Polyamine transport, accumulation, and release in brain

Authors


Address correspondence and reprint requests to Kazuei Igarashi, Graduate School of Pharmaceutical Sciences, Chiba University, 1–33 Yayoi-cho, Inage-ku, Chiba 263–8522, Japan. E-mail: iga16077@p.chiba-u.ac.jp

Abstract

Cycling of polyamines (spermine and spermidine) in the brain was examined by measuring polyamine transport in synaptic vesicles, synaptosomes and glial cells, and the release of spermine from hippocampal slices. It was found that membrane potential-dependent polyamine transport systems exist in synaptosomes and glial cells, and a proton gradient-dependent polyamine transport system exists in synaptic vesicles. The glial cell transporter had high affinities for both spermine and spermidine, whereas the transporters in synaptosomes and synaptic vesicles had a much higher affinity for spermine than for spermidine. Polyamine transport by synaptosomes was inhibited by putrescine, agmatine, histidine, and histamine. Transport by glial cells was also inhibited by these four compounds and additionally by norepinephrine. On the other hand, polyamine transport by synaptic vesicles was inhibited only by putrescine and histamine. These results suggest that the polyamine transporters present in glial cells, neurons, and synaptic vesicles each have different properties and are, presumably, different molecular entities. Spermine was found to be accumulated in synaptic vesicles and was released from rat hippocampal slices by depolarization using a high concentration of KCl. Polyamines, in particular spermine, may function as neuromodulators in the brain.

Abbreviations used
ATP

adenosine triphosphate

CCCP

carbonylcyanide m-chlorophenylhydrazone

HPLC

high-performance liquid chromoatography

Kir channel

inward rectifier K+ channel

MEM

modified essential medium

NMDA

N-methyl-d-aspartate

TCA

trichloroacetic acid

The endogenous polyamines putrescine, spermidine and, in particular, spermine modulate a number of types of ion channels (Williams 1997). Intracellular spermine is responsible for intrinsic gating and rectification of strong inward rectifier K+ (Kir) channels by directly plugging the ion channel pore (Ficker et al. 1994; Lopatin et al. 1994, 1995; Fakler et al. 1995; Nichols and Lopatin 1997). Intracellular spermine also causes inward rectification at some subtypes of Ca2+-permeable AMPA and kainate-type glutamate receptors in the CNS, again by plugging the receptor channel pore (Bowie and Mayer 1995; Kamboj et al. 1995). Extracellular spermine has multiple effects at the N-methyl-d-aspartate (NMDA) subtype of glutamate receptor, including stimulation that increases the size of NMDA receptor currents and voltage-dependent block (Williams et al. 1994; Zhang et al. 1994; Williams 1997; Dingledine et al. 1999).

Some progress has been made in identifying the polyamine-binding sites on Kir channels (Taglialatela et al. 1995; Yang et al. 1995; Kubo and Murata 2001) and on AMPA/kainate and NMDA channels (Williams et al. 1995; Kashiwagi et al. 1997, 2002; Masuko et al. 1999; Panchenko et al. 1999). The role of intracellular polyamines in block of Kir and AMPA channels is well-established, but the in vivo role of extracellular polyamines at NMDA or other receptors is unclear. If extracellular polyamines modulate receptors, and thus synaptic transmission in vivo, the polyamines must be released from neurons or glia and rapidly re-incorporated into those cells. Polyamines are present at relatively high concentrations in the brain, with differences in regional distribution (Seiler 1981). However, little is known about transport, uptake, and release of polyamines in neurons and glia. Previous work has described high-affinity uptake and release of polyamines in the brain (Harman and Shaw 1981; Gilad and Gilad 1991), but the properties and localization of brain polyamine transport systems have not been studied in detail. In this paper, we studied uptake and release of polyamines in brain tissue, and characterized polyamine transport systems in synaptic vesicles, synaptosomes, and glial cells, and the release of spermine from hippocampal slices.

