Inhibition of 5′-nucleotidase by concanavalin A: Evidence for localization on the outer surface of the plasma membrane

Summary 5-Nucleotidase activity of rat liver plasma membrane is markedly inhibited by concanavalin A. Taken together with a unilateral pattern of labelling of concanavalin A binding sites with hemocyanin, the results indicate that an allosteric site of the enzyme is at the outer surface of the membrane.Work supported in part by NIH CA 13145 to D.J. Morré and by U.S. NASA W-12792(05) to J. Shen-Miller. Argonne National Laboratory is operated under the auspices of the U.S. Energy Research and Development Administration. We thank Keri Safranski, Dorothy Werderitish, and G.T. Chubb for technical assistance     

Purification and some properties of cytosol 5′-nucleotidase from rat liver

A 5′-nucleotidase (5′-ribonucleotide phosphohydrolase, EC 3.1.3.5) was highly purified from rat liver. The preparation appeared homogeneous on the criteria of disc-gel electrophoresis.A pH optimum at about 6.5 was observed for all substrates tested. The activity of this enzyme was absolutely dependent on the presence of various bivalent metal salts. The highest V value was attained with MgCl₂ and the concentration at half-enzyme saturation was lowest with MnCl₂. The enzyme had markedly higher affinities for IMP, dIMP, GMP and dGMP than the other 5′-mononucleotides, although V values for all the substrates tested were in the same order of magnitude.The activity of this enzyme was stimulated by various alkali metal salts, some carboxylic acids and adenine nucleotides. When AMP was used as substrate, the substrate-velocity plot was sigmoidal and NaCl, Tris-maleate and ATP stimulated the enzyme by decreasing the sigmoidicity of the plot. When IMP was used as substrate, the substrate-velocity plot was hyperbolic and these three activators stimulated the enzyme by increasing the V and decreasing the K_m value.Some of these results provided consistent evidence for the identity of this enzyme and the cytosol 5′-nucleotidase, the presence of which had been reported in crude preparations from rat liver.     

Distribution of ecto-5′-nucleotidase in the rat liver: effect of anaemia

In the kidney a striking parallel exists between the expression of ecto-5-nucleotidase and of erythropoietin by renal fibroblasts. It was therefore hypothesized that the expression of ecto-5-nucleotidase in fibroblasts might be controlled by oxygen tension. In order to test this hypothesis, we examined the distribution of the enzyme in a tissue which displays a defined zonation in respect to oxygen tension, namely in the liver; anaemia was used in order to exaggerate this zonation. The distribution of ecto-5-nucleotidase was investigated by light and electron microscopy using enzyme and immunohistochemical methods in the livers of healthy and of anaemic rats. Anaemia was produced by haemolysis combined with X-ray irradiation. The enzyme was detected in the bile canaliculi, in the connective tissue of the portal triads and of the central veins, and in fat-storing cells probably corresponding to a special form of fibroblasts. In healthy animals the perisinusoidal ecto-5-nucleotidase activity was slightly higher in the pericentral than in the periportal area of the acinus whereas the inverse was observed for the staining of bile canaliculi. Anaemia provoked an increase of ecto-5-nucleotidase in fat-storing cells in the pericentral zone of the acinus and in fibroblasts around the central veins, resulting in steepended gradients along the sinusoids. The intralobular gradient of ecto-5-nucleotidase in perisinusoidal cells and the effect thereon of anaemia suggest that the expression of the ecto-5-nucleotidase might be directly or indirectly controlled by local oxygen tension.     

Insulin-dependent release of 5′-nucleotidase and alkaline phosphatase from liver plasma membranes

Insulin treatment of isolated liver plasma membranes induced the release of 5′-nucleotidase and alkaline phosphatase. This effect was maximal at physiological hormone concentrations, being 36% and 17% for 5′-nucleotidase and alkaline phosphatase respectively, and was fully mimicked by the phosphatidylinositol specific phospholipase C (PI-PLC), thus confirming the presence of a glycosyl-phosphatidylinositol anchoring-system for these exofacial enzymatic proteins. The complete inhibition of insulin dependent enzyme release by neomycin is strongly supportive of an involvement of membrane-located PI-PLC activity. In addition, the insulin-like effect on enzyme release induced by the GTP non-hydrolysable analog, GTP-γ-S, and its sensitivity to the pertussis toxin are in favour of a mediatory role exerted by the G proteins system, in the transduction of some actions of insulin.     

