Carbohydrate-mediated recognition of a circulating placental alkaline phosphatase-immunoglobulin M complex

Partial breakdown of glycated alkaline phosphatases mediated by reactive oxygen species

The lower levels of serum alkaline phosphatase (AP) activity found in patients with diabetes mellitus apparently originate from the selective disappearance or decrease in bone AP activity in the circulation. Hence, we investigated in vitro the effect of glycation on the activities of five AP isozymes. Aseptic incubation with 25 mmol/L of D-glucose and APs rapidly reduced bone and placental AP activities before those of liver, kidney and intestinal enzymes. The resulting bone and placental AP molecules were clearly glycated, according to the result of aminophenylboronic acid affinity chromatography. Furthermore, Western blotting analysis revealed that the placental AP molecule was fragmented, and its partial cleavage took place at Ala154 on the AP molecule. Since glycation of serum proteins causes the generation of reactive oxygen species, the effects of reactive oxygen species on placental AP activity were assayed, and the results indicated that hydroxyl radicals might be a major factor for the specific inactivation of AP activities. The reduction in AP activity by incubation with glucose in vitro was reversed by the further addition of catalase. Furthermore, ferrous ion with hydrogen peroxide, which generates hydroxyl radicals, had an inhibitory effect on AP activities. These findings suggest that the reduced AP activity in diabetic patients might result from partial cleavage of the bone AP molecule by reactive oxygen species induced by glycation.

Properties of high-molecular-mass placental alkaline phosphatases in normal pregnancy sera

We examined the appearance of high-molecular-mass placental alkaline phosphatases (HPLAPs) in the serum of normal pregnant women by means of polyacrylamide gel electrophoresis (PAGE) in the presence of Triton X-100. The HPLAPs were undetectable or only slightly detectable by PAGE in the absence of Triton X-100. The HPLAPs were detected in all sera sampled during the last trimester of pregnancy. The catalytic activities of total placental alkaline phosphatase (TPLAP) and HPLAPs were correlated (r = 0.96) and the ratio of HPLAPs/TPLAP catalytic activity was 0.20 (0.06) (mean and SD) in 40 serum samples from pregnant women. The HPLAPs appear to be formed from a common dimeric placental alkaline phosphatase (PLAP) (common-PLAP), as judged by the fact that they were formed again after removal of HPLAPs from serum by gel filtration. The formation of HPLAPs was more prominently observed with the faster fractions of gel filtration. The apparent molecular mass of the HPLAPs in pregnancy serum was 720 KDa by gel filtration. HPLAPs were not converted to common-PLAP by phosphatidylinositol-specific phospholipase (PIPL) C and PIPL-D treatments. The HPLAPs were selectively incorporated into liposomes consisting of phosphatidylcholine/cholesterol, and most of the PIPL-D-treated PLAP could from HPLAPs, while a small amount of PLAP could not form HPLAPs. On the other hand, HPLAPs in pregnant women's sera and HPLAPs prepared from partially purified PLAP in vitro could be converted to common-PLAP by brief treatment with subtilisin. However, the highly purified PLAP could not form HPLAPs in the presence of Triton X-100. These results suggest that PIPL-D-resistant and PLAP-associated serum protein may regulate the conversion of PLAP to HPLAP in the presence of Triton X-100.

Intestinal alkaline phosphatase isoforms in rabbit tissues differ in glycosylation patterns

OBJECTIVES

In patients with advanced liver cirrhosis or chronic nephritis, an intestinal alkaline phosphatase (IAP)-like enzyme is ectopically expressed in the liver or kidney. In this study, we used rabbit organs as a human pathological model, because the rabbit liver or kidney expresses an IAP-like enzyme as the predominant isozyme, unlike humans.

METHODS

IAP and the IAP-like enzyme were purified from rabbit intestine and kidney, respectively, by immunoaffinity chromatography using a monoclonal antihuman IAP antibody. Some properties of the IAP and IAP-like enzyme expressed in rabbit organs are compared.

RESULTS

Some of their catalytic and physicochemical properties differed. In particular, the net charge, molecular mass, and hydrophobicity of IAP from rabbit intestine was slightly different from the IAP-like enzyme from rabbit kidney. There was a difference in the sugar chain structure between the two enzymes according to the results of lectin affinity chromatography, and part of the peptide maps differed slightly. However, there was no difference in the peptide maps after treatment with endo-N-acetylglucosaminidase F.

CONCLUSIONS

The primary structures of IAP and the IAP-like enzyme are basically similar, except for the glycosylation process of the respective AP isozymes.

Human tissue non-specific alkaline phosphatases: sugar-moiety-induced enzymic and antigenic modulations and genetic aspects

To investigate the possible role(s) of glycans in human tissue non-specific alkaline phosphatase (TNAP) activity, the iso-enzymes were purified and treated with various exo- and endo-glycosidases. Catalytic activity, oligomerization, conformation and immunoreactivity of the modified TNAPs were evaluated. All TNAPs proved to be N-glycosylated, and only the liver isoform (LAP) is not O-glycosylated. Usually, the kidney (KAP) and bone (BAP) isoenzymes are similar and cannot be clearly discriminated. Differences between the immunoreactivity of KAP/BAP and LAP with a BAP antibody were mainly attributed to the N-glycosylated moieties of the TNAPs. In addition, elimination of O-glycosylations moderately affects the TNAP reactivity. Interestingly, N-glycosylation is absolutely essential for TNAP activity, but not for that of the placental or intestinal enzymes. According to the deduced amino acid sequence of TNAP cDNA, Asn-213 is a possible N-glycosylation site, and our present findings suggest that this sugar chain plays a key role in enzyme regulation. With regard to the oligomeric state of alkaline phosphatase (AP) isoforms, the dimer/tetramer equilibrium is dependent on the deglycosylation of glycosyl-phosphatidylinositol(GPI)-free APs, but not GPI-linked APs. This equilibrium does not affect the AP conformation as observed with CD. With regard to TNAPs, no data were available on the gene expression or nature of the 5'-non-translated leader exon of human KAP, as opposed to BAP and LAP genes. cDNA sequencing revealed that cortex/medulla KAP is genetically related to BAP, and medulla KAP to LAP.