Search

The introduction of intermolecular disulfide bonds by amino acid mutations is an effective method for stabilizing dimeric proteins. X-ray crystal structure of Fab of a therapeutic antibody, adalimumab, revealed the first loop of the CH1 domain to be partially unsolved at position 135–141. To find new sites for the introduction of intermolecular disulfide bonds in adalimumab Fab, Fab mutants targeting the unsolved region were predicted using molecular simulation software. Four Fab mutants, H:K137C-L:I117C, H:K137C-L:F209C, H:S138C-L:F116C and H:S140C-L:S114C, were expressed in the methylotrophic yeast Pichia pastoris. SDS-PAGE analysis of these mutants indicated that H:K137C-L:F209C, H:S138C-L:F116C and H:S140C-L:S114C mutants mostly formed intermolecular disulfide bonds, whereas some H:K137C-L:I117C mutants formed intermolecular disulfide bonds and some did not. Differential scanning calorimetry measurements showed increased thermal stability in all Fab mutants with engineered disulfide bonds. The bio-layer interferometry measurements, for binding of the antigen tumor necrotic factor α, indicated that Fab mutants had less antigen-binding activity than wild-type Fab. In particular, the KD value of H:K137C-L:F209C was ~17 times higher than that of wild-type Fab. Thus, we successfully introduced intermolecular disulfide bonds between the first loop region of the CH1 and CL domains and observed that it increases the thermostability of Fab and affects the antigen-binding activity.

Peptidases or proteinases are now classified into seven families based on the nature of the catalytic residues [MEROPS-the peptidase database (http://merops.sanger.ac.uk/)]. They are aspartic- (first described in 1993), cysteine- (1993), serine- (1993) metallo- (1993), threonine- (1997), glutamic- (2004) and asparagine-peptidase (2010). By using an S-PI (pepstatin Ac) as a probe, a new subfamily of serine peptidase, serine-carboxyl peptidase (sedolisin) was discovered in 2001. In addition, the sixth family of peptidase, glutamic peptidase (eqolisin) was also discovered in 2004. The former peptidase is widely distributed in nature from archea to mammals, including humans. One of these enzymes is related to a human fatal hereditable disease, Batten disease. In contrast, the distribution of the latter peptidases is limited, with most of them found in human or plant pathogenic fungi. One such enzyme was isolated from a fungal infection in an HIV-infected patient. In this review, the background of the findings, and crystal structures, catalytic mechanisms, substrates specificities and distribution of the new peptidase families are described.

We found that CEL-I, a GalNAc-specific C-type lectin isolated from the marine invertebrate Holothuroidea (Cucumaria echinata), induces inducible nitric oxide synthase (iNOS) expression and NO production in RAW264.7 cells. The NO production was inhibited by an iNOS inhibitor, L-NAME, but was not by a lipopolysaccharide (LPS) inhibitor, polymyxin B. In the presence of 0.1-M GalNAc, increased NO production by CEL-I-treated RAW264.7 cells was observed rather than the inhibition. Bovine serum albumin (BSA) significantly inhibited the CEL-I-induced NO production as well as the binding of FITC-labelled CEL-I on RAW264.7 cells. Three MAP kinase inhibitors (specific to extra-cellular regulated kinase, c-jun NH(2)-terminal kinase and p38 MAP kinase) inhibited CEL-I-induced NO production with different extents. Heat-treatment of CEL-I resulted in a decreased activity of CEL-I depending on the temperature. These results suggest that CEL-I induces NO production in RAW264.7 cells through the protein-cell interaction rather than the binding to the specific carbohydrate chains on the cell surface.

We found that CEL-I, a GalNAc-specific C-type lectin isolated from the marine invertebrate Holothuroidea (Cucumaria echinata), induces inducible nitric oxide synthase (iNOS) expression and NO production in RAW264.7 cells. The NO production was inhibited by an iNOS inhibitor, L-NAME, but was not by a lipopolysaccharide (LPS) inhibitor, polymyxin B. In the presence of 0.1-M GalNAc, increased NO production by CEL-I-treated RAW264.7 cells was observed rather than the inhibition. Bovine serum albumin (BSA) significantly inhibited the CEL-I-induced NO production as well as the binding of FITC-labelled CEL-I on RAW264.7 cells. Three MAP kinase inhibitors (specific to extra-cellular regulated kinase, c-jun NH(2)-terminal kinase and p38 MAP kinase) inhibited CEL-I-induced NO production with different extents. Heat-treatment of CEL-I resulted in a decreased activity of CEL-I depending on the temperature. These results suggest that CEL-I induces NO production in RAW264.7 cells through the protein-cell interaction rather than the binding to the specific carbohydrate chains on the cell surface., Journal of biochemistry, 146(2), pp.209-217; 2009

An LC-MS study revealed some heterogeneity in terms of molecular mass of a cysteine-free mutant of dihydrofolate reductase (DHFR) after long storage of the highly purified protein as an ammonium sulfate precipitate, but not in the case of a cysteine- and methioneine-free mutant of DHFR. One-third of the cysteine-free DHFR sample stored for a long time, around 18 months, comprised molecular species with molecular masses increased by 16, 32 and 48 Da. A peptide mapping study revealed that at least one of the methionine residues at positions 1, 16 and 20 was oxidatively modified to a methione-sulfoxide residue, while those at positions 42 and 92 were essentially unaffected. Each of the oxidized species of the DHFR exhibiting different degrees or sites of oxidation was further purified to essentially homogeneity in terms of molecular mass from the stored sample, and its enzyme activity was determined. One oxidized DHFR showed higher activity than that of the non-oxidized enzyme, while the other four oxidized DHFRs showed less activity. This agrees with the observation that the enzyme activity of the stored sample, a mixture in terms of oxidation, was apparently the same as that of the non-oxidized enzyme. This suggests that the activity itself is not a proper measure for quality control of proteins.