Materials and methods

Preparation of synaptic vesicles, synaptosomes, and glial cells

Synaptic vesicles were prepared from rat cerebral tissues according to the procedure described by Kish and Ueda (1989). Whole cerebrum from 6-week-old male Sprague–Dawley rats was homogenized in 10 vol of buffer (140 mm potassium gluconate, 1 mm sodium hydrogen carbonate, 1 mm magnesium acetate, pH 7.2), and centrifuged at 12 000 g for 15 min. The precipitate was resuspended in 20 vol of 6 mm Tris–maleate, pH 8.1, and centrifuged at 43 500 g for 15 min. The supernatant was centrifuged at 200 000 g for 60 min, and right-side-out synaptic vesicles were prepared from the precipitate by hypo-osmotic lysis with 20 vol of 10 mm potassium gluconate by the method of Noremberg and Parsons (1989). The purity of the synaptic vesicles was assessed by glutamate transport activity (Kish and Ueda 1989).

Synaptosomes were prepared from rat cerebral tissues by the method of Fletcher and Johnston (1991). Cerebral tissue from approximately 6-week-old male Sprague–Dawley rats was homogenized in 9 vol of 0.32 m sucrose, and centrifuged at 1200 g for 10 min. The supernatant was loaded onto an equal volume of 1.2 m sucrose and centrifuged at 110 000 g for 20 min. The fraction retained at the interface was collected and loaded onto 10 mL of 0.8 m sucrose. After centrifugation at 110 000 g for 20 min, the synaptosomal pellet was resuspended in 0.32 m sucrose and used for the transport assay.

Glial cells (astrocytes) were prepared from cerebral cortex of 1- to 3-day-old rats and cultured in modified essential medium (MEM, Nissui Co. Ltd, Japan) containing 10% fetal bovine serum according to the method of Noble and Mayer-Pröschel (1998).

Measurement of polyamines

Rat whole cerebrum, synaptic vesicles, synaptosomes, and glial cells were homogenized with 5% (w/v) trichloroacetic acid (TCA) and centrifuged. The supernatant was used for the assay of polyamines. Polyamines (putrescine, spermidine, and spermine) were measured by using a TOSOH high-performance liquid chromoatography (HPLC) system (Tosoh Co., Tokyo) as described previously (Igarashi et al. 1986). Protein content was determined using the 5% TCA precipitate by the method of Lowry et al. (1951), and polyamine content expressed as nmol/mg protein.

Assays for polyamine uptake

The reaction mixture (0.1 mL) for measuring uptake by synaptic vesicles contained 10 mm Tris–HCl, pH 7.8, 0.14 m KCl, 2 mm adenosine triphosphate (ATP), 100 µm[14C]spermine or [14C]spermidine (185 MBq/mmol, Amersham Biosciences, Piscataway, NJ, USA), and synaptic vesicles (200 µg protein). After incubation at 30°C for 1–4 min, the vesicles were collected on membrane filters (HAWP filter, 0.45 µm; Millipore, Bedford, MA, USA) and washed three times with a total of 15 mL of a buffer containing 10 mm Tris–HCl, pH 7.8, 0.14 m KCl, and 1 mm spermine or spermidine. Radioactivity on the filters was determined by using a liquid scintillation spectrometer. Non-specific binding was defined as radioactivity of the reaction mixture kept at 0°C and was subtracted from the total radioactivity to obtain values for specific uptake.

The reaction mixture (0.1 mL) for measuring uptake by synaptosomes contained 10 mm Tris–HCl, pH 7.8, 0.118 m NaCl, 100 µm[14C]spermine or [14C]spermidine (185 MBq/mmol), and synaptosomes (100 µg protein). After incubation at 30°C for 4–12 min, synaptosomes were collected on membrane filters (GF/B filter, 0.45 µm; Whatman, International Ltd, Kent, UK). After washing the filters with 15 mL of a buffer containing 10 mm Tris–HCl, pH 7.8, 0.118 m NaCl and 1 mm spermine or spermidine, radioactivity on the filters was determined by using a liquid scintillation spectrometer.

