Non-exocytotic GABA overflow in rat striatum inhibits gnawing

Life Sciences.,Vol. 61, No. 16, pp. 1593-1601, 1997 CopyTight0 1997 Ekvier science Inc. Printed in the USA. Ail rights resewed om4-32ns/!G7 $17.00 + .cKl

ELSEVIER

PII soo24-3205(97)00738%

NON-EXOCYTOTIC GABA OVERFLOW IN RAT STRIATUM INHIBITS GNAWING Kelly L. Drew’, Terri Fitka’, Yong Hu’ and Urban Ungerstedt* ‘Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, Alaska 99775-7000 USA ‘Department of Physiology and Pharmacology, Karolinska Institute, S 17l-77, Stockholm, Sweden (Received in final form July 14,1997)

The present study tested the hypotheses that spontaneous y-aminobutyric acid (GABA) efflux in anterior rat striatum is 1) independent of intra- and extracellti calcium; and 2) is physiologically relevant. Extracellular dopamine (DA) and GABA were sampled from striatum of awake, freely moving rats using in vivo microdialysis. Although dialysate concentrations of DA were 2 to 3 times greater than GABA and were decreased by at least 70% by removal of calcium, GABA was unatTectedeven in the presence of EGTA or the intracellular calcium chelator APTRA-AM. Functional signiticance of this nonexocytotic pool of GABA was tested by injecting 3-mercaptopropionic acid (3-MPA), an inhibitor of GABA synthesis, into the striatum via a guide cannula sidled alongside a microdialysis probe and measuring subsequent effects on behavior and perfusate concentrations of GABA. Results show that 3-MPA increases gnawing behavior suggesting that basal, non-exocytotic GABA overflow notmally functions to suppress gnawing, key W&S: e&nm, basal release, m&dialysis, microinjection,exocytosis,spontaneousefflux,GABA In anterior rat striatum basal GABA sampled usiig in vivo microdialysisappears to be resistant to decreases in extracellular calcium (1,2, 3,4, 5). Yet, at the same time, basal GABA appears to play a functional role in the striatum. GABA antagonists injected into the striatum produce an increase in stereotypic behaviors and locomotion depending on the location of the injections (6). One purpose of the present study was to determine if extracellular GABA is resistant to decreases in extracellular calcium because of liberation of intracellular calcium pools. This was accomplished by including an intracellular calcium chelator (APTRA-AM) in the calcium free pet&ion fluid. Another purpose was to test the hypothesis that basal, in vhw, efflux of GABA in anterior rat striatum maintains a tonic level of physiologically sign&ant inhibition. We tested this hypothesis by injecting 3-mercaptopropionic acid (3MPA), an inhibitor of GABA synthesis (7, 8), into anterior rat striatum through a removable injection cannula sidled alongside a microdialysis probe. Behavioral effects were recorded and microdialysis samples were analyzed for GABA. Results indicate that 1) extracellular GABA is independent of both intra and extracellti caltium; and 2) that this pool of GABA tonically inhibitsgnawing behavior. These results suggest that a novel, non-exocytotic mechanism exists which maintains a physiologically relevant pool of extracellular GABA. Correspondence: Kelly L. Drew, Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, AK 99775-7000, (907) 474-7190; e-mail: [email protected]; Fax: (907) 474-6967.

