieved by specific lectin-glycan binding.

Monitoring of the interaction between U937 cells and electroactive daunomycin with an arginine-rich peptide.

Daunomycin penetrates the membrane of a U937 cell, which is a human histiocyte-related lymphoma cell. Several arginine-rich peptides have also exhibited a high degree of permeability with these cells. Therefore, we attempted to improve the membrane permeability of daunomycin by coupling it with an arginine-rich peptide. The cell membrane permeability of daunomycin was monitored using voltammetry, because daunomycin is an electroactive compound. First, daunomycin was combined with N-(6-maleimidocaproyloxy)sulfosuccinimide. Second, the cross-linking agent with daunomycin was bound to the cysteine residue of RRRRRRRRGC (peptide-1). The two-step synthesis suppressed the formation of by-products that might have conjugated with the amino groups of peptide-1. After the quinone moieties of daunomycin were reduced using an electrode, an oxidation peak appeared due to the moieties. The peak current of daunomycin with U937 cells had decreased. For the mixture of the daunomycin/peptide-1 probe and cells, the electrode response was smaller than that of daunomycin with the cells. Thus, the membrane penetration of the daunomycin/peptide-1 probe was improved compared with the use of only daunomycin. In addition, the membrane penetration of the probe was measured using fluorescence spectroscopy. The sensitivity of the electrochemical procedure was 100-fold that was obtained by fluorescence spectrometry.

Construction of a peptide with an electroactive daunomycin like a pendant arm to detect ovalbumin.

In this study, a peptide-1 (RNRCKGTDVQAW) constructing lysozyme was conjugated with an electroactive daunomycin in order to voltammetrically detect ovalbumin (OVA). Hetero-bifunctional cross-linking agents with four kinds of ethylene chains in differing lengths were used to bind the peptide-1 and daunomycin. After a cross-linking agent had reacted with an amino group of daunomycin, the compound was introduced into the peptide to the cysteine residue in the peptide using a pendant arm. The OVA was sensed via a change in the electrode response of the daunomycin moiety, based on the binding between the peptide and the OVA. The adsorption of the peptide probe on the electrode increased with increases in the ethylene chain. The binding constants between the peptide probes and the OVA, however, did not depend on the length of the chain. This was because the ethylene chain influenced the binding. When the peptide and the daunomycin were bound using N-(6-maleimidocaproyloxy) sulfosuccinimide, the electrode response of the peptide probe was the most sensitive from among the four cross-linking agents. The calibration curve of the OVA using the peptide probe was linear and ranged from 1.5×10(-11) to 3.0×10(-10)M. Furthermore, this method could be applied to the electrochemical sensing of the OVA in egg whites and in fetal bovine serum.

Galectin-1 deficiency improves axonal swelling of motor neurones in SOD1(G93A) transgenic mice.

AIMS: Galectin-1, a member of the ß-galactoside-binding lectin family, accumulates in neurofilamentous lesions in the spinal cords of both sporadic and familial amyotrophic lateral sclerosis (ALS) patients with a superoxide dismutase 1 gene (SOD1) mutation (A4V). The aim of this study was to evaluate the roles of endogenous galectin-1 in the pathogenesis of ALS. METHODS: Expression of galectin-1 in the spinal cord of mutant SOD1 transgenic (SOD1(G93A) ) mice was examined by pathological analysis, real-time RT-PCR and Western blotting. The effects of galectin-1 deficiency were evaluated by cross-breeding SOD1(G93A) mice with galectin-1 null (Lgals1(-/-) ) mice. RESULTS: Before ALS-like symptoms developed in SOD1(G93A) /Lgals1(+/+) mice, strong galectin-1 immunoreactivity was observed in swollen motor axons and colocalized with aggregated neurofilaments. Electron microscopic observations revealed that the diameters of swollen motor axons in the spinal cord were significantly smaller in SOD1(G93A) /Lgals1(-/-) mice, and there was less accumulation of vacuoles compared with SOD1(G93A) /Lgals1(+/+) mice. In symptomatic SOD1(G93A) /Lgals1(+/+) mice, astrocytes surrounding motor axons expressed a high level of galectin-1. CONCLUSIONS: Galectin-1 accumulates in neurofilamentous lesions in SOD1(G93A) mice, as previously reported in humans with ALS. Galectin-1 accumulation in motor axons occurs before the development of ALS-like symptoms and is associated with early processes of axonal degeneration in SOD1(G93A) mice. In contrast, galectin-1 expressed in astrocytes may be involved in axonal degeneration during symptom presentation.

Construction of an electrode modified with gallium(III) for voltammetric detection of ovalbumin.

