Failure to demonstrate retrograde labelling of cerebellar Purkinje cells after injection of [3H]GABA in Deiters' nucleus

Neurosc/ence Leue~'s, 38 (1983) 23-28 Elsevier Scientific Publishers Ireland Ltd.

23

FAILURE T O D E M O N S T R A T E ~ O G R A D E LABELLING OF i CELLS ~ i N J E C T I O N OF ["~H]GABAI N

DEffF_ ii NUCLEUS L. WIKLUND*, H. KONZLE and M. CUI~NOD

Brain Research Institute, August.ForeI-Strasse I, CH-8029 Ziirich (Switzerland) (Received October 22rid, 1982; Revised version received and accepted April 6th, 1983)

Key words: GABA - transmitter specific retrograde labelling - cerebellum - Purkinje cells - terminal uptake mechanisms After microinjections o f IJHIGABA into the dorsal part of the Deiters' nucleus no retrograde labelling could be observed in the cerebellar Purkinje cells projecting to this nucleus. Tentatively, this failure to demonstrate transmitter specific retrograde labelling is altributed to lack of high affinity uptake mechanisms in the GABAerwc Purkinje axon terminals. These observations indicate a very important role of terminal uptake mechanisms in initiation of transmitter specific retrograde labelling. Anterograde axonal labelling was observed along several tracts originating in the injected area, but was not limited to connections with GABA as likely transmitter candidate.

Transmitter selective retrograde labelling has recently been proposed as an autoradiographic method to identify different amino acid and monoamine neuronal systems [5,21,22]. After injections of [3HIGABA (-r-aminobntyric acid) in rat substantia nigra, Streit and collaborators [21,22] observed selective labelling of the presumed GABAergic striatonigral pathway, suslgestinl~ that the method could be used also for this inhibitory amino acid transmitter. Selective retrograde labelling with [3H]GABA has also been observed in the pigeon isthmo-tectal [10,111 and thaiamo-hyperstriatal [21, as well as the cat subthaiamo-pallidal projection [131. But for none of these connections the transmitter specificity could be corroborated by independent neurochemical data. The projection of Purkinje cells located in the lateral vermis of cerebellar anterior lobe to the dorsal part of the lateral vestibular (Deiters') nucleus [4,24] represents one of the best established GABAergic connections in the brain [3,6,7,14,15,17]. We report here the failure to obtain retrograde labelling of the Purkinje cells projecting to Deiters' nucleus with [3H]GABA, and discuss the implications of this finding for the concept of transmitter specific retrograde labelling. Fifteen female albino rats of the SIV strain (200-220 g body weight) were anaesthetized with Nembutal (40 mg/kg; i.p.). Stereotaxic microinjections of 25 ~Ci [3HIGABA (4-amino-n-[2,3-3Hlbutyric acid, Amersham, 57 Ci/mmol) in 50 nl aqueous solution were delivered into the dorsal part of Deiters' nucleus with a ther*Present _a44ress for all correspondence: Department of Histology, Biskopsgatan 5, S-223 62 Lund, Sweden 0304-3940/83/$ 03.00 © 1983 Elsevier Scientific Publishers Ireland Ltd.

