Proc. Natl. Acad. Sci. USA Vol. 84, pp. 1709-1712, March 1987 Neurobiology

Role of sialic acid in synaptosomal transport of amino acid transmitters (y-aminobutyric acid/acidic amino acids/carrier proteins/) MALGORZATA M. ZALESKA AND MARIA ERECIN'SKA* Departments of Biochemistry and Biophysics and of Pharmacology, University of Pennsylvania, School of Medicine, Philadelphia, PA 19104 Communicated by Robert E. Forster II, November 20, 1986 (received for review November 1, 1986)

ABSTRACT Active, high-afflinity, sodium-dependent up- MATERIALS AND METHODS and of the acidic amino acid take of y-aminobutyric acid Male Sprague-Dawley rats (220-250 g) were used throughout D-aspartate was inhibited by pretreatment of synaptosomes the study. Synaptosomes were isolated according to the with neuraminidase from Vibrio cholerae. Inhibition was of a method of Booth and Clark (22); the final pellet was sus- noncompetitive type and was related to the amount ofsialic acid pended in Krebs/Henseleit/Hepes buffer (140 mM NaCl/5 released. The maximum accumulation ratios of both amino mM KCl/1.3 mM MgSO4/5 mM NaHCO3/1 mM Na2HPO4/ acids (intracellular [amino acid]/extraceliular [amino acid]) 10 mM Tris-Hepes, pH 7.4). All media used for incubations remained largely unaltered. Treatment with neuraminidase were supplemented with 10 mM glucose and 1.27 mM CaCI2. affected neither the synaptosomal energy levels nor the con- Incubation Condition and Treatment with Neuraminidase. centration of internal potassium. It is suggested that the The synaptosomal suspension (about 4 mg of protein per ml) y-aminobutyric acid and acidic amino acid transporters are was preincubated for 20 min at 34WC in a shaking water bath glycosylated and that sialic acid is involved in the operation of in the presence or absence of neuraminidase (from Vibrio the carrier proteins directly and not through modification of cholerae, Calbiochem-Behring Diagnostics). The concentra- driving forces responsible for amino acid uptake. tion of neuraminidase varied from 0.001 to 0.05 unit/mg of synaptosomal protein. After incubation the samples were It has become increasingly clear that the carbohydrate diluted 1:20 with Krebs/Henseleit/Hepes buffer and centri- conjugates with proteins and lipids are involved in a number fuged for 10 min at 8000 x g. The pellet was washed twice by of events that occur at the cell surface. The and centrifugation and finally resuspended in the same buffer (4-5 mg of protein per ml). Aliquots were used for uptake studies, may influence recognition, communication, and measurements of sialic acid content and of metabolite and adhesion (1-3); modulate the function of receptors (4-6) and potassium concentrations. of transport processes (7, 8); and affect antigenic activity and Measurements of Amino Acid Uptake. Synaptosomes sus- cell growth (2, 9). It is also well established that N- pended as described above were preincubated for 10 min at acetylneuraminic acid (sialic acid) is one ofthe nine monosac- 28°C and diluted 1:5 into medium of the same composition charides that form carbohydrate portions of membrane containing either 1-32 ,M D-[3H]aspartate (New England glycolipids and glycoproteins. Among the various organs of Nuclear; specific activity, 14 Ci/mmol; 1 Ci = 37 GBq) or the body, brain has the highest content of sialylglycolipids 0.5-16 ,M [14C]GABA (Amersham; specific activity, 231.6 and sialylglycoproteins (6, 10). These compounds are present mCi/mmol). In some experiments 10 ,M [3H]leucine (New in both the neurons and the glia (3, 10), and the England Nuclear; specific activity, 52.2 Ci/mmol) was used. fraction of the plasma membrane of synaptosomes (nerve- Samples (300 ,ul) were withdrawn at 20, 40, and 60 sec and ending particles) has a high proportion of sialic acid (11-13). centrifuged rapidly (Beckman microfuge) through a layer of These observations suggested the hypothesis that sialyl- silicone oil (specific gravity, 1.03; General Electric). (For glycoproteins play a regulatory role in several aspects of leucine, samples were taken at 1, 2, and 3 min.) Radioactivity synaptic function (14, 15). in the pellets and in the total suspensions was measured in a The termination of neurotransmission at GABAergic Searle Delta 300 liquid scintillation counter using Liquiscint (GABA, y-aminobutyric acid) and glutamatergic synapses in LS-121 (National Diagnostics, Somerville, NJ). The rates of the mammalian central nervous system is accomplished uptake were calculated by linear regression analysis of the through the reuptake of the amino acid transmitters by the increase in radioactivity in the pellets. high-affinity, sodium-dependent transport systems (16-21). Measurements of Maximal Amino Acids Gradients. Synap- The molecules bind sodium and amino acids at the tosomes suspended at 4-5 mg of protein per ml were carrier incubated for 20 min at 28°C with 2 uM D-[3H]aspartate and external surface of the plasma membrane and then transport 2 ,M ['4C]GABA. Aminooxyacetic acid (2 mM) was added them into the cell. One might expect, therefore, that they to prevent of GABA. At the end of the incuba- belong to a class of "transmembrane" proteins that span the tion, 300-,ul aliquots were withdrawn and centrifuged through lipid bilayer of the plasma membrane. Like many other silicone oil. The radioactivity was then measured both in the proteins ofthat type, the amino acid transporters may contain pellets and supernatants. Intrasynaptosomal water content residues within their structure that could be important was measured with tritiated water and [14C]polyethylene for their function. The object of this investigation was to glycol as described (23). explore this possibility using a preparation of rat brain Determination of Sialic Acid. Aliquots (1-2 mg of protein) synaptosomes that contain high-affinity transport mecha- of synaptosomal suspensions that had been incubated either nisms (19-21) for the putative amino acid neurotransmitters. Abbreviation: GABA, y-aminobutyric acid. The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed at: Department of payment. This article must therefore be hereby marked "advertisement" Biochemistry and Biophysics, University of Pennsylvania, Phila- in accordance with 18 U.S.C. §1734 solely to indicate this fact. delphia, PA 19104-6059. Downloaded by guest on September 29, 2021 1709 1710 Neurobiology: Zaleska and Erecin'ska Proc. Natl. Acad Sci. USA 84 (1987) with or without neuraminidase were treated with an equal which is common to a number of acidic amino acids including volume of 10% trichloroacetic acid to precipitate the protein. L-glutamate (30-35). Preincubation of synaptosomes with After centrifugation (10 mint 8000 x g), the pellet was increasing concentrations of neuraminidase resulted in a resuspended in 1.5 mll of0.05 M H2SO4 and hydrolyzed for 1.5 progressive decrease in the rate of uptake of both amino hr at 80'C. The samples were centrifuged, and the resulting acids. The effect was significant (GABA, P < 0.02; D- supernates were analyzed for free by the aspartate, P < 0.05) at the lowest amount of enzyme tested method of Warren (24). A standard of crystalline neuraminic (0.00125 unit/mg of synaptosomal protein) and reached a acid (Sigma) was used as reference. The recovery of known plateau of <50% of the control value at about 0.015 unit of amounts of neuraminic acid, which were added to precipitat- neuraminidase per mg of synaptosomal protein. Further ed synaptosomal protein and carried through the entire increase in the concentration of the enzyme gave only a slight procedure, was 90 ± 4% (mean ± SEM; n = 6). increase in the degree of inhibition. Measurements of Metabolite and Potassium Concentrations. In contrast to the behavior exhibited by GABA and The incubations were terminated by the addition of cold 0.6 D-aspartate, transport of L-[PH]leucine into