Prodigiosins are natural red pigments that have multi-biological activities. Recently, we discovered a marine bacterial strain, which produces a red pigment. Extensive two-dimensional nuclear magnetic resonance and mass spectrometry analysis showed that the pigment is a prodigiosin analogue (PG-L-1). Here, we investigated the effect of PG-L-1 on NADPH oxidase activity in macrophage cells. PG-L-1 significantly inhibited superoxide anion (O(2)(-)) production by phorbol 12-myristate 13-acetate (PMA)-stimulated RAW 264.7 cells, a mouse macrophage cell line. The ED(50) value was estimated to be approximately 0.3 microM. PG-L-1 had no direct scavenging effect on O(2)(-) generated by hypoxanthine/xanthine oxidase system in electron spin resonance-spin trapping determinations, suggesting that this compound directly acts on the O(2)(-) production system, NADPH oxidase, in macrophage cells. We further investigated the effect of PG-L-1 on the behaviour of the cytosolic components of the NADPH oxidase, p67(phox), p47(phox), p40(phox), Rac and protein kinase C (PKC), in PMA-stimulated RAW 264.7 cells. Although PG-L-1 showed no effect on the activation of PKC, the immunoblotting analysis using specific antibodies showed that PG-L-1 strongly inhibits the association of p47(phox) and Rac in the plasma membrane of PMA-stimulated RAW 264.7 cells. These results suggest that PG-L-1 inactivates NADPH oxidase through the inhibition of the binding of p47(phox) and Rac to the membrane components of NADPH oxidase.

We found that CEL-I was a potent cytotoxic lectin. MDCK, HeLa, and XC cells were highly sensitive to CEL-I cytotoxicity and killed in a dose-dependent manner, whereas CHO, L929, and RAW264.7 cells were relatively resistant to CEL-I, and no significant toxicity was observed up to 10 microg/ml. Among these cell lines, MDCK cells showed the highest susceptibility to CEL-I cytotoxicity. A binding study using FITC-labeled CEL-I (F-CEL-I) revealed that the amounts of bound F-CEL-I on the sensitive cell lines were evidently greater than those on the resistant cell lines, suggesting that the different susceptibility of the cell lines to CEL-I cytotoxicity is partly explained by different efficiencies of binding of CEL-I to these cell lines. Interestingly, the cytotoxicity of CEL-I toward MDCK cells was more potent than those of other lectins such as WGA, PHA-L, and Con A, even though these lectins were capable of binding to MDCK cells at comparable levels to CEL-I. Since the cytotoxicity of CEL-I was strongly inhibited by GalNAc, the binding to cell surface specific carbohydrates is essential for the CEL-I cytotoxicity. The trypan blue dye exclusion test indicated that CEL-I caused a disorder of plasma membrane integrity as a relatively early event. CEL-I failed to induce the release of carboxyfluorescein (CF) from CF-loaded MDCK cells as seen for pore-forming hemolytic isolectin CEL-III, suggesting that the primary cellular target of CEL-I may be the plasma membrane, but its action mechanism differs from that of CEL-III. Although CEL-I induced dramatic cellular morphological changes in MDCK cells, neither typical apoptotic nuclear morphological changes nor DNA fragmentation was observed in CEL-I-treated MDCK cells even after such cellular changes. Our results demonstrated that CEL-I showed a potent cytotoxic effect, especially on MDCK cells, by causing plasma membrane disorder without induction of apoptosis.

A missense mutation in the gene of tissue-nonspecific alkaline phosphatase, which replaces aspartic acid at position 289 with valine [TNSALP (D289V)], was reported in a lethal hypophosphatasia patient [Taillandier, A. et al. (1999) Hum. Mut. 13, 171-172]. To define the molecular defects of TNSALP (D289V), this mutant protein in transiently transfected COS-1 cells was analyzed biochemically and morphologically. TNSALP (D289V) exhibited no alkaline phosphatase activity and mainly formed a disulfide-linked high molecular mass aggregate. Cell-surface biotinylation, digestion with phosphatidylinositol-specific phospholipase C and an immunofluorescence study showed that the mutant protein failed to appear on the cell surface and was accumulated intracellularly. In agreement with this, pulse/chase experiments demonstrated that TNSALP (D289V) remained endo-beta-N-acetyl- glucosaminidase H-sensitive throughout the chase and was eventually degraded, indicating that the mutant protein is unable to reach the medial-Golgi. Proteasome inhibitors strongly blocked the degradation of TNSALP (D289V), and furthermore the mutant protein was found to be ubiquitinated. Besides, another naturally occurring TNSALP with a Glu(218)-->Gly mutation was also found to be polyubiquitinated and degraded in the proteasome. Since the acidic amino acids at positions 218 and 289 of TNSALP are thought to be directly involved in the Ca(2+) coordination, these results suggest the critical importance of calcium binding in post-translational folding and assembly of the TNSALP molecule.