24 moinjec~or over Iq min through micropipettes (circa 50 pm outer diameter), which were left in situ for 5 rain after completing the injection. After a su~-ival of 5,6,12 or 24 h, the an;.mals were reanaesthetized and fixed by perfusion of 3.5 or 5~e glutaraidehyde in 0.17 M phosphate buffer, pH 7.4. The dissected brains were stored in the same fixative overnight, and then transferred to 30¢e sucrose in the same buffer+ After s~nking ( I - 2 days), the specimens were sectioned on a freezing microtome (30 pro), and the sections mounted on gelatinized slides. Following defattint through a ~rics of alcohols and xylcne, and d~+ing, the section bearing slides ~ere dipped in Kodak NTB-2 emulsion (diluted ! : I with distilled water). After 3 or 6 ~eeks exposure, the autoradiograms were developed in Dektol, fixed and rinsed. and stained with cresyl violet. The results were evaluated by bright- and darkfield m i c r o c o p y . Injectio~ ~ites appeared a~ a relatively well circumscribed zone of highly increased labelling. ~ho~¢ extent most easily could be evaluated by dark-field microscopy (Fig. l). I+h¢ center of injection x,~as always well localized within the dorsal, magnoccilular part of Deiter~" nucleus, but the diffitsion of the tracer often involved parts of the superior ~estibular and interposed cerebellar nuclei. The penetration of the micropipette had passed through the anterior VI Iobule and caudalmost V lobule. Along the tract a limited ~pread of tracer could be observed, but this did not intcrt+ere ~ith the microscopic in~+estigation of other parts of the anterior or l~O~teliof lobe ¢¢feb¢l|tnll. In all animai~, the cerebellar cortex was carefldly sur~cy,d for signs of retrograde neuronal labelling+ Ilowc~er. after none of the survival times was there any evidence of labelling of I'urkinje cells lot any other taenrone~ located in any part of the cetcbellar ¢orwx) (Fig. 21. Similarly, tire brain~lein from medulla oblongata ~o me~cnCel~l+ahm~a~ JllkeMJgated without e~idence of relt ogtade labellint~ of any of the oth¢l ~lfl'elcllt soilrce,, to the I)eiter~' nucleus (see tel. 12). lit colltr;.IM to the lack of t elrograd¢ neuronal labdlmg, sigt~s of antefoglad¢ labdting ~.,,er¢ regularly observed in the efferent projections of Ih¢ nuclei involved in the I'ltlGABA injection. Elms. in all cases, labelled fibres emerged from the injection site (Fig. 3) and ran in a ~+entr,~-medialdirection towards the paramedian r,.'ticular formation. ['here the,," fiwmcd a bundle of cross-sectioned axons, which could be traced caudally in a position above the inferior olive and, alter longer suri~al times, into the ~entral spinal white matter, l'he.~ fibres were interpreted as c o l responding to the lateral vestibuiospinai In~ct [2.1]. In ca~es where the injections in~ol~cd the ~uperio~ vestibular nucleus. ~trongly labelled ascending fibres were observed in the medial longitudinal fasciculus, and apparent terminal labelling of high density x~as observed over the trochlear nucleus (Fig. 4), while the oculomotor nucleux dispktyed sparser labelling. After injections involving the interposed nuclei, labelled fibres ~et+e pre~ent ~ithin the superior cerebellar peduncle and could be traced to~ards their decussation in mesencephalon, as expected for these presumed ccrebeilofugal fibres (of. refs. 9 attd 23). The most important resuh of the present study is the demonstration that Purkinje

25

Fig. 1. Dark-field micrograph of a |~H]GABA injection site centred on the dorsal magnocellular part of l)t iters' nucleus (arrow). Survival 6 h. Abbreviations: FN, fastigial nucleus; IN, interposed nucleus; I+N, iatera~ nucleus; RB, rest(form body; V, ventricle. Bar -= I ram. Fig° 2. Dark-field micrograph of a section lhrough the anterior lobe of the cerebellum (~ame experiment a~ m Fig. 1). No retrogradely labelled Purkinje cells are to he found in the lateral vermi~ (arrows). Ab+ bre~+iations: III, third Iobule; IV, fourth lobute; V, fifth Iobule of cerebellar vermin. Bat -=0.5 ram. Fig. 3. Anterogradely P,~elled fibres (arrows) of the lateral vestibulospinal tract (LVS;) emergiag from the [+HIGABA injection site (above left}. Survival 5 h. Abbreviation~: G VII, ge~u o~ f~cial aerve; V, ventricle. Bar 0.5 ram+

cells ~¢re not labelled retrograde|y with [-*HIGABA. which was administered in ~helr ~erminal area at a concentration and amount pre~-ionsly found by Streit and co-workers |21.22] ~o result in labelling of striatonigral GABAergic neuroses. At firs~ sight, this observation seems to speak against the concept of transmitter specific retrograde labelling, at least for GABAergic systems. However, in a series of careful investigations. Storm-Mathisen [19.20| studied the distribm~on of high aff/ni~" uptake for GABA in different GABAergic connections. Transcctions of the ,~rimonigral pathway was found to r~csult in marked reductions to similar magni~ad~.~ of both g'mamic acid dccarboxylase activity {i.e. of the GABA syn~hetizing enzyme| and the high affinity uptake of GABA in the substantia nigra. In ~l,e dorsal part of Deiters" nucleus, however, cutting of the Purkinje cell afferent |uput brought abom a drastic decrease of gimamic acid decarboxylase activity ~ i~hout significantly changing the high affinity uptake of GABA in the tissue. (The ~ormal GABA uptake was relatively low in dorsal Deiters" nucleus, making it unlikely ihat a lesion induced change could have been obscured by a high background uptake.) Storm-Mathiscn concluded that. contrary to the situation in

|-L~. 4. A~,.~ a~ ~ c ~ o I ~ of |~H~i~%~ c e ~ e d on ~.i~¢rs" nucleus but spreading to involve abo ~h¢ ,upc~o~ ~c s ~ | a ~ ~ u c k ~ q| ~ h ~r~i~a|L amerogradeL~ |a~eiled t'ibre.~can be followed along the medial