synaptosomes M perchloric acid (final concentration), and the extracts were was not affected by pretreatment with neuraminidase. After neutralized with 3 M K2CO3 in 0.5 M triethanolamine. After a 20-min incubation with the enzyme (0.017 unit/mg of centrifugation, concentrations of metabolites were deter- synaptosomal protein) the rate of leucine uptake was found mined in the clear supernatant fraction by standard enzy- to be 69.9 ± 7.5 pmol/min per mg of protein (mean ± SEM; matic procedures: ATP and creatine phosphate by the meth- n = 3), which was not significantly different from the control od of Lamprecht and Trautschold (25) and ADP and creatine value of 70.5 ± 8.5 pmol/min per mg of protein (mean ± according to Bernt et al. (26). Intra- and extrasynaptosomal SEM; n = 3). potassium was measured by atomic absorption as described The Effect of Neuraminidase Treatment on the Maximum (23). Protein concentration was determined by the biuret Accumulation Ratios of GABA and D-Aspartate and on the standard (27). Synaptosomal Energy and Potassium Levels. The maximum reaction using bovine serum albumin as accumulation ratios of GABA and D-aspartate were mea- sured from the distribution of the radiolabeled amino acids RESULTS between the inside of synaptosomes and the external medium the in samples preincubated with and without neuraminidase. The Effect of Neuraminidase Treatment on Synaptosom- The intracellular [GABA]/extracellular [GABA] ratio was al Uptake of GAIIA and D-Aspartate. The effect of neuramini- 750 ± 42 (mean ± SEM; n = 3) for control conditions and 513 dase pretreatment on the synaptosomal uptake of GABA and ± 41 (mean ± SEM; n = 3) for samples incubated with 0.025 D-aspartate is shown in Fig. 1. It is well documented that unit of neuraminidase per mg of synaptosomal protein. The GABA is transported into synaptosomes on a separate carrier corresponding ratio values for D-aspartate were 1100 ± 100 protein(28), whereas D-aspartate, anonmetabolizeable stereo- and 807 ± 57 (mean ± SEM; n = 3), respectively. isomer of aspartate (29), is taken up by a different system, Simultaneous determinations of the internal level of po- 100- tassium ions revealed no differences between the control and neuraminidase-treated preparations: 289 ± 10 versus 275 ± 15 nmol of K+ per mg of synaptosomal protein (both means 90- + SEM; n = 3), respectively. In addition, synaptosomes incubated with the glycosidase did not differ in their content of the high-energy phosphate compounds. The values found [ATP, 4.6 ± 0.1; creatine phosphate, 6.7 ± 0.4; and creatine, 18.7 ± 0.5 nmol/mg of protein (means ± SEM; n = 3)] were, within the limits of experimental error, the same for prepa- 0) rations preincubated with and without neuraminidase. The Effect of Treatment with Neuraminidase on the Kinetic Parameters of Amino Acid Uptake. To provide an insight into 4- 0 the mechanism of inhibition of amino acid transport caused by treatment of synaptosomes with neuraminidase, kinetic Asp parameters of amino acid uptake were determined for the GABA control and treated preparations. The kinetic constants were calculated from the initial rates of uptake using the double- reciprocal plot analysis (Fig. 2). It can be seen that the general characteristics of the plots are typical of a noncompetitive 0.01 0.02 0.03 0.04 0.05 inhibition, treatment with the glycosidase decreased the Vma Neuraminidase (U/mg synaptosomal protein) values without affecting the Km value. The Effect of Treatment with Neuraminidase on the Release FIG. 1. Inhibition of [14C]GABA and D-[3H]aspartate uptake by of Synaptosomal Sialic Acid. Fig. 3 shows quantitative rela- treatment with neuraminidase. Synaptosonmes (4-5 mg of protein per ml) suspended in Krebs/Henseleit/Hepes buffer and supplemented tionships between the concentration of neuraminidase added with 1.27 mM CaCl2 and 10 mM glucose were incubated for 20 min into the incubation mixture and the amount of sialic acid at 340C with and without various concentrations of neuraminidase released from synaptosomes during the 20-min incubation from Vibrio cholerae. After two washes by centrifugation and a period. The amount of sialic acid released by the treatment resuspension of the pellet in the same buffer, synaptosomes were was calculated as the difference in the total sialic acid found preincubated for 10 min at 280C. Uptake studies were initiated by a in the synaptosomes incubated with and without the enzyme. 