Electrodes modified with gallium(III) complexes were constructed to detect ovalbumin (OVA). For immobilization of a gallium(III)-nitrilotriacetate (NTA) complex, the electrode was first covered with collagen film. After the amino groups of the film had reacted with isothiocyanobenzyl-NTA, the gallium(III) was then able to combine with the NTA moieties. Another design featured an electrode cast with a gallium(III)-acetylacetonate (AA) complex. The amount of gallium(III) in the NTA complex was equivalent to one-quarter of the gallium(III) that could be utilized from an AA complex. However, the calibration curves of OVA using gallium(III)-NTA and gallium(III)-AA complexes were linear in the ranges of 7.0 × 10(-11) - 3.0 × 10(-9) M and 5.0 × 10(-10) - 8.0 × 10(-9) M, respectively. The gallium(III) on the electrode with NTA complex had high flexibility due to the existence of a spacer between the NTA and the collagen film, and, therefore, the reactivity of the gallium(III) to OVA was superior to that of the gallium(III)-AA complex with no spacer.

Voltammetric detection of ovalbumin using a peptide labeled with an electroactive compound.

For this study, a new method was developed to electrochemically detect ovalbumin via its binding with the peptide-1(RNRCKGTDVQAW) in lysozymes. The peptide that exists at the C-terminal of a lysozyme was combined with ovalbumin. When an electroactive compound was introduced to the N-terminal side of the peptide through ethylene gycolbis(sulfosuccinimidyl succinate), the labeled peptide-1 served as a probe for the detection of ovalbumin. The electrode responses of labeled peptide-1 were measured after the labeled peptide-1 and ovalbumin were incubated in a 0.1 M phosphate buffer (pH 5.6). As a result, the electrode response decreased as the concentration of ovalbumin increased. The detection limit of ovalbumin was 2.3 × 10(-11) M as estimated at 3-fold the standard deviation (3σ) (n = 5). Because the steric structure of the peptide and some of the amino acid residues were related to the binding, we prepared a peptide-2, to which the N- and C-terminals of peptide-1 were alternated. The decrease in the response for the labeled peptide-2 was less than that for the labeled peptide-1. In addition, the peak current of a peptide-3, for which the D of peptide-1 was replaced with S, was hardly changed with or without ovalbumin. Therefore, it was clear that the binding was influenced by the steric factors and by the sequence of the peptide. However, a peptide-1 with bis(sulfosuccinimidyl) suberate was designed to investigate the hydrophobic influences on the probe. The change in the peak current was smaller than that of peptide-1 with ethylene gycolbis(sulfosuccinimidyl succinate), which was due to the hydrophobic properties of the alkyl chain between the peptide and the ovalbumin. The proposed method could be applied to the determination of ovalbumin in egg whites. Consequently, the concept becomes an electrochemical sensing method for proteins based on the protein-peptide interaction.

Electrochemical sensing of concanavalin A using a non-ionic surfactant with a maltose moiety.

To electrochemically detect concanavalin A (ConA), a new method was developed using mixed micelles between a non-ionic surfactant with a maltose moiety and electroactive daunomycin. The surfactants, in which the length of the alkyl chain was different, were n-decyl-ß-D-maltoside, n-dodecyl-ß-D-maltoside, and n-tetradecyl-ß-D-maltoside. The measurement principle was due to the micelle breakdown caused by the binding between the ConA and maltose moieties. When ConA was combined with maltose moieties at a concentration of surfactant that was near the critical micelle concentration, the daunomycin that formed the micelles was moved to a solution from the micelles. As a result, the peak current of daunomycin increased as the concentration of ConA was increased. The mechanism was proposed using voltammetry, spectrometry, and gel filtration. The linear range using n-tetradecyl-ß-D-maltoside was 2.0×10(-9) to 8.0×10(-8) M of ConA, and it was the most sensitive in the presence of the three surfactants. To examine whether selective binding took place, measurements with several proteins were carried out. The electrode responses of daunomycin were not influenced by the presence of 5.0×10(-6) M protein. Furthermore, this method could be applied to the determination of ConA in a serum, and to the measurement of sugar chains that can be combined with ConA on the cell surface.

Characterization of galectin-1-positive cells in the mouse hippocampus.

Galectin-1 (gal-1) is one of several well-studied proteins from the galectin families. It is a 14.5 kDa glycoprotein with a single carbohydrate-binding domain. To examine the distribution and properties of gal-1 in the mouse hippocampus, we performed immunohistochemistry using an anti-gal-1 antibody. We found that most gal-1-positive cells showed both NeuN and ß-tubulin III (Tuj-1) immunoreactivity (NeuN: 93%, ß-tubulin III: 88%). Furthermore, we clarified that 77% of gal-1-positive cells expressed somatostatin, 79% of gal-1-positive cells expressed GAD67, 34% of gal-1-positive cells expressed parvalbumin, 5% of gal-1-positive cells expressed calretinin, 2% of gal-1-positive cells expressed calbindin, and 31% of gal-1-positive cells expressed neuropeptide Y in the mouse hippocampus. These results indicate that gal-1 is expressed in interneurons that also express ß-tubulin III and gal-1 may be a novel marker for interneuron subpopulations in the hippocampus.