We have found that a brefeldin A (BFA)-resistant mutant cell line derived from Vero cells (BER-40) is highly resistant to ricin-induced apoptosis as compared with parental Vero cells. In BER-40 cells, all apoptotic events caused by ricin including cytolysis, nuclear morphological changes, and DNA fragmentation occur to a lesser extent than in Vero cells, even though both cell lines show similar sensitivities to ricin-mediated inhibition of protein synthesis. Furthermore, no significant apoptotic signaling events, such as increases in caspase-3 and -9-like activities, release of cytochrome c from mitochondria, or the cleavage of PARP, were observed in BER-40 cells under the conditions at which these changes were evident in Vero cells. Intracellular biochemical changes associated with ricin-induced apoptosis, such as the depletion of glutathione and an increase in free Zn2+, were also less apparent in BER-40 cells than in Vero cells. BER-40 cells were also found to be highly resistant to apoptosis induced by other toxins with different intoxication mechanisms such as diphtheria toxin, modeccin, and anisomycin. These results suggest that the entire apoptotic signal transduction mechanism in BER-40 cells, which may be triggered after the inhibition of protein synthesis by toxins, becomes resistant. Since MDCK cells, a naturally BFA resistant cell line, are highly sensitive to ricin-induced apoptosis, it seems likely that the BFA resistance phenotype may not necessarily lead to resistance to apoptotic cell death. Probably the underlaying BFA-resistance mechanism in BER-40 cells is distinct from that in MDCK cells, and the resistance to ricin-induced apoptosis of BER-40 cells may be a unique phenotype acquired concomitantly with BFA-resistance.

Our previous studies demonstrated that ricin induces the apoptotic death of U937 cells as evidenced by DNA fragmentation, nuclear morphological changes, and increases in caspase-like activities. In this study, we have found that intracellular NAD(+) and ATP levels decrease in ricin-treated U937 cells and that this decrease is followed by the ricin-mediated protein synthesis inhibition. The PARP inhibitor, 3-aminobenzamide (3-ABA), prevents the depletion in NAD(+) and ATP levels and concomitantly protects U937 cells from the lysis that follows ricin treatment. Hence, the protective action of 3-ABA is due to the inhibition of PARP and does not result from its other pharmacological side effects. Moreover, the enzymatic activity of PARP gradually increases and reaches a maximum level after ricin exposure for 3 h, whereas no significant change in activity was observed in untreated cells. However, 3-ABA has no effect on ricin-mediated DNA fragmentation. In addition, immunoblot analysis revealed that significant PARP cleavage occurred more than 12 h after ricin addition, while DNA fragmentation reached a maximum level within 6 h of incubation. Thus, in the case of ricin-induced apoptosis, it appears that PARP cleavage is not an early apoptotic event associated with the onset of apoptosis. Our results suggest that multiple apoptotic signaling pathways may be triggered by ricin-treatment. Probably, the pathway leading to cell lysis via PARP activation and NAD(+) depletion is independent of the pathway leading to DNA fragmentation in which caspases may be profoundly involved. Other protein synthesis inhibitors, including diphtheria toxin and cycloheximide, were less effective in terms of inducing DNA fragmentation and cytolysis, even at concentrations that cause significant inhibition of protein synthesis. Thus, a specific ricin action mechanism through which ribosomes are inactivated may be responsible for the apoptotic events induced by ricin.

Vertebrate skeletal muscle alpha-tropomyosin polymerizes in a head-to-tail manner and binds cooperatively to actin. It has been postulated that the cooperative actin binding is governed by the strength of the head-to-tail interaction. In order to know the relationship between the head-to-tail affinity and actin binding, we studied the properties of tropomyosin variants with single residue substitutions at serine-283, the penultimate residue at the carboxyl terminus that is involved in the head-to-tail interaction. It has been shown that the phosphorylation of serine-283 strengthens the head-to-tail interaction. Viscometry was employed to compare the head-to-tail affinity of tropomyosin variants. Variant S283E showed higher viscosity whereas variant S283K showed lower viscosity compared with the wild type non-phosphorylated alpha-tropomyosin. The results confirm the idea that the interaction is sensitive to the ionic properties of residue 283. The strength of the head-to-tail interaction was assessed directly by sedimentation equilibrium using two pairs of tropomyosin variants designed so that only dimeric interactions were allowed within each pair. From one pair of variants with serine-283, the association constant was determined to be 2.6 x 10(4) M(-1) (SD =1.0 x 10(4)), whereas for the second pair with glutamate-283, the affinity was 3.9 x 10(4) M(-1) (SD =1.6 x 10(4)), slightly stronger than the former, consistent with the results of viscometry. The results indicate that the head-to-tail association is weak as previously implicated from light scattering measurements. Cosedimentation was employed to measure the cooperative actin binding of tropomyosin variants. Although previous results indicated the phosphorylation has no significant influence on the actin affinity, variant S283E shows a lower affinity compared with the control. Variants S283K and S283A show even lower affinities to actin, although these species bind to actin more cooperatively than does variant S283E. The results indicate that the affinity of the head-to-tail interaction between adjacent tropomyosin molecules is weak, and is substantially influenced by an extra charge at residue 283. On the other hand, the interaction with actin, the affinity and the cooperativity in actin binding, is dependent on amino acid residues at 283 and is not simply correlated with the strength of the head-to-tail interaction between Tm molecules in solution.