1594

Non-exocytoticGABA InhibitsGnawing

Vol. 61, No. 16, 1997

Materials and Methods

Microdialysis probes were stereotaxically implanted into the right and lefi anterior striatum of 350 to 400 g male, Sprague-Dawley rats under halothane anesthesia. In the first experiment CMA/l% probes (Stockholm, Sweden) with 2.0 mm x 0.5 mm dialyzing tips were implanted 0.6 mm anterior and 3.5 mm lateral to bregma and -5.5 mm ventral from brain surface (9). In the second experiment probes were constructed using GFSOM regenerated cellulose cuprophan dialysis membrane with a 50,000 MW cutoff (Groningen, The Netherlands) and had dialyzing tips of 3.0 mm x 0.22 mm. These probes were implanted slightly more medial to accommodate the longer length (AFW.6, L=3.2, D--6.5). Finally, for the last experiment, probes were constructed and positioned using the same coordinates as in experiment 2; however, guide cannulae aklixed to the probes were oriented medial to the probe and implanted as a dual barrel assembly (Drew et al., in preparation). Dialysis experiments, conducted in awake, freely moving animals commenced approximately 40 hrs after surgeq. Probes were per&&d with a modifiedRinger’s (m-R) solution (1.2 mM CaCl2,2.7 mM KCI, 148 mM NaCl and 0.85 mM MgClz in experiment 1; with 1.4 mM glucose added in experiments 2 and 3); at a rate of 1.O p.l/min in experiments 1 and 2 and 0.1 pl/min in experiment 3. The lower flow rate was found to reduce variability in behavioral effects of Samples were microinjections made alongside a microdialysis probe (Drew et al., in preparation). collected at 15- or 30-min intervals beghming 1 hr a&r the onset of perlksion. Immediately after commencing flow through the microdialysis probes in experiment 3, an injection cannula, primed with either 3-mercaptopropionic acid (3-W& 80 pdO.5 ~1, pH 4) or vehicle (m-R adjusted to pH 4 with HCl) was inserted into either the right or lefi guide cannula. An injection cannula was inserted on the side with the best flow through the microdialysis probe. In several cases flow on the contralateral side was discontinued due to lower than expected flow rates. Consequently, contralateral effects of 3-MPA injections were not assessed. All procedures involving animals were approved by the University of Alaska Fairbanks’ Animal Care and Use Committee.

ExperimentI Aver two 30-min fractions were collected, the perfusion fluid was switched from m-R to a solution in which the calcium had been omitted or the calcium had been omitted and 0.1 mM EGTA (Sigma, St. Louis, MO) had been added for the duration of the experiment. One-third of each sample was then analyzed for dopamine and the r emaining sample was analyzed for GABA.

Epn’ment 2 After two 30-min baseline samples were collected, the perfbsion fluid was switched to an experimental solution or control. In the experimental group the m-R solution was altered by omitting calcium and adding 0.1 mM 2-aminophenol-N,N,O-triacetic acid, tri(acetoxymethy1 ester) (APTRAAM; Molecular Probes, Eugene, OR). APTRA-AM was dissolved in DMSO giving a final concentration of 1% DMSO, and DMSO (IO/o)was included in the m-R in the control group.

Experiment3 Afler two 15-min fktions were collected, 0.5 pl containing 80 pg of 3-MPA (Sigma, St. Louis, MO) was injected at 0.25 @min using a CMA/lOO microsyringe pump (Stockholm, Sweden). Gnawing was recorded by an observer during the experiment while locomotor activity, rearing, shaking of the body and head and jerks of the head and neck, forelimbs, trunk and hind limbs, as well as circling behavior, were recorded from video tapes. Microdialysis probes were perfused at a rate of 0.1 flmin in one-half of the control group and at 1.O Crymin in the other half As no statistically significant diierences were found in the behavior of control rats perfked at the two flow rates, behavioral data for these groups were combined, however, because perfhsate concentrations of GABA were significantly less concentrated at the higher flow rate, they were not included in the analysis.