The polyamine uptake activity of glial cells (5 × 106 cells/mL) was measured in MEM containing 10% fetal bovine serum using 100 µm[14C]spermine or [14C]spermidine (185 MBq/mmol). During the assay, 1 mm aminoguanidine, an inhibitor of amine oxidase in serum, was added to the medium. After incubation at 30°C for 4–12 min, 0.5 mL cell suspension was loaded onto 0.6 mL of oil layer (corn oil : di-n-butyl phthalate, ratio 3 : 10) and cells were collected by centrifugation through the oil layer as described (Kakinuma et al. 1988). The amount of radioactivity in the cells was measured in 10 mL of Triton–toluene scintillant after sonication with 1 mL of 5% TCA.

Release of [14C]spermine from hippocampus

Rat hippocampus was dissected as described (Glowinski and Iversen 1966) and sliced manually. Slices were labeled in buffer containing 20 mm HEPES–Tris, pH 7.2, 135 mm NaCl, 2 mm CaCl2, 0.2 mm MgCl2, 10 mm glucose and 5 µm[14C]spermine (4.18 GBq/mmol) by incubation at 37°C for 30 min in an atmosphere of 5% CO2. Three to seven slices labeled with [14C]spermine were placed in a Brandel (Gaithersburg, MD, USA) superfusion apparatus. The release of [14C]spermine from hippocampal slices was studied using the method of Baba et al. (1983). Slices were superfused at a rate of 1 mL/min, and fractions were collected every 2 min. After 40 min of superfusion, slices were stimulated with 15–50 mm KCl for 4 min. Radioactivity in each fraction was determined by using a liquid scintillation spectrometer.

Results

Uptake of spermine and spermidine by synaptic vesicles, synaptosomes, and glial cells

As shown in Fig. 1, the initial velocity of spermine and spermidine uptake was much greater in glial cells than in synaptic vesicles and synaptosomes when 100 µm spermine or spermidine was used as a substrate. Spermidine was more efficiently transported into glial cells than spermine, but spermine was more efficiently transported into synaptic vesicles and synaptosomes. The uptake rate versus various concentrations of spermine and spermidine, and the Km and Vmax values are shown in Fig. 2. The Km values for spermine were smaller than those for spermidine, but the Vmax values for spermidine were greater than those for spermine. The Km value for spermidine was much smaller in glial cells than synaptosomes and synaptic vesicles. These results indicate that efficient spermine and spermidine uptake systems, comparable to those of bovine lymphocytes (Kakinuma et al. 1988) and mouse mammary carcinoma FM3A cells (Sakata et al. 2000), exist in synaptic vesicles, synaptosomes, and glial cells although the Km values for spermine and spermidine are relatively high in these three systems.

Figure 1.

Uptake of spermine and spermidine in synaptic vesicles, synaptosomes, and glial cells. Uptake activities were measured as described in Materials and methods. •, Spermine (SPM) uptake; ▵, spermidine (SPD) uptake. The insets show the expanded figure on the vertical axis. Data are shown as mean ± SE of triplicate determinations.

Figure 2.

Concentration dependence of spermine and spermidine uptake by synaptic vesicles, synaptosomes, and glial cells. Km and Vmax values were measured from the data obtained with 2 min incubation according to the Eadie–Hofstee plot by changing the concentration of spermine and spermidine, from 25 µm to 200 µm and 50 µm to 800 µm, respectively. At these concentrations, the uptake by synaptic vesicles, synaptosomes, and glial cells was linear until 2, 4, and 4 min, respectively. Data are shown as mean ± SE of triplicate determinations.