Lipocalin-type prostaglandin D synthase (L-PGDS) is a highly glycosylated member of the lipocalin gene family and is secreted into various human body fluids. We comparatively analyzed the structures of asparagine-linked sugar chains of human L-PGDS produced by recombinant Chinese hamster ovary cells and naturally occurring human urine and amniotic fluid. After the sugar chains were liberated by hydrazinolysis followed by N-acetylation, they were derivatized with 2-aminobenzamide. All of the sugar chains of three L-PGDSs occur as biantennary complex-type sugar chains. Most of the sugar chains of three samples were fucosylated on the inner most N-acetylglucosamine residue. Although the sugar chains of the recombinant L-PGDS do not contain any bisecting N-acetylglucosamine residues, 58% and 34% of the fucosylated-sugar chains of amniotic fluid and urine L-PGDSs, respectively, contain bisecting N-acetylglucosamine residues. The sialic acid residues occur solely as Siaalpha2-->3Gal groups of the recombinant L-PGDS; the sialic acid residues of other L-PGDS occur as both Siaalpha2-->3Gal and Siaalpha2-->6Gal groups. Variations in L-PGDS glycosylation may prove useful as markers to further elucidate the role of L-PGDS glycoforms in different tissues.

In the rat liver, two mRNAs for serine:pyruvate (or alanine:glyoxylate) aminotransferase are generated from a single gene by alternative transcription initiation. The longer mRNA encodes a precursor of a mitochondrial enzyme that has a mitochondrial targeting signal at the N-terminus and is translocated into mitochondria. The shorter mRNA encodes a peroxisomal enzyme of mature size that is imported into peroxisomes. We have been interested in the mechanism of selective targeting to mitochondria of the precursor protein that also contains a peroxisomal targeting signal in the molecule. In this study, we examined the effect of the mitochondrial targeting signal on the conformation of the protein and on the function of the peroxisomal targeting signal in the precursor molecule. The results suggest that the mitochondrial targeting signal causes the conformation of the protein to become unfolded and that this conformational change in turn causes repression of the putative peroxisomal targeting signal contained in the precursor protein.

Although the depletion of reduced glutathione (GSH) has been observed in a variety of apoptotic systems, little is known about the mechanism of GSH depletion. In this study we used polarized MDCK cells to study the GSH flux during ricin-induced apoptosis. Here we report that the specific accumulation of GSH occurred in the basolateral medium during ricin treatment with similar kinetics to in apoptotic changes such as an increase in caspase-3 like activity and DNA fragmentation, while there was no significant increase in the GSH level in apical medium. These results suggest that GSH efflux occurred through a GSH-specific channel or transporter located in the basolateral membrane domain of polarized MDCK cells undergoing apoptosis. Treatment with other protein toxins such as modeccin, Pseudomonas toxin, and diphtheria toxin, which can induce apoptotic cell death, also resulted in selective GSH efflux from the basolateral side. Thus, GSH efflux through a specific transporter may be a common step of apoptosis induced by these toxins, while these toxins have different intoxication mechanisms leading to protein synthesis inhibition. Pretreatment of cells with Z-Asp-CH(2)-DCB, a caspase family inhibitor, inhibited ricin-induced basolateral GSH efflux as well as DNA fragmentation, suggesting that the activation of caspases, i.e. those that are inhibited by Z-Asp-CH(2)-DCB, is implicated in the opening of the GSH transporter.

The addition of CEL-III to sensitive MDCK cells preincubated with 8-anilino-1-naphthalenesulfonate (ANS) caused an increase in the fluorescence intensity of the probe. The increase in the ANS fluorescence caused by CEL-III was Ca2+-dependent and strongly inhibited by 0.1 M lactose, indicating that Ca2+-dependent binding of CEL-III to specific carbohydrate receptors on the plasma membrane is responsible for this phenomenon. In contrast, no significant effect of CEL-III on the ANS fluorescence was observed in CHO cells, which are highly resistant to CEL-III cytotoxicity. In MDCK cells, energy transfer from tryptophan residues to bound ANS molecules was observed in the presence of CEL-III, but not in CHO cells. Furthermore, the amount of ANS bound to MDCK cells increased as the concentration of CEL-III increased. Therefore, a simple interpretation is that the CEL-III-induced increase in ANS fluorescence is attributable to an increase of the hydrophobic region in the plasma membrane where ANS could bind. Immunoblotting analysis of proteins from cells treated with CEL-III indicated that CEL-III oligomers were irreversibly bound to the cells, and the amount of oligomer bound to MDCK cells was much greater than that bound to CHO cells under any conditions tested. The oligomerization may be accompanied by an enhancement of the hydrophobicity of CEL-III molecules, which in turn provides new ANS-binding sites. The difference in susceptibility of MDCK and CHO cells to CEL-III cytotoxicity may be due to a difference in oligomerization of bound CEL-III.