Vol. 61, No. 16, 1997

Non-exoqtotic

GABA Inhibits Gnawing

1595

Histology At the end of each experiment animals were sacrificed by an overdose of pentobarbital. Brains were removed and sliced on a cryostat. Sketches were made of the positions of the microdialysis probes following examination of the frozen tissue. To further veritjl the location of 3MPA injections in 4 animals from experiment 3, injection cannulae were removed, primed with thionin and minserted. Slices of these brains were mounted on slides and Thionin was injected exactly as descrii for 3MPA stained with hematoxylin-eosin. Lkymmine assq Dopamine was analyzed by HPLC with electrochemical detection as described in Drew et al., (3); the limit of detection of this assay was 0.02 pmol dopamine. GABA awy GABA was analyzed usiig HPLC with electrochemical detection and pre-column derivitization with OPA and t-butyl thiol as described by Kehr and Ungerstedt (10) or by a modified version of this method adapted for use with an ESA 5 1OOAelectrochemical detector. In the latter case, the apparatus consisted ofanHPLCpump(Waters,510),aRheodyne7125injector,anucleosilC18,3Cun, 1OOx4mmcolumn with a 10 x 4 mm guard column (Keystone Scientific, Inc.), an ESA 5 1OOAelectrochemical detector with a 5011 analytic cell and a Spectra-Physics (4270) integrator. Mobile phase wnsisted of 50 mM sodium acetate, 1.0 mM EDTA, pH 5.4 and 51% acetonitrile and was pumped at 0.8 ml&n. The working electrode was set at a potential of 0.40 V. The OPA/t-butyl thiol reagent wntained 2OmM OPA (dissolved in 0.1M NaHC03, pH 9.6, 50% methanol) and 2% t-butyl thiol (Fluka, St. Louis, MO) and was added to samples in a 1: 10 volume ratio. Small volumes (1.5 pl) collected at 0.1 ul/min were diluted with water to a tinal volume of 15 pl. After addition of reagent, samples were mixed at room temperature for 80 set and injected. To determine the effect of 3-MPA on the chromatograms, standards were diluted in m-R containing lOA to 10’ M 3-MPA. GABA was quantified by comparing peak height to an externalstandard. StOtiStiCS

Per&sate wncentrations of GABA and dopamine in experiments 1 and 2 were analyzed using 3-way ANOVA (Crisp Crunch Software) with treatment designated as a between subjects variable and side of brain and time of sample collection treated as within subjects variables. Significant ANOVAs were followed by Newman-Keuls post-hoc wmparisons or t-tests where indicated. All p values refer to the interaction between treatment and time unless otherwise stated. Results from the tirst 2 experiments are A similar analysis was performed on GABA shown as the means of the left and right sides wmbiied. data in experiment 3 except that only samples wllected on the side of injection were analyzed so side was not included in the analysis. All behavioral data recorded in experiment 3 were analyzed by Km&alWallis test of 3 groups. A significant ditference between groups was followed by a Mann-Whitney comparison between each 3-MPA group and controls. Results Figure 1 shows 1) that per&ate concentrations of GABA are 2-3 times those of dopamine and 2) that spontaneous efflux of GABA and dopamine is di&rentially atfected by decreases in extracellular calcium. While dopamine was reduced at least 70% to below our 1.O nM limit of detection, GABA was not at&&d by removal of calcium from the per&ion fluid even in the presence of EGTA Figure 2 shows that removing calcium and adding 0.1 mM APTRA-AM to the perfbsion medium, likewise, had no signiticant et&t on perfbsate concentrations of GABA Gnawing behavior was significantly increased compared to the vehicle group 15 min following intrastriatal injection of 3-MPA (pcO.02, Fig. 3). Per&ate wncentrations were wnwmitantly reduced

15%

3025 -T

20 --

i c

1 5 --

‘;;

lO--

O---O

M-R

O--O

no Ca2+

n---n

no Ca2+,

Vol. 61, No. 16, 1997

Gnawing

GABA lnhiits

Nowmcytotic

T 3o 0.1

mM

EGTA

M 2

5 -0 --

-60

-30

0

+o

-_-----------_-_

/--------30

60

t

120

90

150

180

210

TIME (min)

Fig. 1 P&sate concentrations of GABA (solid lines) and dopamine (dotted lines) expressed as means +/- SEM are shown for 2 samples collected before and 6 samples collected during perfusion with the indicated solutions. Samples collected from right and letI striatum were comb&d for a total of 4 to 8 observations per group. O-O

no Ca2+

O--O

Vehicle

, 0.1

mM APTRA-AM,

(l%DMSO

in M-R),

n=5 n=5

30 25 f‘ \f 2 s

20 15 ,.

-5t:

-60

I -30

:

1 I 0

I

I 30

:

I 60

:

/ 90

:

I 120

:

; 150

:

1 180

:

, 210

TIME (min)

Fig. 2 P&sate concentmtions of GABA (mean +/-SEM) before, during and a&r perfbion with 0.1 mM APTRA-AM dissolved in m-R solution in which calcium has been omitted, or vehicle (Loo/o DMSO) diluted in m-R containing 1.2 mM calcium. Hokontal line above the x-axis indicates duration of perbion with APTRA-AM or vehicle solution.