Lectin-induced inhibition of plasma membrane 5′-nucleotidase: Sensitivity of purified enzyme

To determine whether the lectin-induced inhibition of plasma membrane 5′-nucleotidase resulted from direct interaction of the lectin with the enzyme or indirectly from a membranous change due to lectin binding to other membrane glycoproteins, the enzyme was purified and its sensitivity tested in the absence of other membrane components. A 5000 fold purification was achieved by solubilization in Lubrol PX followed by gel filtration (Sephadex G-100), anion exchange (DEAE-Biogel A) and selective adsorption (hydroxylapatite) chromatography. The purified enzyme was even more sensitive to inhibition by high concentrations of concanavalin A, wheat germ agglutinin or Rincinus communis agglutinin than was the membrane-bound enzyme indicating that inhibition is due to direct binding of the lectins to the glycoprotein enzyme itself. Divalent succinyl Con A inhibited neither form of the enzyme suggesting the need for crosslinking for inhibition by the native lectin. The purified enzyme could not be activated by low concentrations of lectins which stimulated the membrane bound enzyme.     

Properties of a 5′-nucleotidase purified from mouse liver plasma membranes

1. Extraction of a mouse liver plasma-membrane fraction with a detergent buffer, N-dodecylsarcosinate-Tris buffer (sarcosyl-Tris buffer), solubilized 90% of the protein and 70% of the 5'-nucleotidase activity. 2. The proteins of the sarcosyl-Tris buffer extract were fractionated by a rate-zonal centrifugation in a sucrose-detergent gradient. The major protein peak sedimented ahead of phospholipids, which mainly remained in the overlay. Glycoproteins were separated ahead of the protein peak. 3. The 5'-nucleotidase activity peak was associated with 5% of the protein applied to the gradient, and contained relatively few protein bands. 4. The 5'-nucleotidase was purified further by gel filtration on Sepharose and Sephadex columns equilibrated with sarcosyl-Tris buffer, to give a single glycoprotein band on sodium dodecyl sulphate-polyacrylamide-gel electrophoresis. The purified enzyme was lipid-free. 5. Electrophoresis in polyacrylamide gels in sarcosyl-Tris buffers showed that the enzymic activity was coincident with the protein band. 6. The molecular weight suggested for the enzyme activity by gel filtration or centrifugation in sucrose gradients was 140000-150000. Sometimes, a minor enzyme peak of lower molecular weight was obtained. 7. Polyacrylamide-gel electrophoresis in sodium dodecyl sulphate indicated that as the polyacrylamide concentration was increased from 5 to 15%, the apparent molecular weight of the enzyme decreased from 130000 to 90000. 8. The evidence that 5'-nucleotidase is composed of two active and similar, if not identical, glycoprotein subunits and the role of detergent in effecting the separation of membrane proteins and glycoproteins are discussed. 9. Substrate requirements, pH optima and the nature of inhibition by an analogue of adenosine diphosphate are reported.     

Distribution of 5′-nucleotidase in the renal interstitium of the rat

Summary The hydrolysis of 5-AMP by 5-nucleotidase is the main source of adenosine. In various tissues adenosine is a local mediator adjusting the organ work to the available energy. In the kidney it regulates renal hemodynamics, glomerular filtration rate and renin release via specific receptors of the arteriolar walls. By immunocytochemistry we identified interstitial and tubular sites of 5-nucleotidase in the rat kidney. In the interstitium the enzyme was detected only in the cortical labyrinth, the compartment that comprises all arteriolar vessels besides other putative targets of adenosine. The 5-nucleotidase-positive cells of the interstitium were identified as fibroblasts. The fibroblasts are in close contact with the tubules as well as with the vessels. Thus, any 5-AMP released by the tubules into the interstitial space would be converted to adenosine in the direct vicinity of its assumed targets. Adenosine produced by tubular cells would hardly have access to its known targets, since 5-nucleotidase is restricted to the luminal cell surface. Pathological events affecting the fibroblasts might influence renal function by modifying the interstitial adenosine production.     