We subsequently examined the driving force for polyamine uptake. As shown in Fig. 3, addition of 10 µm carbonylcyanide m-chlorophenylhydrazone (CCCP), an ionophore that allows permeation of monovalent cations, inhibited spermine uptake by synaptic vesicles, synaptosomes, and glial cells. Addition of 1 µm bafilomycin A1, a specific inhibitor of vacuolar H+-ATPase (Bowman et al. 1988), only inhibited the uptake by synaptic vesicles, but not by synaptosomes and glial cells. Addition of 5 mm azide, an uncoupler of respiratory chain in mitochondria, inhibited spermine uptake by synaptosomes and glial cells, and removal of 2 mm ATP inhibited the uptake by synaptic vesicles. These results indicate that spermine uptake in synaptic vesicles is dependent on the proton gradient produced by H+-ATPase, and the uptake in synaptosomes and glial cells depends on the membrane potential across the plasma membrane.

Figure 3.

Effect of various inhibitors or ATP depletion on spermine uptake in synaptic vesicles, synaptosomes, and glial cells. Spermine uptake was measured in the presence and absence of 10 µm CCCP, 2 mm ATP, 1 µm bafilomycin A1, or 5 mm NaN3 as described in Materials and methods. The control (100%) values are 46.3, 22.5, and 65.4 pmol/min/mg protein for synaptic vesicles, synaptosomes, and glial cells, respectively. Data are shown as mean ± SE of triplicate determinations.

The effects on polyamine uptake of various basic amino acids, amines, and polyamines were measured using concentrations 10-fold greater than the substrate (Fig. 4). The uptake of spermine was inhibited by spermidine and vice versa, suggesting that uptake of spermine and spermidine is catalyzed by the same transport protein in any one system (synaptic vesicles, synaptosomes, and glial cells), although transporters in different systems have different properties. Uptake of spermine and spermidine by glial cells and synaptosomes was inhibited by histidine and agmatine, and weakly by putrescine and histamine. Furthermore, uptake by glial cells was inhibited by norepinephrine. On the other hand, uptake by synaptic vesicles was weakly inhibited by putrescine and histamine, but not by histidine, agmatine, and norepinephrine. The results suggest that the polyamine transporter in synaptic vesicles has a relatively rigorous substrate specificity that recognizes polyamines (putrescine, spermidine, and spermine) and histamine. These results also suggest that polyamine transporters in synaptic vesicles, neuronal cells, and glial cells are different from each other, although it is possible that these three transport systems contain common components.

Figure 4.

Effects of basic amino acids and amines on polyamine uptake in synaptic vesicles, synaptosomes, and glial cells. Uptake of spermine (SPM) and spermidine (SPD) was measured in the presence and absence of 1 mm basic amino acids or amines. The control (100%) values (pmol/min/mg protein) for uptake are 46.3 for spermine and 38.2 for spermidine in synaptic vesicles, 22.5 for spermine and 20.3 for spermidine in synaptosomes, and 65.4 for spermine and 85.2 for spermidine in glial cells. Data are shown in mean ± SE of triplicate determinations.

Because spermine transport in synaptic vesicles was inhibited by histamine, the properties of this polyamine transporter were compared with those of vesicular monoamine transporters which also recognize histamine. The activities of vesicular monoamine transporters are inhibited by reserpine and amphetamine (Erickson et al. 1996). As shown in Fig. 5, spermine transport was inhibited strongly by reserpine, weakly by 5-hydroxytryptamine and methamphetamine, which has similar structure to amphetamine, and the Ki value of histamine for inhibiting spermine transport was 1.8 mm. Thus, the polyamine transporter in synaptic vesicles may have some features in common with the monoamine transporters.

Figure 5.

Effects of histamine, 5-hydroxytryptamine, reserpine, and methamphetamine on spermine uptake in synaptic vesicles. Uptake of spermine was measured in the presence of various concentrations of each compound as shown. The control (100%) value for spermine uptake was 45.3 pmol/min/mg protein. Data are shown in mean ± SE of triplicate determinations.