1597

Non-exocytotic GABA Inhibits Gnawing

Vol. 61, No. 16, 1997

following 3-MPA injection. Unfortunately, GABA could not be analyzed in one tkxtion collected 15 to 30 min following the injection due to interference of 3-MEA. In standards, lo-’ and lo4 M 3-MPA produced a large negative front on the chromatogram siiar to that observed in samples collected 15 to 30 min following 3-MEA injection; the front was absent with 1W’and lOa M 3-MEA. 3-MPA (Iu’ and lOa M) did not a&ct peak heights of standards (0.3 pmol GABA analyzed in the presence of l@’ or 10" M 3-MEA yielded a mean (+SEM) concentration of 0.33kO.05 pmol, n=4); and therefore, appeared not to atfect quantification of GABA at concentrations that did not produce negative fronts on the chromatograms that obscured the GABA peak. C

A

20

20-

17

3-MPA

,5--

0

5 5

IO--

2 0 5 --

t

B 200 175 2 5 2 2

150 17.5 100

*

15

I

50 mtrring

25 0

data’ 0

15

30

45

TIME (min)

60

75

I 90

125 100 75 50 2s 0 0

15

30

45

60

75

90

TIME (min)

Fig. 3 Effects of local injections of 3-MEA (A and B) and vehicle (C and D) from 3 l-33 min (mdicated by arrow and line above x-axis) on gnawing behavior and pertkate concentrations of GABA. Gnawing is expressed as medians with 15th and 85th percentiles shown above each samplingperiod (3-MEA, n=7; controls, n=14). GABA data are shown as means +/- SEM (3-MEA, n=5; controls, n=7). ‘Data not included in analysis because 3-MPA interfered with GABA assay in 3 out of 5 samples; + FO.05 compared to controls; * l60.05 compared to first interval. In the sample collected in the following 15-min fraction, per&ate concentrations of GABA were reduced by approximately 60%. Note that petfixate concentrations were higher in samples collected at 0.1 pknln(~ t 3) than at 1.Ouknin (experiments 1 and 2) due to higher relative recovery at the lower flow rate. Figure 4 shows chromatograms illustrating the effects of 3-MEA on perILsate concentrations of GABA.

1598

Non-exocytotic

GABA Inhibits Gnawing

Vol. 61,No. 16,1997

Fig. 4 Chromatograms showing from left to right, a standard of 0.3 pmol GAFM, a sample collected during the 3-MPA injection and samples collected during the next 2 sampling periods.

Examination of cryostat sectioned brains showed that all microdialysisprobes were placed anterior to the decussation of the anterior commissure. Probe placement was normally distributed between 0.2 and 1.0 mm anterior to bregma with the greatest frequency of probes local&d approximately 0.6 mm anterior to bregma (9). Injection sites were consistently located medial and dorsal to the most ventral aspect of the probe tract as predicted by the dimensions of the dual probe design and length of injection csnnula (Fig. 5). Discussion

Results reported here suggest that basal GABA overflow in anterior rat striatum is both nonexocytotic and physiologically signitlcant. Extmcellular GABA is resistant to decreases in calcitmr even when an intmcehular calcium chelator is included in the perfusion medium. APTRA-AM (50 @vi) markedly attenuates evoked synaptic responses (11). The concentmtion of AFTRA-AM applied via microdialysisin the present study was twice the effective in vitro concentration. Although the possibility exists that we did not achieve effective intmcehular concentrations in vivo, this is unlikely given the calchm~afhnity, binding rate and selectivity of APTRA-AM (11). other reports of reduced calciuminduced decreases in dialysate concentrations of GABA (12, 13) have been shown previously to be due to higher than physiologic concentmtions of calcium in the pretreatment perfusion fluid (5, 14). In areaa of the brain outside of the striatum where GAsAergic neurons are more tonically active some studies

Vol. 61, No. 16, 1997

Non-exocytoticGABA InhibitsGnawing

1599

suggest extracellular calcium plays a role in basal overflow (15, 16, 17) while others do not (18, 19). Failure for ATTRA-AM to atkct GABA overflow in rat striatum argues against the possibility that liberation of intracellular calcium maintains GABA overtlow when extracellular calcium is reduced.