The biosynthesis of basement membrane heparan sulfate proteoglycan (HSPG), known as perlecan, in ACC3 cells established from a adenoid cystic carcinoma of the human salivary gland was studied using metabolic labeling and immunoprecipitation with discriminative antibodies specific for HSPG core protein. Treatment of immunoprecipitated HSPG with HNO2, heparitinase, and chondroitinase ABC revealed that ACC3 cells synthesized HSPG molecules composed of 470-kDa core protein and heparan sulfate but not of chondroitin sulfate. The core protein was shown to contain complex type N-linked oligosaccharides by digestion with N-glycanase and endoglycosidase H. Pulse-chase experiments showed that the mature form of HSPG was formed in the cells in 30 min and released into the medium thereafter. Degradation of HSPG was also found in the chase period of 3 h. In time course experiments, HSPG was found to be synthesized maximally at day 4 after plating, deposited in the cell layer maximally at day 6, and secreted maximally at day 8. This was also confirmed by immunofluorescence, Northern blotting, and in-situ hybridization. The results indicate that ACC3 cells synthesize, secrete and degrade basement membrane type HSPG, which is analogous to those produced by other cell types, and that the biosynthesis and secretion of HSPG in ACC3 cells are strictly regulated by the cell growth, that may be reflected in the characteristic histology of adenoid cystic carcinomas.

Pseudomonas carboxyl proteinase (PCP), isolated from Pseudomonas sp. 101, is the first example from a prokaryote of unique carboxyl proteinases [EC 3.4.23.33] which are insensitive to aspartic proteinase inhibitors, such as pepstatin, diazoacetyl-DL-norleucine methylester, and 1,2-epoxy-3(p-nitrophenoxy)propane. To identify the catalytic residue(s) of PCP, chemical modification was carried out using carboxyl residue-specific reagents, carbodiimides. PCP was inactivated effectively by N,N'-dicyclohexylcarbodiimide (DCCD) with pseudo-first-order kinetics. For the inactivation, 0.7 mol DCCD was involved per 1 mol PCP. The effects of pH and methanol on the inactivation showed that two carboxyl residues (Asp and/or Glu) were involved in the reaction. The inactivation by DCCD was prevented by a competitive inhibitor, tyrostatin, or a synthetic substrate in a concentration-dependent manner. Based on these data, differential labeling of PCP with DCCD was carried out: Firstly, PCP was treated with cold DCCD in the presence of tyrostatin. After removal of the tyrostatin, which covered the substrate binding site, by dialysis, the PCP was treated with [14C]DCCD to label carboxyl residue(s) essential for its function. Two labeled peptides were isolated by HPLC from a trypsin digest of cold- and [14C]DCCD modified PCP. On analysis of their amino acid sequences, it was revealed that the [14C]DCCD was bound to Asp140 and Glu222 of PCP, respectively. Based on these data, it was strongly suggested that Asp140 and Glu222 of PCP were involved in its catalytic function or substrate binding.

We investigated the involvement of caspases and serine proteases in apoptotic cell death induced by ricin, modeccin, diphtheria toxin, and Pseudomonas toxin in U937 cells. We found that caspase-3- and caspase-6-like activities, but not caspase-1-like activity, increased during toxin-induced apoptosis. Z-D-CH2-DCB, a caspase-like inhibitor, completely inhibited the generation of caspase-3- and caspase-6-like activities and blocked all features of apoptosis induced by toxins: nuclear morphological changes, DNA fragmentation, and cytotoxicity. However, three caspase-specific inhibitors, Ac-YVAD-CHO, Ac-DEVD-CHO, and Ac-VEID-CHO, had no effect, even though Ac-DEVD-CHO and Ac-VEID-CHO inhibited the increased caspase-3- and caspase-6-like activity, respectively. These results suggest that the generation of caspase-3- and caspase-6-like activities is redundant, and other caspases distinct from caspase-3 and -6 may be important in toxin-induced apoptosis. Furthermore, serine protease inhibitor, 3,4-dichloroisocoumarine (DCI), abolished the apoptotic cell death and DNA fragmentation caused by toxins, without affecting the increased caspase-3- and caspase-6-like activities. Our results suggest that multiple proteases with different preferences for apoptotic substrates participate in toxin-induced apoptotic death of U937 cells.

During a cyanocysteine-mediated dissection study of dihydrofolate reductase, a peptide fragment with a molecular mass of 18 Da less than expected was found as a major reaction product when the dissection reaction was applied to a Lys-cyanocysteine linkage. Detailed characterization of the dissection products by protease digestion, peptide sequencing, liquid chromatography/electrospray ionization mass spectrometry, and capillary electrophoresis suggested that the by-product was generated via a lactam ring formation through the intramolecular nucleophilic attack of the epsilon-amino group on the carbonyl carbon of the Lys-cyanocysteine linkage. We have also demonstrated the occurrence of intermolecular attack of an alpha-amino group of glycine on the carbonyl carbon of the X-cyanocysteine linkage to form a new X-Gly linkage, which should be a useful reaction for specific modification of proteins at the C-terminal.

