Expanding the clinical relevance of the 5′-nucleotidase cN-II/NT5C2

Purine metabolism is depending on a large amount of enzymes to ensure cellular homeostasis. Among these enzymes, we have been interested in the 5′-nucleotidase cN-II and its role in cancer biology and in response of cancer cells to treatments. This protein has been cited and studied in a large number of papers published during the last decade for its involvement in non-cancerous pathologies such as hereditary spastic paraplegia, schizophrenia, and blood pressure regulation. Here, we review these articles in order to give an overview of the recently discovered clinical relevance of cN-II.     

5′-nucleotidase activity in cultured cell lines. Effect of different assay conditions and correlation with cell proliferation

Summary The 5′-AMPase activity of the ectoenzyme 5′-nucleotidase has been measured in a variety of cell lines, using intact cells. Human cell types showed two orders of magnitude higher enzyme activity than mouse cell lines. The ectoenzyme is inhibited by adenosine 5′-(α,β-methylene) diphosphate and Concanavalin A. A different extent of 5′-nucleotidase lectin inhibition was observed in the studied cell lines, suggesting that the corresponding ectoenzymes are glycoproteins with a different type or degree, or both, of glycosylation. The 5′-nucleotidase activity increased during subculture and decreased after cell transformation. Generally, the 5′-nucleotidase activity was two-to five-fold higher in monolayer than in suspension cell culture. A relation between cell growth and 5′-AMPase activity was also observed. Enzyme activity increased at the end of the lag phase (glioblastoma cells) or during the exponential phase (the other two cell lines). After confluence, the activity decreased to the initial or even lower range of activity. Observed activity variations with cell proliferation correlate with modifications of 5′-AMPase activity during subculture. This work was supported by grant no. PR84-0359 from the Comisión Asesora de Investigación Científica y Técnica (Spain).     

Comparison of 5′-nucleotidase activities of isolated plasma membranes of two ascites cell variants

• 1.1. The glycogen-containing ascites cell line was found to have a 3–5 times higher 5′-nucleotidase specific activity than the glycogen-free variant, resulting in different substrate affinity constants of K_m = 0.14mM and K_m = 0.69mM respectively.
• 2.2. These activity differences were due to true 5′-nucleotidase as shown by its inactivation through specific inhibitors such as concanavalin A and α,β-methylene adenosine diphosphate.
• 3.3. Substrate specificity of the enzyme was similar in both cell lines, but differences were observed with respect to the pH optimum and stability.

     

Effect of sodium deoxycholate on 5′-nucleotidase

The 5′-nucleotidase localized in rat liver plasma membranes was purified to a single protein, which contained phospholipid. The molecular weight and the sedimentation constant were about 150 000 and 7 S in the presence of sodium deoxycholate, while the enzyme protein was aggregated when the preparation was dialyzed thoroughly. The purified 5′-nucleotidase exhibited the same properties as the 5′-nuelcotidase in plasma membranes. The 5′-nucleotidase activity was increased by the addition of various bile salts or by the solubilization of membranes with trypsin, papain or phospholipase C. The solubilized and aggregated forms of the enzyme showed different substrate specificity for nucleotides, pH optimum, heat stability and $\text{K}{}_{\mathrm{<mtext>m</mtext>}}$. The purified enzyme catalyzed an exchange reaction between AMP and adenosine, which was diminished by the addition of sodium deoxycholate.     

Interactions of lectins with membrane glycoproteins effects of concanavalin A on 5′-nucleotidase

Concanavalin A causes a biphasic modification of the activity of the plasma membrane enzyme 5′-nucleotidase. The first stimulatory phase occurs from 0 to 0.05 μM concanavalin A, the second inhibitory phase at higher concentrations. The curve relating binding of ¹²⁵I-labelled concanavalin A and concentration of native lectin is similarly biphasie. The two phases likely result from occupation of distinct families of binding sites. When the enzyme is extracted from the membrane, the stimulatory phase disappears. Thus, the high affinity binding sites responsible for this phase depend upon the intact membrane structure while the others do not.     

