保田 立二

We developed a sensitive method for detection of glycosphingolipid (GSL) antigen(s) on the cell surface. As a model of GSL antigen, ganglioside GD3 was used. An IgM monoclonal antibody (DSG-1) specific for ganglioside GD3 was preincubated with standard inhibitor liposomes containing ganglioside GD3. Then carboxyfluorescein-entrapped indicator liposomes containing ganglioside GD3 and complement were added. Release of the marker from the indicator liposomes was specifically inhibited by inhibitor liposomes. The assay system was simple, sensitive, reproducible, and semiquantitative. Pg to ng of ganglioside GD3 could be detected. Furthermore, ganglioside GD3 on the cells was investigated with SK-MEL-28 human melanoma cell line and human red blood cells (HRBC). When SK-MEL-28 melanoma with ganglioside GD3 was used as an inhibitor, specific inhibition was observed. However, HRBC without ganglioside GD3 showed no significant inhibition. The marker release was 50% inhibited by 1.4 x 10(6)SK-MEL-28 melanoma cells/ml. The amount of ganglioside GD3/melanoma cell was estimated to be at least 1.1 x 10(-14) g from the standard curve made with the liposomes containing 10% epitope density of ganglioside GD3. This assay system may be useful for detection of GSL antigen on the cell.

To identify ARIX gene and PHOX2B gene polymorphisms in patients with congenital superior oblique muscle palsy, 3 exons of the ARIX gene and PHOX2B gene were sequenced by genomic DNA amplification with polymerase chain reaction (PCR) and direct sequencing in 31 patients with congenital superior oblique muscle palsy and in 54 normal individuals. A family with a father and one daughter each having congenital superior oblique muscle palsy was also included in this study. Eleven patients with congenital superior oblique muscle palsy had heterozygous nucleotide changes in the ARIX gene, including 4 patients reported on previously. One patient with atrophy of the superior oblique muscle had a new change of T-4G in the promoter region of the ARIX gene. The other 6 patients had a heterozygous nucleotide change of G153A in the 5'-untranslated region (UTR) of the exon 1 of the ARIX gene. These nucleotide changes of the ARIX gene, taken together, had a significant association with congenital superior oblique muscle palsy(P = 0.0022). One patient and 5 patients had heterozygous nucleotide changes of A1106 C and A1121 C in exon 3 of the PHOX2B gene, respectively, while these changes were absent in the normal individuals. Two patients had both the G153A change in the 5'-UTR of exon 1 of the ARIX gene and the A1121 C change in exon 3 of the PHOX2B gene. In conclusion, the polymorphisms of the ARIX gene and PHOX2B gene may be genetic risk factors for the development of congenital superior oblique muscle palsy.

A carboxyfluorescein (CF)-enveloping soybean phosphatidylcholine liposome was used as a model of physicochemical damage of biomembranes. The liposomes were exposed to a metal-chelate complex [2 mM of ferric nitrilotriacetate (FeNTA) or cupric nitrilotriacetate (CuNTA)] plus a reductant (2 mM of ascorbate or various concentrations of reduced glutathione), and CF release from damaged liposomal membranes and the generation of thiobarbituric acid-reactive substances (TBARS) were measured. In the presence of a reducing agent, both FeNTA and CuNTA stimulated markedly CF release and an increase in the TBARS level, while in the absence of a reducing agent both of the chelate complexes showed little CF release and TBARS. The effects of H2O2 addition to the reaction system containing liposome with FeNTA or CuNTA plus ascorbate were also examined. The CF release was slightly increased by the addition of a smaller dose (0.5 mM) of H2O2 and it was inhibited by 8 mM of H2O2. A similar result was obtained in the TBARS test. These results suggest that FeNTA- or CuNTA-mediated lipid peroxidation can damage liposomal membranes physicochemically, and the redox reaction of the chelated metal itself is more important than a Fenton-type reaction in the process.

We have already developed the liposome immune lysis assay (LILA) for the determination of C-reactive protein (CRP) by employing an inhibition method and a sandwich method. We herein report a new LILA system involving the use of monoclonal antibodies-bearing liposomes. We established five monoclonal antibodies to CRP antigen, AC-1, -2, -3, -4, -5 which had the capacity to activate complement and form antigen-antibody complex. Each of these antibodies was covalently coupled to carboxyfluorescein-entrapped multilamellar liposomes. When the liposomes were incubated with CRP antigen in the presence of guinea pig complement, CRP antigen-dependent liposome lysis was observed but the sensitivity was not great enough for practical use. On the other hand, when liposomes coupling two monoclonal antibodies (AC-1, AC-2) which recognized distinct CRP antigenic determinants were employed in the assay, the sensitivity increased compared with that using only one monoclonal antibody, and the detectable concentration range was 5-300 ng/ml. These results indicated that the combination of two or more monoclonal antibodies which recognize distinct CRP antigenic determinants is effective for increasing the sensitivity of the assay.