Stimulation of (Na+K+)-ATPase of rat liver plasma membrane by amino acids

The effect of twelve l-amino acids on the activity of liver plasma membrane (Na⁺K⁺)-ATPase has been tested. Histidine and arginine significantly enhanced the activity. The activtion by histidine showed saturation kinetics with an apparent K_a of about 8 mM, and was evident over a wide range of Na⁺ concentrations. The same amino acid did not significantly affect the Mg²⁺-dependent ATPase activity.     

Spin-label studies on rat liver and heart plasma membranes: do probe-probe interactions interfere with the measurement of membrane properties?

The structures of purified rat liver and heart plasma membranes were studied with the 5-nitroxide stearic acid spin probe, I(12,3). ESR spectra were recorded with a 50 gauss field sweep, and also with a new technique which "expands" the spectrum by (1) recording pairs of adjoining peaks with a smaller field sweep and (2) superposing the common peaks. The hyperfine splittings measured from the "expanded" spectra were significantly more precise than those obtained from the "unexpanded" spectra. Both procedures were used to study the effects of various I(12,3) probe concentrations on the spectra of liver and heart membranes, as well as the effects of temperature and CaCl2 additions on the spectra of liver membranes, and revealed the following: The polarity-corrected order parameters of liver (31 degrees) and heart (22 degrees) membranes were found to be independent of the probe concentration, if experimentally-determined low I(12,3)/lipid ratios were employed. The absence of obvious radical-interaction broadening in the unexpanded spectra indicated that "intrinsic" membrane properties may be measured at these low probe/lipid ratios. Here, "intrinsic" properties are defined as those which are measured when probe-probe interactions are negligible, and do not refer to membrane behavior in the absence of a perturbing spin label. At higher I(12,3)/lipid ratios, the order parameters of liver and heart membranes were found to substantially decrease with increasing probe concentration. The increase in the "apparent" fluidity of both membrane systems is attributed to enhanced radical interactions; however, an examination of these spectra (without reference to "low" probe concentration spectra) might incorrectly suggest that radical interactions were absent. For the membrane concentrations employed in these studies, the presence of "liquid-lines" (or "fluid components") in the unexpanded ESR spectra was a convenient marker of high probe concentrations. A thermotropic phase separation was observed in liver membranes between 19 degrees and 28 degrees. Addition of CaCl2 to liver plasma membrane [labelled with "low" I(12,3) concentrations] increased the rigidity of the membrane at 31 degrees and 37 degrees, without inducing a segregation of the probe in the bilayer. Previously reported data are discussed in relation to these results, and suggested minimal criteria for performing membrane spin label studies are included.     

Ecto-5′-nucleotidase is expressed by pericytes and fibroblasts in the rat heart

Ecto-5-nucleotidase is anchored at the outer surface of cell membranes and thus its reaction product adenosine is released into the extracellular space. Extracellular adenosine displays via specific receptors a wide range of physiological effects in heart. There are discrepancies in the literature concerning the distribution of ecto-5-nucleotidase in heart. Since we suspected that these may be due to technical problems, in the present study on ecto-5-nucleotidase in rat heart we attempted to circumvent some technical pitfalls. Good preservation of the tissue with open capillary lumina, providing a clear identification of endothelium, was obtained by perfusion fixation. At the light microscopic level, the distribution of ecto-5-nucleotidase studied by enzyme histochemistry and immunohistochemistry using a monoclonal and a polyclonal antibody yielded congruent results. The enzyme was rather homogeneously distributed throughout the myocardium, with a slightly higher incidence of stained cells in the outer thirds than in the inner third of the wall. Consistently high levels of ecto-5-nucleotidase were seen only in interstitial cells. The walls of large vessels and heart muscle cells were constantly negative for ecto-5-nucleotidase. The endothelia of capillaries were mostly negative but a few profiles occasionally displayed a weak immunoreaction. The interstitial cells staining positive for ecto-5-nucleotidase could be identified as pericytes and as fibroblasts according to their shapes and localizations. The immunoreactivity of fibroblasts was confirmed by electron microscopy. These data indicate that adenosine may be formed extracellularly in the interstitium of the myocardium, where it would have direct access to important targets such as myocytes, arterioles and nerve endings.     