Calf intestinal 5′-nucleotidase

The 5′-nucleotidase activity isolated from calf intestinal mucosa has been resolved into three fractions by DEAE-cellulose column chromatography. One of these fractions was obtained free of nonspecific phosphatase activity and was used for further studies. This enzyme has maximal activity between pH 6.0–6.5 in Tris-acetate buffer and at 8.0 in sodium diethyl barbiturate buffer. Complete inactivation of the enzyme occurs in the presence of 0.001 m EDTA. Manganese ions enhance the activity to the greatest extent although this metal ion can be essentially replaced with Co⁺⁺ or Mg⁺⁺. Further studies showed a high specificity for the hydrolysis of 5′-mononucleotides. K_m values determined at pH 8.0 were 5.0–5.8 × 10⁻⁵m for AMP and 5.0–8.0 × 10⁻⁵m for GMP. At pH 6.5 the K_m value for AMP is in the range 14–20 × 10⁻⁵m.     

Isolation and characterization of the ecto-5′-nucleotidase from a rat glioblastoma cell line

Summary 5-Nucleotidase has been purified from rat glioblastoma cells (Rugli cells). The enzyme has been solubilized from plasma membranes by using Triton X-100 and CHAPS. Two affinity chromatographies on concanavalin A and 5-AMP-Sepharose render the purified enzyme with a high specific activity (76.36 mol AMP-min^–1-mg^–1). The purified enzyme gives a single polypeptide band on SDS-PAGE with an apparent molecular mass of 74 kDa. Active forms with an apparent molecular mass of 135 kDa and 268 kDa are observed when the purified enzyme is analyzed by gel filtration in the presence of either 0.6% sodium deoxycholate or 0.1% Triton X-100, respectively. The purified 5-nucleotidase presents optimum activity at pH 7.8–8.1 either in the presence or in the absence of Me²⁺. A linear Arrhenius plot is observed in the 25–46° C temperature range and an activation energy of 33.7 KJ/mol is calculated. The enzyme is inhibited by EDTA; the activity is partially restored by different divalent cations as Zn²⁺, Mn²⁺, and Co²⁺. The hydrolysis of nucleosides 5-monophosphate shows Michaelis kinetic. The enzyme is inhibited by nucleosides di- and triphosphate. 5-Nucleotidase is a glycoprotein, being its activity inhibited at different extent by various lectins.     

Properties of 5′-nucleotidase in rat heart sarcolemma

The activity of 5′-nucleotidase (5′-ribonucleotide phosphohydrolase, EC 3.1.3.5) was examined in membrane fractions isolated by hypotonic shock-LiBr treatment (fraction HL) and sucrose gradient separation (fraction S) of rat ventricle homogenate. The enzyme activity in these two fractions differed significantly in several respects. In fraction HL, 5′-nucleotidase had a high affinity for AMP (K_m 35 μM), and ATP was a potent competitive inhibitor. In contrast, the 5′-nucleotidase displayed by fraction S showed a low substrate affinity (K_m 130 μM) and less sensitivity to ATP. Treatment of membranes with trypsin and neuraminidase markedly stimulated 5′-nucleotidase in fraction HL, whereas only a modest effect was observed in fraction S. Exposure of the membranes to Triton X-100 resulted in a 60% and 10% increase in the enzyme activity in fractions HL and S, respectively. The characteristic activity ratios of 5′-nucleotidase at 200 μM relative to 50 μM AMP in fractions HL and S were modified by alamethicin in an opposite way and became identical. Although concanavalin A almost completely inhibited the 5′-nucleotidase activity in both membrane preparations at a concentration of 2 μM, Hill plots of the data on concanavalin A inhibition revealed a coefficient of 2.2 for fraction S and 1.1 for fraction HL. The differences in 5′-nucleotidase activity of the two membrane fractions are considered to be due to differences in the orientation of the vesicles of the sarcolemmal preparations. These results suggest that two distinct catalytic sites for 5′-nucleotidase are present at the intra and extracellular surface of the rat heart sarcolemma.     

The role of ecto-5′-nucleotidase/CD73 in glioma cell line proliferation

The antioxidant activity of β-carotene and oxygenated carotenoids lutein, canthaxanthin, and astaxanthin was investigated during spontaneous and peroxyl-radical-induced cholesterol oxidation. Cholesterol oxidation, measured as generation of 7-keto-cholesterol (7-KC), was evaluated in a heterogeneous solution with cholesterol, AAPH, and carotenoids solubilized in tetrahydrofuran and in water, and in a homogeneous solution of chlorobenzene, with AIBN as a prooxidant. The formation of 7-KC was dependent on temperature and on cholesterol and prooxidant concentrations. All the carotenoids tested, exhibited significant antioxidant activity by inhibiting spontaneous, AAPH- and AIBN-induced formation of 7-KC, although the overall order of efficacy of these compounds was astaxanthin > canthaxanthin > lutein = β-carotene. The finding that carotenoids exert protective effects on spontaneous and free radical-induced cholesterol oxidation may have important beneficial effects on human health, by limiting the formation of atheroma and by inhibiting cholesterol oxidation in food processing or storage.     