27

the striatonigral system, the reuptake mechanisms for GABA are not concentrated in the Purkinje cell axon terminals. Taken together with the autoradiographic results of Streit and collaborators [21,22] and those of the present investigation, these observations indicate that terminal uptake mechanisms may play a pivotal role for the initiation of transmitter selective retrograde labelling. Independent support for this hypothesis comes from observations in serotoninergic connections, where selective retrograde labelling with [~H]serotonin can be blocked by pharmacological inhibition of serotonin high affinity uptake [1]. Clearly, the lack of transmitter reuptak¢ mechanisms demonstrated by some neurones utilizing amino acid transmitters has to be taken into account in studi,..~ employing specific retrograde tracing techniques, and failure to observe retrograde labelling alone does not allox~ us to exclude that the substance tested may function as a transmitter in the investigated connection. The anterograde autoradiographic labelling observed in the course of these investigations involved several projections: the lateral ~estibulospinal tract, as well as ascending projections from interposed cerebellar and superior vestibular nuclei. There is no evidence in the literature to suggest GABA as a transmitter of the lateral vestibulospinal or the efferent cerebellar projections. Indeed. the excitatory nature of these connections 18,251 argues against mediation by this usually inhibitory transmitter. For the projection of the superior vestibular nucleus to the trochlrar nucleus, however, there is compelling neurochemical a~d physiological evidence for GABA serving as a transmitter 116,181. Taken together, these obser,,.ations indicate thai anterograde labelling after injection of I~ttIGABA is relatively unspecit'ic. IIo~'ever, as originally suggested by Ihntl and Ktin/le [llJ], it may be that GABAcrgi¢ systems like the ~,estibulo-troehlear connection demonstrate a more pro. noun¢¢d labelling due to particular affinity to their ov,n transmitler. This stud)' x,-as supported by the Swiss National Foundation lGrants 3..~0~.79 and 3.506.791 and the Dr. Eric Siack-Gyr-Foundalion. We thank Dr. J. Slorm-Mathisen for useful ¢ommenls on the manuscript. For experl technical assislanc¢ we ~armly thank Armin l-~ih (surgery), Marianna Slella, Ruth Rich (hislology), Johann Kiin~li I,',hotography) and ingrid Stedman (typing). ! ,~,tancda, S., Itobdlicr, P.. Buda, M. and Pujol, J.-l:., Retrograde axonal trampoi; Iollossntl~ mic~' Iron,, of I'ltlsetotonin in the olfactory bulb. I. Ih~:hrmic:d slud~, llram Re',, 196 I lgS(t) 411.~ .~i ~. 2 Ba~noli. P., Burkhallcr, A., +~llt'il, P. afld ('uenod..~J., I ' I I I ( ; A B A r¢lrograde labeling +,t +leuton~ n+ |he pigeon Ihalamo-Wul+,,I pathmay, Ar¢l+. ilaL B i o l , 121 11914.tl 4 +~ ~1. :t ( ban Pala~. V+, Nila+er. ( L , Palay, S.I .. Behlf¢id. X " ' /.immcrman, L A+. ++~u. J ~. ai+d { }'l}onohue, I . I . , ChemJ,:al I;elerogcnilv its ¢¢rebella~ P'u; kmje cell,,: i-',,i~,t,.'n¢¢ and ~.IeXI~,IL'iI~L"~.t| l.'lutanlk' ~h.'id d~.'¢albOX~|~,,¢-like and illolllia|, like imnmn,,,e,~cii~ ,tic,,, P,oc. nat. Ac:,d..~cl. U . ~ .X.. "1~ t191';11 ""~I? "7"7,tl. 4 (of~.ala. N. and Pomp¢iam~. O.. Idrntifieatton of ¢¢,¢bella[ et~rli¢o~¢*,lll~ular n~:urou,, rt:lrog;ad¢l'~ labeled ~ilh hmseradi~h pero~lda~¢, Ncuro~c~en¢¢, 4 t 1979) ~tf' .~1.~.