GDNF promotes neurite outgrowth and upregulates galectin-1 through the RET/PI3K signaling in cultured adult rat dorsal root ganglion neurons.

Galectin-1 (GAL-1), a member of a family of ß-galactoside binding animal lectins, is predominantly expressed in isolectin B4 (IB4)-binding small non-peptidergic (glial cell line-derived neurotrophic factor (GDNF)-responsive) sensory neurons in the sections of adult rat dorsal root ganglia (DRG), but its functional role and the regulatory mechanisms of its expression in the peripheral nervous system remain unclear. In the present study, both recombinant nerve growth factor (NGF) and GDNF (50 ng/ml) promoted neurite outgrowth from cultured adult rat DRG neurons, whereas GDNF, but not NGF, significantly increased the number of IB4-binding neurons and the relative protein expression of GAL-1 in the neuron-enriched culture of DRG. The GAL-1 expression in immortalized adult rat Schwann cells IFRS1 and DRG neuron-IFRS1 cocultures was unaltered by treatment with GDNF, which suggests that GDNF/GAL-1 signaling axis is more related to neurite outgrowth, rather than neuron-Schwann cell interactions. The GDNF-induced neurite outgrowth and GAL-1 upregulation were attenuated by anti-GDNF family receptor (RET) antibody and phosphatidyl inositol-3'-phosphate-kinase (PI3K) inhibitor LY294002, suggesting that the neurite-outgrowth promoting activity of GDNF may be attributable, at least partially, to the upregulation of GAL-1 through RET-PI3K pathway. On the contrary, no significant differences were observed between GAL-1 knockout and wild-type mice in DRG neurite outgrowth in the presence or absence of GDNF. Considerable immunohistochemical colocalization of GAL-3 with GAL-1 in DRG sections and GDNF-induced upregulation of GAL-3 in cultured DRG neurons imply the functional redundancy between these galectins.

Electrochemical assay of concanavalin A-ovalbumin binding on magnetic beads.

To monitor protein-glycoprotein interactions on magnetic beads, the present study developed an electrochemical assay of the binding between concanavalin A (ConA) and ovalbumin (OVA). The system was a powerful model that could be used to evaluate cell junctions. ConA with an electroactive daunomycin was immobilized on 6 different sizes of magnetic beads (diameter: 1.0-8.9 µm) through a cross-linking agent. Six sizes of OVA-beads (diameter: 1.0-8.9 µm) were also prepared using a similar method. The binding was evaluated using an oxidation peak of ConA with daunomycin because ConA recognized OVA with α-mannose residues. When binding took place on the beads' surface, the peak current was decreased due to the electroactive moieties being covered with OVA. When ConA/daunomycin-OVA binding was evaluated, the change of the peak current obtained by the beads (diameter: 8.9 µm) modified with ConA and daunomycin was the greatest in the presence of OVA-modified beads (diameter: 2.5 µm). In contrast, particle agglomeration was observed for the smallest beads (diameter: 1.0 µm) with ConA/daunomycin and OVA. The results suggested that ConA-OVA binding depended on the size of beads. Thus, this method could be applied to measure protein-glycoprotein interactions on the cell surface.

Voltammetric sensing of phosphoproteins using a gallium(III) acetylacetonate-modified carbon paste electrode.

The voltammetric detection of phosphoproteins was developed using a gallium(III) acetylacetonate-modified carbon paste electrode. Because phosphate groups of the protein interacted with the gallium(III) ion, the protein was accumulated on the electrode surface. A hexaammine ruthenium(III) ion, which combined with the functional groups, was used to monitor the interaction. When phosvitin and hexaammine ruthenium(III) ions were incubated in 0.1 M acetate buffer (pH 3.2), a reduction peak of hexaammine ruthenium(III) ion at the electrode decreased as the concentration of the protein increased. In contrast, an increase in the peak current was observed with a plain carbon paste electrode. These results were caused by a competitive reaction of the phosphate groups with the hexaammine ruthenium(III) and gallium(III) ions. In the presence of α-, ß- and κ-caseins, the electrode response decreased due to the order of the numbers of phosphate groups. This method could be applied to the sensing of phosphoproteins at the 10(-10) M level.