1:5 dilution of the preparation into a medium containing 2 AM With no exogenous neuraminidase added, synaptosomes [14C]GABA (A) and 2 /AM D-[3H]aspartate (D). Aliquots (300 /l) were incubated 20 min released 4-5 nmol of sialic acid per mg of withdrawn after 20, 40, and 60 sec and centrifuged through silicone As illustrated in the amount of sialic acid was a Delta 300 protein. Fig. 3, oil. The radioactivity in the pellets counted in Searle released from synaptosomes increased as a function of the liquid scintillation counter. Values are means ± SEM for three or four separate experiments. Results are expressed as percentages of amount of enzyme added, reaching a plateau at about 0.01 respective control values (GABA, 0.21 ± 0.01; D-aspartate, 0.45 unit of neuraminidase per mg protein. It is interesting to note 0.03 nmol/min per mg of protein, n = 4). U, unit. the obvious similarity of this curve with those shown in Fig. Downloaded by guest on September 29, 2021 Neurobiology: Zaleska and Erecifiska Proc. Natl. Acad. Sci. USA 84 (1987) 1711 40- ._ - 5, 0.- 0. A Neuraminidase -2 30

E 0 x C. 20 E x E 7@ 10 l c >" Z 0 0.01 0.02 0.03 0.04 0.05 -0.4 -0.2 Neuraminidase (U/mg synaptosomal protein) 1[ D-Asp ], (pM-1) FIG. 3. Neuraminidase-induced release of sialic acid from 14 B synaptosomal preparations. Synaptosomes were treated with neuraminidase as described in the legend to Fig. 1. After being Neuraminidase washed and suspended, the final pellet was precipitated with tri- C chloroacetic acid and hydrolyzed in 0.05 M H2SO4. The content of 10 sialic acid was determined by the procedure of Warren (24). The 0. 0) 10 quantity of sialic acid released was calculated by subtracting the amount found in the preparation incubated with neuraminidase from E that incubated in its absence. Values presented are means ± SEM for x 8- three or four separate experiments. Neu Ac, neuraminic acid; U, unit. E 6- x brane sialic acid. The for this E simplest explanation phenom- C 4 enon is that both the GABA and the acidic amino acid transporters are glycosylated and that sialic acid constitutes I an integral part of the transport mechanism. It is well known that synaptosomal uptake of GABA and acidic amino acids is by a cotransport system with Na' (see -0.5 0 0.5 2 refs. 17, 19, and 31 for review) and that the driving force for amino acid accumulation is provided by a combination of the 1/[ GABA ], (IlM-') transmembrane electrical potential and the sodium concen- FIG. 2. Kinetic parameters of D-[3H]aspartate and ['4C]GABA tration gradient (36-40). Hence, any reduction in either ofthe uptake after pretreatment of synaptosomes with and without driving forces would necessarily result in a decrease in amino neuraminidase. Synaptosomes were incubated with (o) or without acid transport. However, our results show that the final (e) neuraminidase (0.017 unit/mg of protein) and further treated as maximum accumulation ratios of both GABA and aspartate described in the legend to Fig. 1. For uptake measurements aliquots (which are directly related to the magnitude of the driving were diluted 1:5 into media containing 1, 2, 4, 8, and 16 ,uM forces) are largely unaffected by the treatment with neura- radioactive D-aspartate (A) or 0.5, 1, 2, 4, and 8 AM radioactive GABA (B). The results are presented as double-reciprocal plots of minidase; neither are the synaptosomal energy levels and the the initial rates of uptake versus the amino acid concentration. concentration of internal potassium. [The latter observation Kinetic constants were as follows. (A) Km, 5.3 1.8 and 1.3 AM; V..ax, 70- nmol/min per mg of protein for control and neuraminidase-treated GABM synaptosomes, respectively. (B) Km, 2.8 /AM; Vmax, 0.56 and 0.39 nmol/min per mg of protein, for the two conditions, respectively. 