AF64A affects septal choline acetyltransferase but not parvalbumin immunoreactive cells

Brain Research Bulletin, Vol. 35, No. 3, pp. 217-220, 1994 Copyright 0 1994 Elsevier Science Ltd P&t& in the USA. Ail rights resewed 0361-9230194 $6.00 + .OO

Pergamon

AF64A Affects Septal Choline Acetyltransferase But Not P~alb~~n I~~~oreactive Cells XIN WEZN DONG

ISRAEL HANIN AND STANLEY

A. LORENS’

Department of Pharmacology and Experimental Therapeutics, Loyola University Chicago, Stritch School of Medicine, Maywood, IL 60153 Received 7 December 1992; Accepted 27 April 1994 DONG, X. W., I. HANIN AND S. A. LORENS. AF64A affects sepal choline acetylr~nsferase but not parvalbumin immunoreacrivec&s. BRAIN RES BULL 35(3) 217-220,1994.--Rats received bilateral in~~~~ven~~1~ (ICV) infusions ofeither AFS4A (I.5 ~~~~~~t~~~~;n = 9) or vehicle (3.0 ~~ven~~le; n = 7). Four weeks tater, the ar&mIs were anesthetized and their brains processed to visualize and quamify choline a~etyI~ansfe~e (ChAT) ~mmuno#active (IR) and p~~~~n-~R GABAergic neurons in the septal complex by irnrnun~yt~~rnis~ (PAP method). AF64A significantly reduced the num~r of ChAT-IR perikarya in the medial septum (28%), ventral limb of the diagonal band of Broca (30%). and horizontal limb of the diagonal band of Broca (20%). but did not affect the number of p~albumin-containing GABAergic neurons in any of the septal subdivisions. These results provide further evidence that AF44A is a selective cholinotoxin. Choline acetyltransferase AF64A Septal complex

Cholinotoxin

and GABAergic cell bodies are dis~but~ throughout the medial septal nucleus and diagonal band of Broca (2,4,8,10,11,15,17) and are the source of fibers that innervate the hip~~pa1 fo~ation (2,8,10,1 I, 17). Neurons ~on~ning the calcium binding protein, p~~bu~n, have a similar dis~but~on in the septal complex (1,5,g). It was recently demonstrated that parvalbnmin was coiocalized with GABA in these septal perikarya, and that the septal GABA neurons that project to the hippocatnpus also contain parvalbumin (&IO). Further, less than 1.O% of septal cholinergic neurons contain GABA (4,lO). Ethylcholine ~i~dini~ ion (AF@A) is a cholinotoxic agent that is taken up by the high-unity choline transport system and produces neural degeneration (6,9,15). When injected ICV, AF64A leads to dose-dependent reductions in hippocampal choline acetyltransferase (ChAT) and acetylcholinesterase activities, high-affinity choline transport, and acetylcholine 1eveIs (9,13). In addition, AF64A resnlts in the toss of ChAT ~mrn~o~ac~ve (IR) cell bodies in the septal complex (6,15). other brain regions are not affected by ICV treatment with AF64A (14). The objective of the present study was to determine whether AF64A selectively destroys septohipp~~p~ cholinergic neurons. If this is true, then the ICV admin~s~ation of AF64A should result in the loss of septal ChAT-IR cell hodies, as previously reported (6,15), but spare septal parvalbumin-IR GABA perikarya. We injected bilaterally a cholinotoxic dose of AFf54A (1.5 nmol/lateral ventricle) and counted the number of ChAT and parvalbumin-IR neurons in the septal complex 28 days postoperatively. AF64A led to sig~fic~t reductions (20-30%) in the C~~~~~~G~~