5 Cmhm~ M., Balpm~ P+, Beaudet, A., Rustioni, A., W'ddumi, L. and ~reil, P., Tratmnitt~ specific getrolpade ~ of neurons, in S.L. PalaY and V. Clum-Palay 0Eds.), Cymchemical Methods in Neuroanatmny, Alan R. Liss. New York, 1982, pp. 17-44. 6 c,,i~ DX, ~A.W. a d Fd~,. O.. GABA a,d i ~ k m of t m ~ ' m~rone~ main m ~ . 23(1970) I I?'l~l). "I Foanum. F,, Storm-Madm~ J. and Wallm'~ F.. C.dmam~ decasbo~71~ in m~'biu~ nom'ons. A study of d~ enzy~ in Purkinje odl axom and Bamons in the c~, Brain P~.. 20(1970) ~9-270. $ Gn~emr, S., HonlP, T. mid Luad, S., The v ~ tract. Effects on alpha-mmonemmms in the Imnlmmcral cord in the cat, Exp. Brum P,es., l0 (1970) 94-120. 9 Hawoian, A.L, Mm~pmt, L.C. and Youe~ P.A., ~ projections in the rat: an amoradiolprapJ~ and dqlene~ion study, L comp. NeuroL, 197 (1981) 217-236. 10 Hum, 5.P. and Kfmzke,H., Selectiveuptake and transport of label within three ideutifkd neuronal systems after injection of I~HIGABA into the p J ~ optic t e a m : an amoradiograpltic and GOtlli study-. J. comp. NeuroL. 170 (1976) 173-190. ! I Hint. S.P.. Strdto P., Kunzlc, H. and CuC-nod, M.o Characterization of the pigeon isthmo.tectal pathway by- 5~'lect~-euptake and retrograde movement of radioactive compounds and by Gollli-like HRP, Brain Rcs+, 129 (1977) 197-212. 12 Ito. J., Sasa. M+, Matmok~, I. and Takami. S, Afferent projection from reticular nuclei, inferior olive and cerebellum to lateral vestib~dar nucleus of the mt as demotmraled by horseradish peroxid,x~, Brain I~'s.. 231 (191r~)427-432. 13 Nauta, H,LW, and C~nod, M,, Perikaryal c:ll lahelling in the subthalami¢ nucleus following the injection of {JHIGABA imo the pallidal complex: An autorndiolgraphics, mly in cat. Neurogienc¢, "/(!~2) 2725-27J4, 14 Obala. K+. Ito, M . Ochi, R. and Sam~, N.. Phannacotogical pgopenicl of the ~ p t i ¢ inhibition by Purkinje cell as~'qJ and the action of y-aminolmtyric acid on Deit~s' nouronl, E~p. Brain Rcs.. 4 (19fi7) 43-57. 15 Oerlel, WJt., Schm¢chel, U.E., MulP~ini. E.. Tapp~, M.L. and Kopin, LJ., Immunocytochemical localization of lllulamte decarbosyl~ in rat c e r ~ l u m with a pew amid'rum, N~uro~euce, 6 (191]1) 271~27~5. 16 Prechl. W,. Baker, R, and OkmlM, Y,, Evkknce for GABA a~ the ~yn~plic transmiller of Ihe illhibitory vc~dibulo-ocular pathway. Exp, Brain Res.. 18 (1974) 415~4,~;, 17 Ribak, C.E.. Vanghn. J.E. and Saito, K., lmmunocytochemical Iocali~,ationof Slutami¢ acid d¢carbo~ylase in neuronal ~ n a l a follo~inlg colchicine inhibition of axonal tran~pmt, Brain R~., 140 (1978) 315-332. 18 Roffler-Tarlow, S. and Tarlow. E,, Reduction of GABA synthesis following lesions of inhibitory vestibulo-trochlear pathway, Br~n Res., 91 (19"/5) 326-330. 19 Storm-Matbi~.n, J., Hish afemity uptake of GABA in premmed GABAerlgic nerve endin~ in rat brain, Brain Res,, 84 (t97¢,/409-427, 20 Stmm-Mathi~eu, J., Distribution of the components of the GABA system in neuronal tissue: cerebellum and hipp~ampus - effects of axotomy. In E. Roberts. T.N. Chase and T.B. Tower (Eds.), GABA in Nervous System Function, Raven Press, New York. 1976, ~,p, 149-168. 21 Strcit, P,, Selective retroiwdde labeling indicating the trar, smitter of neuronal pathways. J. comp, Neurol,, 191 (19~0) 429-463. ~.2 Streit, P.. Knecht, E, and Cub'nod. M.. Transmitter-specific retrograde lab~linlgin the striatonigral and raphe-nisral pathways, Science. 205 (1979) 306-308. 2,3 Verhmlrt, WJ.C., A Stereotaxic Atlas of the Brain Stem of the Cat, Parts ! and Ii, Van Gorcum, A . e . , tg~: Voogd+ J,, ~ e Cerebellum of the Cat, Van ~ r c u m , Asseu, 1967, 215 pp. 25 Yosbida, M, and Hirma, i., The mode of cerebello-thalamicrelay transmission investigated with intracellular recording from cells in the ventrolateral nucleus of cat's thalamus, Exp, Brain Res., i0 (1970) 121-139,