The activity of membrane-bound alkaline phosphatase (ALP) expressed on the external surface of cultured murine P19 teratocarcinoma and human HL-60 myeloblastic leukemia cells was studied at physiological pH using p-nitrophenylphosphate (pNPP) as substrate. The rate of substrate hydrolysis catalyzed by intact viable cells remained constant for eight successive incubations of 30 min and was optimal at micromolar substrate concentrations over the pH range 7.4-8.5. The value of apparent K(m) for pNPP in P19 and HL-60 cells was 120 microM. Hydrolytic activity of the ecto-enzyme at physiological pH decreased by the addition of levamisole, a specific and noncompetitive inhibitor of ALP (K(i) P19 = 57 microM; K(i) HL-60 = 50 microM). Inhibition of hydrolysis was reversed by removal of levamisole within 30 min. Retinoic acid (RA), which promotes the differentiation of P19 and HL-60 cells, induced levamisole-sensitive ecto-phosphohydrolase activity at pH 7.4. After its autophosphorylation by ecto-kinase activity, a 98-kDa membrane protein in P19 cells was found to be sensitive to ecto-ALP, and protein dephosphorylation increased after incubation of cells with RA for 24 h and 48 h. Orthovanadate, an inhibitor of all phosphatase activities, blocked the levamisole-sensitive dephosphorylation of the membrane phosphoproteins, while (R)-(-)-epinephrine reversed the effect by complexation of the inhibitor. The results demonstrate that the levamisole-sensitive phosphohydrolase activity on the cell surface is consistent with ecto-ALP activity degrading both physiological concentrations of exogenously added substrate and endogenous surface phosphoproteins under physiological pH conditions. The dephosphorylating properties of ecto-ALP are induced by RA, suggesting a specific function in differentiating P19 teratocarcinoma and HL-60 myeloblastic leukemia cells.

Alkaline phosphatase (ALP) is a glycoenzyme that is highly expressed during carcinogenesis and is induced by retinoic acid (RA) in various cells. We investigated the effects of RA on N-linked glycosylation of the tissue nonspecific liver/bone/kidney- type of ALP (L/B/K ALP), on ALP transcripts, and on total protein glycosylation in two neuronal cell lines, P19 and NG108CC15, and in primary cultures of neonatal rat brain. ALP activity was determined in cell extracts and found to be induced by RA. Tunicamycin was used at various concentrations to inhibit protein N-glycosylation. After treatment of cells with low concentrations (0.1 and 0.125 microgram/ml) of tunicamycin for 48 h, uninduced and RA-induced ALP activity declined while incubation with a protease inhibitor restored activity, indicating that the L/B/K ALP bear N-linked oligosaccharide chains important for maintaining enzymatic activity. Interestingly, ALP activity in RA-treated cultures was less inhibited by tunicamycin compared to untreated controls suggesting that RA may have an impact on ALP N-glycosylation. To investigate effects of RA on ALP glycosylation further, incorporation of [(14)C]mannose and [(35)S]methionine into ALP protein was determined in the presence or absence of RA. The ratio of mannosylation and biosynthesis demonstrate that incubation of cells with RA increased [(14)C]mannose incorporation into ALP molecules. Also, the release of free [(14)C]mannose from ALP molecules relative to the amount of protein by N-Glycanase was increased in RA-treated cultures. In addition, mannosylation of total protein was found to be induced in cells after exposure to RA. Analysis of biosynthesized ALP monomers revealed that RA increased glycosylation of the polypeptides. Furthermore, tunicamycin decreased the stability of ALP mRNA, an effect that was reduced by cotreatment with RA. Thus, the degree of N-glycosylation of the L/B/K ALP as well as mRNA and protein levels of this enzyme are affected by RA. The P19 cell line provides a useful model system to study the molecular mechanism(s) underlying the action of RA on glycosylation during neuronal differentiation further.

Several known and some new retinoids were synthesized and their in vivo activity was investigated by an assay, based on induction of alkaline phosphatase in P19 teratocarcinoma cells, human prostate carcinoma cells and primary cultures of neonatal rat heart cells. The assay used in this study was found to be reproducible and useful for rapid screening of retinoids for biological activity. Two newly synthesized compounds exhibit high biological activity. The biological potency of the compounds was compared to their ability to bind to recombinant retinoic-acid receptor a and to cellular retinoic-acid-binding protein I determined by Charsorb-binding assay. mRNA of both retinoic-acid-binding proteins could be detected in the three cell lines investigated. As expected from the number of different retinoic-acid receptors, the results suggest that retinoids do not need to bind retinoic-acid receptor a nor cellular retinoic-acid-binding protein I in order to exhibit biological activity, but most retinoids investigated show a clear correlation between binding to these proteins and their biological activity.

UDP-N-acetylglucosamine:dolichyl-phosphate N-acetylglucosamine-1-phosphate transferase (GPT) is the first enzyme in the dolichol pathway of protein N-glycosylation, and is implicated in the developmental programmes of a variety of eukaryotes. In the present study we describe the effects of all-trans-retinoic acid (RA) on the levels of GPT protein and enzymic activity, and on the transcription rate of the GPT gene, in mouse P19 teratocarcinoma cells. RA caused a dose-dependent and protein-synthesis-dependent induction of enzyme activity. The maximum induction of GPT activity (about 3-fold) required 2 days of exposure to 1 microM RA. Induced GPT activity also resulted in an increase in the rate of incorporation of [3H]mannose into Glc3Man9GlcNAc2. Enzymic activities paralleled GPT gene expression. The GPT gene was induced (2-fold) after 7 h of RA treatment. An approx. 3-fold increase in a 48 kDa GPT protein and approx. 4-fold increases in the levels of three GPT transcripts (1.8, 2.0 and 2.2 kb) were observed after 2 days of RA treatment. The enhanced levels of GPT protein and mRNAs began to decline 3 days after the initiation of differentiation, and GPT expression was down-regulated during cellular differentiation. GPT activity decreased about 2. 8-fold to a constant level in differentiated P19 cells. The results indicate that the RA-induced enzyme activity was mainly determined by increased transcription of the GPT gene. RA-treated P19 cells were about 4-fold more resistant to tunicamycin, a fungal antibiotic which inhibits GPT, than were control cells. In addition, GPT activity in membranes from RA-treated P19 cells exhibited approx. 4-fold increased resistance to tunicamycin compared with activity in membranes from untreated control cells, demonstrating that resistance to tunicamycin is correlated with induced GPT activity. Furthermore, increased GPT activity had regulatory significance with regard to the rate of incorporation of [3H]mannose into Glc3Man9GlcNAc2-P-P-dolichol and into glycoproteins. Together, the data provide additional insights into the hormonal regulation of GPT and present evidence that the RA-mediated induction of GPT has a regulatory impact on the dolichol pathway.