Diurnal variation in 5′-nucleotidase activity in rat liver

Summary The diurnal variation of 5-nucleotidase activity in periportal and pericentral areas of rat liver parenchyma has been determined with quantitative histochemical means. 5-Nucleotidase activity was estimated using microdensitometry in cryostat sections after being incubated with a medium according to Wachstein and Meisel (1957). It appeared that 5-nucleotidase activity was significantly higher in pericentral areas than in periportal areas throughout the daily cycle and showed a maximum at the end of the light period. It was concluded that 5-nucleotidase activity may be related with the capacity to diminish messenger RNA resulting in protein breakdown.     

Cytochemical studies on the localization of 5′-nucleotidase in the acinar cells of the rat salivary glands

Summary The cytochemical localization of 5-nucleotidase (5-AMPase), and its validity, were investigated in parotid and submandibular acinar cells of a rat. Biochemical determinations showed that adequate treatment with glutaraldehyde could minimize the loss of enzymatic activity, and that 5-AMPase and non-specific alkaline phosphatase (-GPase) possessed different pH optima.The cytochemical distribution of the reaction products from the 5-AMPase activity was distinct from those of -GPase. 5-AMPase activity was localized on the surface membranes of acinar, ductal and myoepithelial cells of both salivary glands. -GPase activity was evenly distributed on the entire plasma membranes of myoepithelial cells and on the basal plasmalemma of acinar cells. The reaction products, which appeared on the luminal and lateral plasma membranes of the acinar cells, were presumed to reflect the presence of 5-AMPase, while those on the myoepithelial surface and basal plasma membranes of the acinar cells demonstrated both 5-AMPase and -GPase.The results indicate that 5-AMPase activity can be utilized as a reliable marker enzyme of plasma membranes in the salivary acinar cells.     

Kinetics of the 5′-nucleotidase and the adenosine deaminase in subcellular fractions of rat brain

Suspensions of rat brain microsomes, synaptosomes, and synaptic vesicles were able to convert adenosine to inosine by means of adenosine deaminase. Isosbestic points of this transformation, at 222, 250 and 281 nm, remained unchanged with time-course. This fact suggests that adenosine deaminase (ADA, E.C. 3.5.4.4) is located on the surface of the vesicles whereas purine nucleoside phosphorylase (PNP, E.C. 2.1.2.4) is located inside the vesicles. Kinetic parameters of the particulate 5-nucleotidase (5N, E.C. 3.1.3.5) and adenosine deaminase were analogous to those of the cytosolic enzymes. These results suggest that soluble and particulate enzymes represent different pools of the same molecular species.     

New ionic targets of 3,3′,5′-triiodothyronine at the plasma membrane of rat Sertoli cells

The functions of Sertoli cells, which structurally and functionally support ongoing spermatogenesis, are effectively modulated by thyroid hormones, amongst other molecules. We investigated the mechanism of action of rT₃ on calcium (⁴⁵Ca²⁺) uptake in Sertoli cells by means of in vitro acute incubation. In addition, we performed electrophysiological recordings of potassium efflux in order to understand the cell repolarization, coupled to the calcium uptake triggered by rT₃. Our results indicate that rT₃ induces nongenomic responses, as a rapid activation of whole-cell potassium currents in response to rT₃ occurred in <5 min in Sertoli cells. In addition, the rT₃ metabolite, T₂, also exerted a rapid effect on calcium uptake in immature rat testis and in Sertoli cells. rT₃ also modulated calcium uptake, which occurred within seconds via the action of selective ionic channels and the Na⁺/K⁺ ATPase pump. The rapid response of rT₃ is essentially triggered by calcium uptake and cell repolarization, which appear to mediate the secretory functions of Sertoli cells.     

Lymphocyte plasma membranes II. Cytochemical localization of 5′-nucleotidase in rat lymphocytes

Cytochemical studies of thymic and splenic lymphocytes from rats showed that 5′-nucleotidase was restricted to the plasma membranes. Isolated plasma membranes contained the highest specific activity of 5′-nucleotidase of any cellular fractions. The results indicate that this enzyme can be used as a plasma membrane marker for lymphocytes.