We have studied the biosynthesis and intracellular transport of tissue-nonspecific alkaline phosphatase (TNSALP) transiently expressed in COS-1 cells. Mutations were introduced into TNSALP to examine the effects of a single amino acid substitution on the activity and biosynthesis of TNSALP. The cells expressing wild-type TNSALP exhibited more than 200-fold higher alkaline phosphatase activity than untransfected ones. Pulse-chase experiments showed that TNSALP was synthesized as a 66-kDa endoglucosaminidase H (Endo H)-sensitive form and converted to EndoH-resistant forms with heterogenous molecular masses ( approximately 80 kDa), which finally appeared on the cell surface as judged by digestion with phosphatidylinositol-specific phospholipase C (PI-PLC). In contrast, a TNSALP with a Glu218-->Gly mutation exhibited no phosphatase activity at all and the 66-kDa Endo H-sensitive form was the only molecular species throughout the chase in the transfected cells. In accordance with this finding, digestion with PI-PLC and immunofluorescence observation confirmed that this mutant was never expressed on the cell surface. Another mutant with a Ala162-->Thr substitution, which naturally occurs in association with a lethal hypophosphatasia, exhibited a low activity and only a small fraction of the 66-kDa form acquired Endo-H resistance and reached the cell surface. Since the wild-type and the mutant TNSALPs were labeled with [3H]ethanolamine, a component of glycosylphosphatidylinositol (GPI), it is unlikely that the impaired intracellular transport of the two mutants is due to a failure in their modification by GPI. Interestingly, the 66-kDa Endo H-sensitive form of the TNSALP mutants but not that of the wild-type, was found to form an interchain disulfide-bonded high-molecular-mass aggregate within the cells. These results suggest that impaired intracellular transport of the TNSALP (Ala162-->Thr) molecule caused by its aggregation is the molecular basis for the lethal hypophosphatasia carrying this mutation.

Pseudomonas carboxyl proteinase (PCP), isolated from Pseudomonas sp. 101, and Xanthomonas carboxyl proteinase (XCP), isolated from Xanthomonas sp. T-22, are the first and second examples of unique carboxyl proteinases [EC 3.4.23.33] which are insensitive to aspartic proteinase inhibitors, such as pepstatin, diazoacetyl-DL-norleucine methylester, and 1,2-epoxy-3(p-nitrophenoxy)propane. The substrate specificities of PCP and XCP were studied using a series of synthetic chromogenic peptide substrates with the general structure, P5-P4-P3-P2-Phe-Nph-P2'-P3' (P5, P4, P3, P2, P2', P3': a variety of amino acids, Nph is p-nitro-L-phenylalanine, and the Phe-Nph bond is cleaved). PCP and XCP were shown to hydrolyze a synthetic substrate, Lys-Pro-Ala-Leu-Phe-Nph-Arg-Leu, most effectively among 28 substrates. The kinetic parameters of this peptide for PCP were Km = 6.3 microM, Kcat = 51.4 s-1, and kcat/Km = 8.16 microM-1.s-1. The kinetic parameters for XCP were Km = 3.6 microM, kcat = 52.2 s-1, and kcat/Km = 14.5 microM-1.s-1. PCP showed a stricter substrate specificity than XCP. That is, the specificity constant (kcat/Km) of each substrate for PCP was in general0.5 microM-1.s-1, but was drastically improved by the replacement of Lys by Leu at the P2 position. On the other hand, XCP showed a less stringent substrate specificity, with most of the peptides exhibiting reasonable kcat/Km values (1.0 microM-1.s-1). Thus it was found that the substrate specificities of PCP and XCP differ considerably, in spite of the high similarity in their primary structures. In addition, tyrostatin was found to be a competitive inhibitor for XCP, with a Ki value of 2.1 nM, as well as for PCP (Ki = 2.6 nM).

Xanthomonas carboxyl proteinase (XCP), isolated from Xanthomonas sp. T-22, is the second example of the unique carboxyl proteinases [EC 3.4.23.33] which are insensitive to the classical aspartic proteinase inhibitor. The gene coding for XCP was cloned, sequenced, and expressed in Escherichia coli. The XCP gene contains an open reading frame of 2,481 base pairs encoding a protein of 827 amino acid residues with a M(r) of 83,677. The XCP was synthesized as a large precursor consisting of three regions: NH2-terminal prepro (N-Prepro) (237 amino acid residues); mature XCP (398 a.a.residues); and COOH-terminal pro (C-Pro) (192 a.a. residues). The N-Prepro and mature XCP regions had no sequence similarity to any other proteins reported so far, except the carboxyl proteinase from Pseudomonas sp. 101 [Oda, K., Takahashi, T., Tokuda, Y., Shibano, Y., and Takahashi, S. (1994) J. Biol. Chem. 269, 26518-26524]. The C-Pro region showed high similarity to COOH-terminal regions of other microbial proteinase precursors. E. coli carrying a plasmid containing the cloned wild-type XCP gene produced an 84-kDa protein. This protein was processed into a mature, active form under acidic conditions. This process was completely blocked by tyrostatin, an XCP-specific inhibitor from Kitasatosporia sp. 55, indicating an autocatalytic processing. The purified recombinant XCP had the same characteristics as authentic XCP except for the NH2-terminal amino acid sequence. When the mutant XCP gene truncated in the C-Pro region was expressed in E. coli, an expected 64-kDa protein was detected in the cells, and also processed into the 42-kDa active form under the acidic conditions. Thus, the C-Pro region was not essential for the formation of active mature XCP.