Accumulation of spermine in synaptic vesicles

Given that there is a polyamine transporter in synaptic vesicles, we examined whether accumulation of spermine takes place in these vesicles. As shown in Table 1, spermine was accumulated in synaptic vesicles, in which the spermine content (8.28 nmol/mg protein) was highest among whole cerebrum, synaptic vesicles, synaptosomes, and glial cells. The concentration of spermine was 1.5–2.8 mm assuming that 1 mg protein corresponds to 3–5.5 µL of cell volume (Watanabe et al. 1991; Miyamoto et al. 1993). Furthermore, the polyamine content of synaptosomes was relatively high, consistent with the idea that polyamines may play important roles in neurons. The polyamine content in glial cells was lower than that in whole cerebrum, even though glial cells have a more efficient polyamine uptake system than synaptosomes. It is conceivable that glial cells have a low rate of synthesis or a high rate of turnover of polyamines, accounting for their low polyamine content.

Table 1.  Polyamine content of various fractions in rat brain
FractionPutrescineSpermidine
(nmol/mg protein)
Spermine
  1. Polyamines were extracted by 5% trichloroacetic acid from each fraction, and measured by HPLC as described in Materials and methods. Data are shown as mean ± SE of triplicate determinations.

Whole cerebrum0.06 ± 0.014.25 ± 0.153.10 ± 0.40
Synaptic vesicles0.02 ± 0.014.66 ± 0.688.28 ± 1.02
Synaptosomes0.15 ± 0.047.14 ± 1.795.76 ± 1.75
Glial cells0.10 ± 0.032.80 ± 0.282.40 ± 0.35

Release of spermine from hippocampus

Experiments were carried out to determine whether spermine was released from hippocampal tissues under a depolarized state. For these experiments, hippocampal slices were pre-loaded with [14C]spermine, depolarized by superfusion with a buffer containing 15–50 mm KCl, and the release of [14C]spermine was measured. As shown in Fig. 6, the release of [14C]spermine was dependent on the number of slices and the concentration of KCl. The results suggest that spermine is released from neurons by depolarization, perhaps from synaptic vesicles, and could act as a neuromodulator.

Figure 6.

Depolarization-induced release of spermine from rat hippocampal slices. KCl-induced release of [14C]spermine in rat hippocampal slices was measured as described in Materials and methods. (a) A total of 50 mm KCl-induced spermine release was measured using three (•), five (▵), and seven (□) slices. (b) Spermine release was evoked by 15 (•), 30 (▵), and 50 mm (□) KCl at five slices. Data are shown as the mean of duplicate determinations.

Discussion

Recently, there has been considerable interest in the possible role of polyamines as modulators of a number of types of ion channels (Williams 1997; Dingledine et al. 1999). However, the localization and cycling of polyamines in the brain is not well understood. It has been reported that polyamines are taken up by slices of rat cerebral cortex (Harman and Shaw 1981), by cultured cerebellar astrocytes (Dot et al. 2000), and by synaptosomes (Gilad and Gilad 1991).

In this study, we have clarified some aspects of polyamine cycling in brain by systematically studying uptake and release of polyamines. We confirmed the existence of an efficient polyamine uptake system in glial cells. However, the Km values for spermine and spermidine were 30–50 times higher under our experimental conditions compared to a previous report (Dot et al. 2000). The Km values that we measured for polyamine transport in synaptosomes were also much higher than previously reported (Gilad and Gilad 1991). This may be due to methodological difference between the current work and previous studies. In the present work, non-labeled spermine or spermidine was included in the washing buffer for measurement of polyamine transport in synaptic vesicles and synaptosomes, and glial cells were collected by centrifugation through an oil layer. Thus, under our experimental conditions, non-specific binding of [14C]-polyamines is greatly diminished.