Fig. 5 Dii showing position of microdialysis probes and injection cannula in anterior striatum. A physiological role for this non-exocytotic, extracellular pool of GABA is supported by other studies showing that intrastriatal administration of GABA antagonists alter behavior (6) and enhance acetylcholine release (20). Increases in gnawing behavior following 3-MPA injection in the present study are likewise consistent with a physiologic role for this pool of GABA. Gnawing behavior appeared to increase within 10 min of injection and persist for up to 25 min; however, the increase was statistkahy sign&ant only for a brief period 15 to 18 min following 3-MPA injection. Delay of behavioral response could be due to diision of 3-MPA or to GABA turnover rate. Considering the location of the injections, it is unlikely that di&sion of 3-MPA outside the striatum accounted for the behavioral effects. Bilateral injections of the GABA antagonist picrotoxin slightly posterior and ventral to the present site, was found by Scheel-Kruger (6) to induce stereotyped head movements, s&&g, and episodic licking/bitingactivities. Delay in the behavioral response is more likely explained by the rate of GABA turnover. Intraperitoneal injection of 3-MPA (9Omgkg) was shown to reduce whole tissue concentrations of GABA in rat striatum to 72,67, and 48% of control 4, 7 and 15 min following ip administration (8). The dose of 3-MPA administered here is comparable with other microinjection studies (21). Although 3-MPA is also known to inhibit GABA transaminase (7), the present results suggest 3-MF’A decreases per&sate concentrations of GABA. Unfortunately difkion of 3-MPA into the perfhsate 15 to 30 min atIer the injection, when 3-MPA concentrations were likely highest around the probe, interfered with the GABA chromatography. Thus, we were unable to determine maximum decreases in extracellular GABA and little was gained by injecting the drug next to the microdialysisprobe rather than including it in the perfbsion fluid. A better approach might be to use a chromatographic technique described by Kehr et al., (22) that utiis 3-MPA in the precolumn derivitizing reagent. More importantly, 3-MPA could have competed with t-butyl thiol in the precolumn derivitization of samples in the present study and reduced GABA peak heights. We saw no evidence of this when standards were prepared in concentrations of 3-MPA that did not obscure the GABA peak. Consequently results suggest that 3-MPA produced at least a 60 % reduction in extracellular GABA. other studies siiarly

1600

Non-exmytotic GABA Inhibits Gnawing

Vol. 61, No. 16, 1997

show that basal per&sate c.oncentratioflsof GABA sampled from rat striatum are decreased dramatically following systemic administration of 3-MPA (10) or excitotoxic lesions (23,24). In conclusion, an extracellular pool of GABA in rat striatum is resistant to decreases in calcium even in the presence of EGTA and the intracellular calcium chelator APTRA-AM (11). This same pool of GABA inhibits gnawing behavior. These results are consistent with a functional role for a non-exocytotic pool of extracellular GABA. Whether this pool is “released” via voltage dependent uptake reversal (25, 26, 18, cf. 27); is synthesized extracellularly (28,29), or originates tiom some other source has yet to be determined. The 60% decrease in pe&sate GABA reported here is greater than the 20% decrease produced by tetrodotoxin (5) susgesting that voltage dependent reversal of carrier mediated uptake cannot account for all of the calcium-independent GABA overflow. Other mechanisms of depolarization such as inhibition of ATP-sensitive potassium channels (30) could, however, drive voltage dependent carrier mediated release of GABA, a non-exocytotic process. Whatever mechanism proves to maintain spontaneous GABA overflow, results reported here suggest that this pool of GABA is physiologically significant and warrants fiuther investigation. Acknowledgments We gratefully acknowledge support from the National Institute of Aging (R55 AG09483) and National Science Foundation (IBN 9121221) to K.L.D. as well as support for T.F. from the National Science Foundation, Alaska Native Student Internship Program (OCE-9016113). We would like to thank Dr. Peter G. Osborne for reviewing the manuscript, MS Ahna Davis for secretarial assistance, and Dr. Fatima Matos for her gifl of cuprophan membrane.