GABA

Immunocytochemistry

Parvalbumin

number of ChAT-IR but not of p~~bn~~-~ neurons in the medial septal nucleus and diagonal band of Broca. These results support the view that AF64A is a selective cholinotoxin for the septohip~~pal system. METHOD Animafs Male Fischer 344 rats (324-356 g at the time of surgery) were obtained from Harlan Sprague-Dawley Inc. (Indianapolis, IN). The animals (n = 16) were housed individ~~ly in an AAAL+AC approved facility, and maintained on a 12 L:12 D cycle. Food and water were available ad lib. Surgery Each rat was anesthetized with ~ntob~bit~ sodium (X, m& kg, If), then placed in a Kopf s&reotaxic instrument with the incisor bar set 3.3 mm below the interaural plane. Rats received bilateral ICV infusions of either AF64A (1.5 nmo~ven~cle~ n = 9) or vehicle (3.0 @ventricle; n = 7), We selected this dose of AF64A because in a previous study (15) it produced the greatest loss of septal cholinergic neurons. The vehicle and AF64A solutions were made fresh on the day of surgery, as previously detailed (13). The vehicle consisted of pH adjusted water and was prepared in the same manner as the AF64A solution (pH 7.4). The AF64A or vehicle solution was injected into each lateral ventricle using the following coordinates: 0.8 mm caudal, + 1S

’ Requests for reprints should be addressed to Dr. S. A. Lorens, Laboratory of Behavioral Pharmacology {Bldg. 135), Loyola University Medical Center, 2160 South First Avenue, Maywood, IL 60353.

DONG. HANIN AND LORENS

21x

TABLE

1

AF44A (I 5 NMOIJVENTRICLE, BILATERALLY) KEDUCES THE NUMBER OF CHOLINE ACETYLTRANSFERASE (ChAT) BUT NOT PARVALBUMIN IMMUNOREACTIVE PERIKARYA IN THE SEFTAL AREA GlWp

Parvalbumin Vehicle AF64A ChAT Vehicle AF64A

MS

VDFJ

hDB

1177 ” 20 1141 t 18

454 2 15 445 -c 16

343 L 21 390 I 29

1006t15 723 t 22* (-28%)

735 5 23 513 2 IO”

5002 11 402 i: lo* (-20%)

( -30%)

Data presented as group mean 5 SEM (vehicle, n = 7; AF64A, n = 9) number of immunoreactive (IR) perikarya. Numbers in parentheses indicate the percent reduction in ChAT-IR neurons compared to the vehicle controls. MS = medial septum. vDB = vertical limb of the diagonal band of Broca. hDB = horizontal limb of the diagona1 band of Broca. * Significantly, f > 6.98, p < 0.0001, less than in vehicle group (Student’s r-test).

mm lateral, and 3.6 mm ventral to bregma. The AF64A and vehicle solutions were infused ( 1.O gamin) via 28 gauge cannulae connected to a CMA/lOO microsyringe pump (Carnegie Medicin AB, Solna, Sweden). Following completion of the drug or vehicle infusions, the cannulae were left in situ for 5.0 min prior to being removed. Histology

On the 28th day postoperatively, each animal was anesthetized with sodium pentobarbital (90 mg/kg, IP) and perfused via the aorta with heparanized (300 U) phosphate-buffered saline (PBS; 200-250 ml at 650 ml/h) followed by 4.0% paraformaldehyde and 0.1% glu~aldehyde in 0.1 M PBS (450-500 ml at 600 ml/h). The brains were removed and postfixed for 1.0 h in the same fixative con~ning 20% sucrose. The brain was trimmed into a 6.0 mm block, beginning 2.0 mm caudal to the frontal pole, then stored in buffer as detailed previously (15). The tissue blocks were frozen onto chucks, placed in a Hacker-Bright Model No. 5030 cryostat/microtome (Hacker Instruments, Fairfield, NJ), and allowed to equilibrate at 16°C for 1.0 h. Five adjacent coronal sections (40 pm) were obtained at 0.2 mm intervals between the initial decussation of the corpus callosum and 1.0 mm caudal to the ICV injection site. The first section was stained by the cresyl violet technique, the second and third were used for ChAT and parvalbumin immunohistochemistry, and the fourth and fifth sections served as controls (primary antibodies omitted). To vizualize ChAT-IR and p~~b~in-IR cell bodies in adjacent sections through the septal complex, the tissue was processed as detailed previously (10,lS). In brief, the free-floating sections were washed extensively in PBS and then exposed sequentially to: a) the primary antibodies: rabbit antiparvalbumin [ 1:SOOdilution; generously supplied by Dr. K. Baimbridge, see (3)] and monoclonal rat anti-ChAT (concentration of 2.0 &ml; Boehringer-Mannheim Biochemicals, Indianapolis, IN) for 2-3 days; b) the secondary antibodies: goat antirabbit IgG (150; Jackson Immunoresearch, Avondale, PA) and goat antirat IgC ( 150) for 5 h; and, c) rabbit PAP complex (1: 100; StembergerMeyer, Jarretville, MD) and rat PAP complex (1: 100) overnight.