It is notable that the polyamine transporters in synaptosomes and glial cells have broad substrate specificity. They recognize agmatine, histidine, and histamine as well as polyamines (spermine, spermidine, and putrescine). There is a previous report that agmatine can be transported by a polyamine transporter in mammalian NIH3T3 cells (Satriano et al. 2001). In addition, the polyamine transport system in glial cells may also recognize norepinephrine. It is also possible, although less likely, that norepinephrine inhibits polyamine transport indirectly via a signaling cascade after activation of adrenergic receptors. Polyamines were efficiently transported into glial cells, but we have not observed a multiphasic pattern of kinetics under our experimental conditions, suggesting that a single kind of transporter catalyzes polyamine uptake in glial cells. We also found a polyamine transport system in synaptic vesicles. This transport system has a rather strict substrate specificity. It recognizes only histamine in addition to polyamines (spermine, spermidine, and putrescine).

Although polyamine transport systems in prokaryotes have been characterized at the molecular level (Igarashi and Kashiwagi 1999), those on the plasma membrane in eukaryotic cells have not. However, the properties of polyamine transporters have been characterized in various cells and tissues (Seiler et al. 1996). In many cases, including bovine lymphocytes, polyamines are taken up with Km values in the micromolar or high nanomolar range, whereas polyamine transport systems in synaptosomes and glial cells had relatively low affinities for polyamines (Km: 10–100 micromolar range). However, the Vmax values of polyamine transporters in synaptosomes and glial cells were greater than that in bovine lymphocytes (Kakinuma et al. 1988). This suggests that high concentrations of polyamines, if released into synapses after depolarization, are effectively taken up in neurons and glial cells.

In relation to the polyamine transporter in synaptic vesicles, we have described multiple, ATP-driven proton gradient-dependent polyamine transport systems on the vacuolar membrane in Saccharomyces cerevisiae (Tomitori et al. 1999, 2001), whose properties are similar to those of the transporter in synaptic vesicles. The vacuolar transporters of S. cerevisiae have 12 transmembrane segments similar to vesicular monoamine transporters (Erickson et al. 1996). At present, histamine is thought to be transported by monoamine transporters. However, the Ki value of histamine for the monoamine transporter was 4.7 mm, and was about 50 times higher compared to the Ki value of 5-hydroxytryptamine, dopamine, and norepinephrine (Erickson et al. 1996). Histamine may be more efficiently transported by the polyamine transporter than by monoamine transporters. The polyamine transporter on synaptic vesicles was also inhibited by reserpine but not by methamphetamine. Together with its sensitivity to histamine, these observations suggest that the polyamine transporter in synaptic vesicles may have some features in common with the monoamine transporters.

Our results, taken together, suggest that there is cycling of polyamines in the brain, and that polyamines, in particular spermine, could function as neuromodulators (Fig. 7). Spermine accumulates in synaptic vesicles and is released from brain tissues (presumably from neurons) under depolarization. Spermine released from neurons is efficiently taken up by neurons and glial cells. Through circulation of polyamines, intracellular polyamines block Kir channels, and Ca2+-permeable AMPA/kainate receptors and extracellular polyamines may modulate NMDA receptors. There are also reports that NMDA-induced neurotoxicity due to Ca2+ influx is inhibited by α-difluoromethylornithine, an inhibitor of polyamine biosynthesis (Markwell et al. 1990; Kish et al. 1991).

Figure 7.

Model of polyamine circulation. Polyamine transporters expressed in synaptic vesicles, neuronal cells and glial cells each have different properties. Polyamines, especially spermine, are accumulated in synaptic vesicles and are released by depolarization. Intracellular polyamines cause block and rectification of Kir channels and some subtypes of AMPA/kainate receptors. Extracellular polyamines may modulate NMDA receptors.

Acknowledgements

This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan, by the Pharmacological Research Foundation, Tokyo, by the Futaba Electronics Memorial Foundation, Chiba, Japan and by the National Institutes of Health grant R01-NS35047.

Ancillary