References 1. J.A GIRAULT, U. SPAMPINATO, H.E. SAVAKI, J. GLOWINSKI and M.J. BESSON, Neuroscience 19 1101-l 108 (1986). 2. B.H.C. WESTERINK, H.M. HOFSTEEDE, G. DAMSMA and J.B. DE VRIES, NauynSchmiedeberg’s Arch P-1. 337 373-378 (1988). 3. K.L. DREW, W.T. O’CONNOR, J. KEHR and U. UNGERSTEDT, Life Sci. 45 1307-1317 (1989). 4. B.H.C. WESTERINK and J.B. DE VRIES, Nauyn-Schmiedeberg’s Arch Pharmacol. 339 603-607 (1989). 5. K.L. DREW and U. UNGERSTEDT, J. Neurochem. 57 1927-1930 (1991). 6. J. SCHEEL-KRUGER, Acta Neurologica Scandinavica 73 (1986). 7. C. LAMAR, Jr., J. Neurochem. 17 165-170 (1970). and 5. STORM-MATHISEN, 8. A KARLSSON, F. FONNUM, D. MALTHE-SORENSSEN Biochemical Pharmac~l. 23 3053-3061 (1974). 9. G. PAXINOS and C. WATSON, m Rat Brain in Stereotaxi Coordinates. Academic Press, Inc., New York (1986). 10. J. KEHR and U. UNGERSTEDT, 5. Neurochem. 51 1308-1310 (1988). 11. I. SPIGELMAN, M. TYMlANSKI, M.C. WALLACE, P.L. CARLEN and A.A. VEL-, Neuroscience 75 559-572 (1996). J. Neurosci. 12. P.G. OSBORNE, W.T. O’CONNOR, K.D. DREW and U. UNGERSTEDT, Methods. 34 95-105 (1990). P. KALEN, C. LUNDBERG, K. WICTORIN, E. ROSENGREN and A. 13. KCAMPBELL, BJORKLUND, BrainRes. 614 241-250 (1993). 14. B. MOGHADDAM and B.S. BUNNEY, J. Neurochem. 53 652-654 (1989). 15. AJ. BOURDELAIS and P.W. KALIVAS, J. Neurochem., 58 231 l-2320 (1992).

Vol. 61, No. 16, 1997

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30.

Non-exocytotic GABA Inhibits Gnawing

1601

AE. HERBISON, RP. HEAVENS, RG. DYER, J. Nemo&m. 55 1617-1623 (1990). H.L. ROWLEY, K.F. MARTIN and CA MARSDEN, J. Neurosci. Methods 57 93-99 (1995). M.J. DURING, K.M. RYDERand D.D. SPENCER Nature 376 174-177 (1995). M. QUME,P.S. WHITTONandL.J.FOWLER, J. Neurochem. 642256~2261(1995). P. DEBOER and B.H.C. WESTERINK, J. Neurochem. 62 70-75 (1994). C.A. PEANO, C.A. SHONIS, G.H. DILLON and T.G. WALDROP, Brain Res. Bull. 28 107-l 13 (1991). J. KEI-IR, CMA Application Note #16, CMAMicrodialysis AB, Stockholm, Sweden (1993). S.P. BUTCHER and A HAMBERGER, J. Neurochem. 48 713-721 (1987). J.J. MEANA B. JOHANSSON, M. HERRERA-MARSCHITZ, W.T. O’CONNOR M. GOINY, F.E. PARKINSON, B.B. FREDHOLM and U. UNGERSTEDT, Brain Res. 596 65-72 (1992). S. BERNATH and M.J. ZIGMOND, Neuroscience 27 563-570 (1988). J.-P. PIN and J. BOCKAERT, J. Neurosci. 9 648-656 (1989). S. BERNATH, Prog. Neurobiol. 38 57-91 (1992). R TAPIA, Regulation of ghzmate decarboqhe activity in glutamine, glutamate and GABA in the central nervous system. L. Hertz, E. Kvamme, E. McGeer and A Schousboe (Eds.), 113128, Alan R Liss Inc., New York (1983). J.G. WOOD, B.J. MCLAUGHLIN and J.E. VAUGHN, GABA in Nervous System Function. E. Roberts, T.N. Chase and D.B. Tower @ds.), 133-148, Raven Press, New York (1976). M.J. DURING, P. LEONE, K.E. DAVIS, D. KERR and RS. SHERWIN, J. Clin. Invest. 95 2403-2408 (1995).