Plasmapheresis with a dextran sulfate column is a treatment for patients with hypercholesteremia. When proteins bound to the column during the treatment were fractionated to prepare some known proteins, we found a 57 kDa glycoprotein designated GP57 which showed a new N-terminal amino acid sequence. Western-blot analysis of human plasma revealed that only a 120 kDa protein, GP120, reacted with anti-GP57 antibody. Since GP120 and GP57 had an identical N-terminal amino acid sequence, GP120 is probably the intact form of GP57. The isoelectric point of GP120 was 6.8. N-Glycanase treatment decreased the molecular weight of GP120 by 15 kDa. Neuraminidase and O-glycanase, however, did not affect the molecular weight. Amino acid sequence analyses of the lysylendopeptidase digest of GP120 revealed significant homology to the heavy chains of inter-alpha-trypsin inhibitor (ITI) family. Since GP120 showed no bikunin sequence, and chondroitinase treatment and alkaline treatment of GP120 did not affect its molecular weight, we concluded that GP120 was not a complex with bikunin. We designated GP120 as IHRP (ITI heavy chain-related protein).

C-terminus of S19 ribosomal protein (RP S19) endows the cross-linked homodimer with a dual effect on the C5a receptor in leucocyte chemoattraction; agonistic effect on the monocyte receptor, and antagonistic effect on the neutrophil receptor. C5a exhibits the uniform agonistic effect on this receptor of both cell types. We have currently prepared a recombinant C5a-chimeric protein bearing the C-terminus of RP S19 (C5a/RP S19 chimera) to be used as a substitute of the RP S19 dimer. In vitro, this chimera similarly inhibited the intracellular Ca 2+ mobilization of neutrophils induced by C5a to the RP S19 dimer did. In the guinea pig skin, 10 −7 M C5a/RP S19 chimera exhibited an inhibitory capacity to the neutrophil infiltration induced by 3 × 10 −7 M C5a without enhancing monocyte infiltration. In reverse passive Arthus reaction, the neutrophil infiltration associated with plasma extravasation was significantly reduced by the simultaneous administration of 10 −7 M C5a/RP S19 chimera with antibodies. The C5a/RP S19 chimera is a useful tool not only to examine the molecular mechanism that underlies the functional difference of the C5a receptor between monocytes and neutrophils, but also to prevent C5a-mediated hyper-response of neutrophils in acute inflammation.

Our previous studies demonstrated that CEL-I, an N-acetylgalactosamine (GalNAc)-specific C-type lectin purified from the marine invertebrate Cucumaria echinata (Holothuroidea) showed potent cytotoxicity to several cell lines such as HeLa, MDCK and XC cells. In this study, we found that CEL-I induced increased secretion of tumour necrosis factor-alpha (TNF-alpha) and granulocyte colony stimulation factor (G-CSF) by mouse macrophage cell line RAW264.7 cells in a dose-dependent manner, whereas this cell line was highly resistant to CEL-I cytotoxicity. The cytokine-inducing activity of CEL-I was stronger than that of phytohaemagglutinin (PHA-L). A binding study using FITC-labelled CEL-I (F-CEL-I) indicated that the amount of bound F-CEL-I on RAW264.7 cells was greater than that of F-PHA-L, suggesting that the greater activity of CEL-I to induce cytokine secretion by RAW264.7 cells is partly due to the higher binding ability. Since the cell binding and cytokine-inducing activity of CEL-I were partly but significantly inhibited by the specific sugar (GalNAc), it is considered that the binding of CEL-I to cell-surface-specific saccharide moieties, which may be recognized by CEL-I with higher affinity than GalNAc, is essential for the induction of cytokine secretion. The secretion of TNF-alpha and G-CSF from CEL-I-treated RAW264.7 cells were almost completely prevented by brefeldin A (BFA), whereas increase in mRNA levels of these cytokines were not affected by BFA. Bio-Plex beads assay suggested that temporal increase in phosphorylation of extracellular-regulated kinase (ERK), c-jun NH(2)-terminal kinase (JNK) and p38 MAP kinase occurred at relatively early time following CEL-I treatment. Furthermore, the secretion of TNF-alpha and G-CSF were inhibited by specific inhibitors for these MAP kinases. These results suggest that the intracellular signal transduction through the activation of MAP kinase system is involved in CEL-I-induced cytokine secretion.