Treatment of ovarian cancer cell lines with 5-aza-2′-deoxycytidine upregulates the expression of cancer-testis antigens and class I major histocompatibility complex-encoded molecules

Purpose  To test the hypothesis that decrease in DNA methylation will increase the expression of cancer-testis antigens (CTA) and class I major histocompatibility complex (MHC)-encoded molecules by ovarian cancer cells, and thus increase the ability of these cells to be recognized by antigen-reactive CD8⁺ T cells. Methods  Human ovarian cancer cell lines were cultured in the presence or absence of varying concentrations of the DNA demethylating agent 5-aza-2′-deoxycytidine (DAC) for 3–7 days. The expression levels of 12 CTA genes were measured using the polymerase chain reaction. The protein expression levels of class I MHC molecules and MAGE-A1 were measured by flow cytometry. T cell reactivity was determined using interferon-γ ELISpot analysis. Results  DAC treatment of ovarian cancer cell lines increased the expression of 11 of 12 CTA genes tested including MAGE-A1, MAGE-A3, MAGE-A4, MAGE-A6, MAGE-A10, MAGE-A12, NY-ESO-1, TAG-1, TAG-2a, TAG-2b, and TAG-2c. In contrast, DAC treatment decreased the already low expression of the MAGE-A2 gene by ovarian cancer cells, a finding not previously observed in cancers of any histological type. DAC treatment increases the expression of class I MHC molecules by the cells. These effects were time-dependent over a 7-day interval, and were dose-dependent up to 1–3 μM for CTA and up to 10 μM for class I MHC molecules. Each cell line tested had a unique pattern of gene upregulation after exposure to DAC. The enhanced expression levels increased the recognition of 2 of 3 antigens recognized by antigen-reactive CD8⁺ T cells. Conclusions  These results demonstrate the potential utility of combining DAC therapy with vaccine therapy in an attempt to induce the expression of antigens targeted by the vaccine, but they also demonstrate that care must be taken to target inducible antigens. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Implications of a 5′-nucleotidase inhibitor in human leukemic cells for cellular aging and cancer

5′-Nucleotidase activity of normal human embryonic lung fibroblasts (IMR-90) was found to be inhibited by the homogenates of seven different cell lines originated from patients with different kinds of leukemia and of fresh lymphocytes from a patient with Sezary syndrome (circulating T-cell lymphoma). About 97% of the inhibiting activity was found in the soluble fraction of RPMI 8402 cells, a cell line originated from the lymphocytes of a patient with acute lymphocytic leukemia. This inhibiting activity was not destroyed by dialysis, heating at 56°C for 30 min, nor digestion with RNAase or DNAase. About 85% of the inhibiting activity was destroyed by digestion with papain at 37°C for 1 h and it was destroyed completely by heating at 100°C for 30 min. When the heated (56°C for 30 min) soluble fraction of RPMI 8402 cells was mixed with the homogenate of IMR-90 cells, it had no effect on the activities of alkaline, neutral or acid phosphatases, nor of $\text{N-}\text{acetyl}\text{-β-}\text{d}\text{-glucosaminidase}$ or cytochrome c oxidase of IMR-90 cells. Preincubating the mixed samples for 1, 20 and 45 min, respectively, before adding the substrate, the heated soluble fraction of RPMI 8402 cells did not increase the percentage of inhibition for 5′-nucleotidase of the homogenate of IMR-90 cells. No inhibition of other enzyme activities was observed under similar conditions. These data suggest that the inhibiting activity is due to a protein(s) that is not a protease. The inhibiting activity was found in a single peak after the soluble fraction was fractionated by Sephadex G-100 chromatography and sedimentation centrifugation. The molecular weight of the inhibitor was found to be approx. 35 000 by comparing its retention volume and sedimentation rate with those of proteins of known molecular weight. The present study suggests that the previously reported undetectability of 5′-nucleotidase in permanent cell lines could be due to the presence of a protein inhibitor for 5′-nucleotidase in these human leukemic cell lines. It also supports the hypothesis that the increased 5′-nucleotidase activity in normal senescent cells in vitro may be a control in cellular aging that is missing from leukemic cells in vitro.     