60- Asp Values are means of three independent experiments with SD within 5% range. -0 5 0 _ 4 0 1. When the amount of sialic acid released from synapto- somes by treatment with various concentrations of neura- 0 30' minidase was plotted against the degree of inhibition of GABA and D-aspartate uptake by preparations treated with , 20 the corresponding concentrations ofthe enzyme straight lines shown in Fig. 4 were obtained. The linear regression analysis 10 of these plots gave best fit to straight lines with correlation coefficients of 0.984 for GABA and 0.985 for D-aspartate. 10 20 30 DISCUSSION Neu Ac release ( nmol mg synaptosomal protein) The main finding of the present study is that incubation of FIG. 4. Relationship between the extent of inhibition of amino acid uptake and the loss of sialic acid from synaptosomes. The results nerve-ending particles (synaptosomes) from rat brain with presented in Figs. 1 and 3 were used in the analysis. The lines were the specific exoglycosidase neuraminidase decreases the fitted by the least-squares method. The data for aspartate were found ability of synaptosomes to take up the amino acid to give best fit to a straight line with a slope of 2.19 and a correlation neurotransmitters GABA and D-aspartate through the high- coefficient, r, of 0.985 whereas those for GABA gave a slope of 1.70 affinity transport systems in parallel with the loss of mem- and an r of 0.984. Neu Ac, neuraminic acid. Downloaded by guest on September 29, 2021 1712 Neurobiology: Zaleska and Erecinska Proc. Nati. Acad. Sci. USA 84 (1987) is in agreement with the earlier finding (41) that the activity vous System, eds. Margolis, R. U. & Margolis, R. K. (Plenum, of the Na/K ATPase is insensitive to neuraminidase treat- New York), pp. 165-184. ment.] It seems, therefore, that inhibition of amino acid 11. Brunngraber, E. G., Dekirmenjian, H. & Brown, B. D. (1967) transport a consequence of a direct Biochem. J. 103, 73-78. is effect of the loss of 12. Dekirmenjian, H. & Brunngraber, E. G. (1969) Biochim. sialic acid on the carriers themselves and is not caused by an Biophys. Acta 177, 1-10. indirect influence of altered ionic gradients on the uptake 13. Dicesare, J. L. & Rapport, M. M. (1973) J. Neurochem. 20, mechanism. 1781-1783. Although the mechanism of inhibition of GABAo and 14. Rahmann, H., Rosner, H. & Breer, H. (1976) J. Theor. Biol. aspartate uptake by neuraminidase treatment is not clear at 57, 231-237. the present time, some relevant speculations can be offered. 15. Barchi, R. L. (1983) J. Neurochem. 40, 1377-1385. It has been shown that the sialic acid residues reside at the 16. Iversen, L. L. (1971) Br. J. Pharmacol. 41, 571-591. ends of the carbohydrate side chains and hence are fully 17. Martin, D. L. (1973) J. Neurochem. 21, 345-356. exposed to the external environment. 18. Curtis, D. R. & Johnston, G. A. R. (1974) Ergeb. Physiol. Our results demon- Biol. Chem. Exp. Pharmakol. 69, 97-188. strate that only the high-affinity, Na'-dependent transport of 19. DeFeudis, F. V. (1975) Annu. Rev. Pharmacol. 15, 105-130. amino acids is affected by removal of the surface sialic acid 20. Fonnum, F. (1978) Amino Acids as Chemical Transmitters whereas the uptake of leucine, which is Na' independent (Plenum, New York). (42), remains unaltered. Furthermore, the Na'-dependent 21. Fagg, G. E. & Lane, J. D. (1979) Neuroscience 4, 1015-1036. uptake of serotonin is also decreased by preincubation with 22. Booth, R. F. G. & Clark, J. B. (1978) Biochem. J. 176, neuraminidase (41). Therefore, it is possible that the nega- 365-370. tively charged sialic acid residues either form a part of the 23. Troeger, M. B., Rafalowska, U. & Erecifiska, M. (1984) J. binding site for Na+ or, alternatively, that they are involved Neurochem. 