Phospholipase Cdelta4 (PLC delta4) gene has been cloned from the cDNA library of regenerating rat liver. Using PLC delta4 gene-disrupted mice (PLC delta4(-/-)), we studied a role of PLC delta4 during liver regeneration after partial hepatectomy (PH). In PLC delta4(-/-), liver regeneration occurred in an apparently normal way. However, BrdU-indices indicated that PLC delta4 gene disruption delayed the onset of DNA synthesis by 2 h. Noticeably, the BrdU-indices in PLC delta4(+/+) remained rather constant throughout S phase, 25-35%, whereas in PLC delta4(-/-), it fluctuated drastically from 25% at 34 h to 65% at late S, 42 h after PH. This fact showed that PLC delta4 gene disruption caused a higher synchronization of cell proliferation. The mRNA for PLC delta4 in PLC delta4(+/+) appeared at late G1, and the expression continued throughout S phase. PLC activity increased transiently in chromatin at the late G1 and S phases in only PLC delta4(+/+), but not in PLC delta4(-/-). The specific increases in PLC activity well correlated with the transient increases of protein kinase C (PKC) alpha in chromatin of PLC delta4(+/+). PKC epsilon also increased transiently in chromatin from PLC delta4(+/+) at late S. It is concluded that PLC delta4 regulates the liver regeneration in cooperation with nuclear PKC alpha and epsilon.

Tripeptidyl peptidase I (TTP-I), also known as CLN2, a member of the family of serine-carboxyl proteinases (S53), plays a crucial role in lysosomal protein degradation and a deficiency in this enzyme leads to fatal neurodegenerative disease. Recombinant human TPP-I and its mutants were analyzed in order to clarify the biochemical role of TPP-I and its mechanism of activity. Ser280, Glu77, and Asp81 were identified as the catalytic residues based on mutational analyses, inhibition studies, and sequence similarities with other family members. TPP-I hydrolyzed most effectively the peptide Ala-Arg-Phe*Nph-Arg-Leu (*, cleavage site) (k(cat)/K(m) = 2.94 microM(-1).s(-1)). The k(cat)/K(m) value for this substrate was 40 times higher than that for Ala-Ala-Phe-MCA. Coupled with other data, these results strongly suggest that the substrate-binding cleft of TPP-I is composed of only six subsites (S(3)-S(3)'). TPP-I prefers bulky and hydrophobic amino acid residues at the P(1) position and Ala, Arg, or Asp at the P(2) position. Hydrophilic interactions at the S(2) subsite are necessary for TPP-I, and this feature is unique among serine-carboxyl proteinases. TPP-I might have evolved from an ancestral gene in order to cleave, in cooperation with cathepsins, useless proteins in the lysosomal compartment.

Clostridium perfringens type A strains that produce alpha-toxin cause gas gangrene, which is a life-threatening infection with fever, pain, edema, myonecrosis and gas production. Intramuscular injection of the toxin or Bacillus subtilis carrying the alpha-toxin gene causes myonecrosis and produces histopathological features of the disease. Immunization of mice with alpha-toxin or fragments of the toxin prevents gas gangrene caused by C. perfringens. The toxin possesses phospholipase C (PLC), sphingomyelinase (SMase) and biological activities causing hemolysis, lethality and dermonecrosis. These biological activities are closely related to PLC and/or SMase activities. However, there is yet some uncertainty about the biological activities induced by the PLC and SMase activities of alpha-toxin. Based on the isolation and characterization of the gene for alpha-toxin and a comparison of the toxin with enzymes of the PLC family, significant progress has been made in determining the function-structure of alpha-toxin and the mode of action of the toxin. To provide a better understanding of the role of alpha-toxin in tissue damage in gas gangrene, this article summarizes current knowledge of the characteristics and mode of action of alpha-toxin.

CEL-I is a C-type lectin isolated from the Holothuroidea Cucumaria echinata. This lectin shows very high N-acetylgalactosamine-binding specificity. We constructed an artificial gene encoding recombinant CEL-I (rCEL-I) using a combination of synthetic oligonucleotides, and expressed it in Escherichia coli cells. Since the recombinant protein was obtained as inclusion bodies, the latter were solubilized using urea and 2-mercaptoethanol, and the protein was refolded during the purification and dialysis steps. The purified rCEL-I showed comparable hemagglutinating activity to that of native CEL-I at relatively high Ca(2+)-concentrations, whereas it was weaker at lower Ca(2+)-concentrations due to decreased Ca(2+)-binding affinity. rCEL-I exhibited similar carbohydrate-binding specificity to native CEL-I, including strong GalNAc-binding specificity, as examined by hemagglutination inhibition assay. Comparison of the far UV-CD spectra of recombinant and native CEL-I revealed that the two proteins undergo a similar conformational change upon binding of Ca(2+). Single crystals of rCEL-I were also obtained under the same conditions as those used for the native protein, suggesting that they have similar tertiary structures. Although native CEL-I exhibited strong cytotoxicity toward cultured cells, rCEL-I showed low cytotoxicity. These results indicate that rCEL-I has a tertiary structure and carbohydrate-binding specificity similar to those of native CEL-I. Howeger, there is a subtle difference in the properties between the two proteins probably due to the additional methionine residue at the N-terminus of rCEL-I.