Absence of 5′-nucleotidase in porcine polymorphonuclear leucocyte membranes

A fraction enriched in plasma membranes from porcine polymorphonuclear leucocytes, isolated by sucrose density centrifugation was shown to possess considerable AMP hydrolysing activity (150 nmol/min per mg protein). However all of this activity could be inhibited using excess $\text{p-}\text{nitrophenyl}$ phosphate in the incubation medium. Furthermore the hydrolysis of AMP by the membrane was unaffected by the 5′-nucleotidase inhibitor α,β-methyleneadenosine diphosphate and by the lectin concanavalin A, another potent inhibitor of 5′-nucleotidase. An antibody against mouse liver 5′-nucleotidase also did not inhibit the activity. These results suggest that the hydrolysis of AMP by porcine polymorph membranes is not accomplished by a specific 5′-nucleotidase and the necessity for distinguishing between true 5′-nucleotidase and non-specific phosphatase activity is discussed.     

Assignment of ecto-5′-nucleotidase to human chromosome 6

Summary Ecto-5-nucleotidase activity (5NT) was measured on whole cells of 26 human x Chinese hamster hybrids. Concordance analysis showed 100% correlation between enzyme activity and inheritance of human chromosome 6. This observation was confirmed by a segregation analysis in which cells of a hybrid containing chromosome 6 were stained by indirect immunofluorescence for HLA Class 1 antigen and sorted by a fluorescence-activated cell sorter (FACS). Cells in the HLA^- compartment were cloned and expression of HLA and 5NT was determined. Of nine clones, three were HLA^-, 5NT^- and six were HLA⁺, 5NT⁺, supporting the linkage of 5NT to chromosome 6.     

Heterogeneity in 5′-nucleotidase activity of mouse peritoneal macrophages

Summary After stimulation of the mouse peritoneal cavity with newborn calf serum (NBCS), four types of monocyte and macrophage were distinguished on the basis of peroxidase (PO) patterns. Cytochemically, these cells showed strong heterogeneity in 5-nucleotidase (5N) activity. Monocytes and monocyte-derived macrophages with PO activity in granules lacked 5N activity. Resident macrophages (with PO activity in RER and nuclear envelope) generally had significant 5N activity on the plasma membrane, the pattern showing close correlation with the biochemical findings. The group of PO-negative macrophages comprised both 5N-negative and 5N-positive cells. These findings suggest two possibilities, i.e., that monocytes (5N-)transform via PO-negative cells (5N-/+) into resident macrophages (5N+), or that the monocytes and monocyte-derived macrophages and the resident macrophages represent separate lineages. The fourth type of macrophage, the exudate-resident cell (with PO activity both in granules and in the RER and nuclear envelope), occurred only in low numbers and very late after NBCS stimulation, and is therefore considered not to be a transitional cell between monocytes and resident macrophages.     

Localization of 5′-nucleotidase activity in the parotid acinar cells of a rat treated with isoproterenol

Summary Cytochemical localization of 5- nucleotidase (AMPase) has been investigated in the parotid acinar cells of rats at various stages of exocytic secretion induced by an administration of isoproterenol (IPR).In the resting stage, the acinar cells show AMPase activity located on the baso-lateral and luminal plasmalemma, and in the earliest secretory stage the luminal plasma membranes are devoid of the enzymatic activity. However, these particular regions exhibit AMPase activity during the advanced stages of secretion, and the AMPase positive membranes become absorbed into the cytoplasm by endocytic activity. The absorbed membrane components then seem to be degraded by the action of lysosomes.The intracellular fate of the endocytic vacuoles has been examined by the aid of ferritin particles introduced retrogradely through ductal lumina. Ferritin containing vacuoles are distributed in the cytoplasm, and these droplets change into secondary lysosomes. No tracer particles are recognized in the internal space of the Golgi lamella and its associated vesicles.The results suggested that in the exocytic secretion of parotid acinar cells, AMPase originating from plasma membrane intermingles with the membranes derived from secretion granules, and is translocated into cytoplasm by an endocytic mechanism. The internalized membrane components are, at least partly, degraded by lysosome action.