42, 1735-1742. in the "active" conformation ofthe carrier 24. Warren, L. (1959) J. Biol. Chem. 234, 1971-1975. protein that binds 25. Lamprecht, W. & Trautschold, I. (1974) in Methods ofEnzy- the cation. Since for productive substrate translocation the matic Analysis, ed. Bergmeyer, H. U. (Academic, New York), carriers must be fully loaded, i.e., they must acquire both the pp. 2101-2110. amino acid and the Na' (43, 44), those protein molecules that 26. Bernt, E., Bergmeyer, H. U. & Mollering, H. (1974) in Meth- contain less sialic acid, and consequently bind Na' less ods ofEnzymatic Analysis, ed. Bergmeyer, H. U. (Academic, efficiently, would be unlikely to participate in the transport New York), pp. 1772-1776. cycle. Our finding that the Vmax for the amino acid transport, 27. Gornall, A. G., Bardawill, C. J. & David, M. M. (1949) J. Biol. but not the Km, is altered by treatment with neuraminidase is Chem. 177, 751-766. consistent with this suggestion. 28. Balcar, V. J. & Johnston, G. A. R. (1973) J. Neurochem. 20, that 529-539. Finally, it should be pointed out although our results 29. Davies, L. P. & Johnston, G. A. R. (1976) J. Neurochem. 26, suggest that the GABA and acidic amino acid transporters are 1007-1014. glycoproteins, they do not prove it. The possibility remains 30. Balcar, V. P. & Johnston, G. A. R. (1972) J. Neurochem. 19, that depletion of sialic acid residues of other membrane 2657-2666. proteins or ofglycolipids causes structural changes in the two 31. Peterson, N. A. & Raghupathy, E. (1972) J. Neurochem. 19, carrier proteins that interfere with amino acid transport. 1423-1438. However, the observation that sialoglycoproteins of synaptic 32. Snyder, S. H., Young, A. B., Bennett, J. P., Jr., & Mulder, plasma membrane are more susceptible to hydrolysis with A. H. (1973) Fed. Proc. Fed. Am. Soc. Exp. Biol. 32, exogenous neuraminidase than the (13) seems to 2039-2047. 33. Johnson, J. L. (1978) Prog. Neurobiol. 10, 155-202. suggest that the former carbohydrate conjugates play the 34. Watkins, J. C. & Evans, R. H. (1981) Annu. Rev. Pharmacol. more important role in the events described in this work. Toxicol. 21, 165-201. 35. Erecifiska, M. & Troeger, M. B. (1986) FEBS Lett. 199, 95-99. This work was supported by Grant NS 10939 from the National 36. Martin, D. L. (1976) in GABA in the Nervous System Func- Institutes of Health. tion, eds. Roberts, E., Chase, T. N. & Towers, D. B. (Raven, New York), pp. 347-385. 37. Blaustein, M. P. & King, A. C. (1976) J. Membr. Biol. 30, 1. Schwarz, R. T. & Datema, R. (1982) Adv. Carbohydr, Chem. 153-173. Biochem. 40, 287-379. 38. Pastuszko, A., Wilson, D. F. & Erecin'ska, M. (1982) J. Biol. 2. Hakomori, S. (1981) Annu. Rev. Biochem. 50, 733-764. Chem. 257, 7514-7519. 3. Edelman, G. M. (1983) Science 219, 450-457. 39. Erecifiska, M., Wantorsky, D. & Wilson, D. F. (1983) J. Biol. 4. Jacobs, S. & Cuatrecasas, P. (1983) Annu. Rev. Pharmacol. Chem. 258, 9069-9077. Toxicol. 23, 461-479. 40. Kanner, B. I. (1983) Biochim. Biophys. Acta 726, 293-316. 5. Hayes, G. R. & Lockwood, D. H. (1986) J. Biol. Chem. 261, 41. Dette, G. A. & Wasemann, W. (1978) Hoppe-Seyler's Z. 2791-2798. Physiol. Chem. 359, 399-406. 6. Wiegant, H. (1982) Adv. Neurochem. 4, 149-221. 42. Bennett, J. P., Jr., Mulder, A. H. & Snyder, S. H. (1974) Life 7. Fishman, P. H. & Brady, R. 0. (1976) Science 194, 906-915. Sci. 15, 1045-1056. 8. Guidotti, G. (1976) Trends Biochem. Sci. 1, 11-13. 43. Wheeler, D. D. (1980) J. Neurosci. Res. 5, 323-337. 9. Ledeen, R. (1985) Trends NeuroSci. 8, 169-174. 44. Nelson, M. T. & Blaustein, M. P. (1982) J. Membr. Biol. 69, 10. Mahler, H. R. (1979) in Complex Carbohydrates of the Ner- 213-223. Downloaded by guest on September 29, 2021