We compared the transfection efficiency of four types of positively charged liposomes composed of (i) N-(α-trimethylammonioacetyl)-didodecyl-D-glutamate chloride (TMAG), dilauroylphosphatidylcholine (DLPC), and dioleoylphosphatidylethanolamine (DOPE) (1:2:2 molar ratio); (ii) 3β [N-(N′, N′-dimethylaminoethane)-carbamoyl] cholesterol (DC-Chol) and DOPE (3:2 molar ratio); (iii) dimethyldioctadecylammonium bromide (DDAB) and DOPE (1:2.2 molar ratio); (iv) N-[1-(2,3-dioleyloxy) propyl] -N,N,N-trimethylammonium chloride (DOTMA) and DOPE (1:1, w/w; lipofectin). Luciferase gene was used as a reporter gene. Among the cationic liposomes used, the liposomes composed of TMAG, DOPE and DLPC showed a much higher efficiency of plasmid DNA entrapment than the other cationic liposomes tested. In the absence of serum, the cationic multilamellar vesicles (MLV) and small unilamellar vesicles (SUV) composed of TMAG, DOPE and DLPC gave highly efficient transfection. On the other hand, MLV, dehydration-rehydration vesicles (DRV), and SUV liposomes prepared with the mixtures of DC-Chol and DOPE showed similar levels of transfection efficiency. However, the cationic liposomes composed of DDAB and DOPE showed inferior efficiency, whether in the form of DRV, SUV or MLV. The transfection efficiency of lipofectin was also low. In the presence of serum, on the other hand, a considerable (about 30-50%) amount of transfection activity was still observed at 10% fetal calf serum in the cationic MLV and SUV composed of TMAG, DOPE and DLPC. Cationic MLV, composed of TMAG, DOPE and DLPC, can transfect plasmid DNA, not only in the adherent cell lines but also in the suspension cell lines. These findings indicate that the transfection efficiency of cationic liposomes is affected by the lipid composition, the type of liposome, or the presence or absence of serum. They also indicate that the cationic liposomes containing TMAG, DOPE and DLPC are efficient vectors for gene transfer into cells.

<P>Phospholipid vesicles, also known as liposomes, were examined for their ability to act as a drug carrier to the brain. 9-Amino-1,2,3,4-tetrahydroacridine (THA), a centrally acting acetylcholinesterase inhibitor, was used as a model drug. THA was encapsulated in dehydration-rehydration vesicles (DRV) composed of egg yolk phosphatidylcholine, cholesterol and dipalmitoyl-phosphatidic acid (molar ratio, 10/10/1) and injected into the heart of mice. The toxicity and side effects of THA were reduced by encapsulation in liposomes. The THA concentration in the mouse brain after injection of THA-encapsulated DRV at a dose of 2 mg/kg remained higher than that of free THA at the same dose. Effective concentration of THA in the brain was also prolonged by the use of liposomes, although accumulation of THA in the spleen and kidney was observed. We, therefore, concluded that liposomes are useful as carriers of drugs to the brain.</P>

Monoclonal antibodies were raised against urine proteins from diabetic patients. An antibody, YO-2, stained three protein bands with apparent molecular weights of 66, 49, and 36 kDa. These bands were not reactive with an anti-human albumin antibody. The urine levels of YO-2-reactive antigen in the normal control were 0.97 +/- 0.37 U/g-Cr (units per gram of urine creatinine) (mean +/- SD). Those of the normo-, micro-, and macroalbuminuric diabetic patients, respectively, were 1.38 +/- 1.36, 2.87 +/- 2.07, and 3.92 +/- 3.33 U/g-Cr. They were significantly higher in the micro- and macroalbuminuric patients. The urine levels of YO-2-reactive antigen had no significant correlation with the urine albumin levels and hemoglobin A1c. We concluded that; a) monoclonal antibody YO-2 recognized a non-albumin urine antigen increasingly excreted in diabetic patients with nephropathy, b) recent glycemic control of diabetes would not significantly affect the urinary excretion rate of YO-2-reactive antigen, and c) the excretion rate and probably the mechanism of YO-2-reactive protein differed from those of albumin. The urine levels of YO-2-reactive antigen could be a clinical marker of diabetic nephropathy.