All incubations were carried out at CC. After extensive washing. the immunoperoxidase reaction was developed using 0.06% 3,3diaminobenzidine (DAB; Sigma, St. Louis, MO) and DAB containing 0.01% hydrogen peroxide, then intensified by immersion in 0.1% osmium tetroxide (Sigma) for 30 s. Tris-buffered saline (50 mM, pH 7.4) containing 1.O% normal goat serum and 0.5% Triton X-100 was used for all washes and antibody dilutions. Sections were mounted and dried onto gelatinized glass slides, defatted in alcohol and xylene, then coverslipped with Permount.

The p~valbumin-IR and ChAT-IR cells in the medial septum, and in the vertical and horizontal limbs of the diagonal band of Broca, as described by Kiss et al. (lo), were examined using a Leitz Dialux 20 microscope and counted using a Baush and Lomb microprojector (X40). Cell counts were made from five sections obtained at 0.2 mm intervals through the septal area. The cell number reported for each division, thus, is representative of the total number of parvalbumin-IR and ChAT-IR perikarya in vivo within each division. RESULTS

All of the animals survived the surgery and appeared healthy at the time of sacrifice. The rats weighed between 346-381 g. No group differences were observed. As seen in Table 1, AF64A led to significant ZO-30% reductions in the number of ChAT-IR cell bodies in each of the three septal subdivisions. There were no differences between the vehicle and AF64A-treated groups in the number of parvalbuminIR perikarya observed (Table I). Figure I shows the histological effects of AF64A on ChAT-IR perikarya in the septal complex. Figure 2 demonstrates the absence of any morphological changes in septal parvalbumin-II? neurons after AF64A treatment. DISCUSSION

ICV administration of AF64A results in a substantial reduction in the number of ChAT IR but not of p~albu~n-IR cell

FIG. 1. Photomicrographs (x20) in the same coronal plane of septal choline acetyltransferase immunoreactive (PAP/DAB method) neurons in animals that had received ICV injections of vehicle (A; 3.0 fll/ventricle) or AF64A (B; 1.5 nmol/ventricle) 28 days previously. Compared to the vehicle injection (A), AF64A (B) administration resulted in a major reduction in the number of septal cholinegic neurons.

219

SEPTAL ACh AND GABA CELLS: AF64A EFFECTS

FIG. 2. Photomicrographs (X I5 in A and D; X63 in B and E; X 160 in C and F) of parvalbumin immuno~active (PAP/DAB method) neurons in the same coronal plane through the septal complex of rats injected ICV with vehicle (A-C; 3.0 $/ventricle) or AF64A (D-F; 1.5 nmo~ven~cie) 28 days earlier. These sections were obtained immediately adjacent to those shown in Fig. 1. Although AF64A produced ventricular hypettrophy (D) and substantial reductions in septal cholinergic neurons (Fig. lB), it failed to affect the morphology or number of septal parvalbumin immunoreactive cell bodies (D-F). As a reference, the arrows indicate a blood vessel common to the three adjacent micrographs for each rat. Abbreviations: ac, anterior commissure; cc, corpus catlosum; v, lateral

bodies in the septal complex. These findings support the view that AF64A is a selective cholinotoxin that induces degeneration of septobip~ampal cholinergic neurons. In a previous study we studied the dose-dependent effects of ICV AF64A on the number of ChAT-IR perikarya in the four subdivisions of the septal complex as defined by Amaral and Kurz (2). We found that the intermediate ChAT subgroup was the most sensitive to AF64A. The lowest dose (0.5 nmohventricle) examined produced a 3 1% reduction in the number of ChATIR cell bodies in this region. Higher doses of AF64A did not result in additional reductions in intermediate subgroup cholinergic neurons. Only the highest dose (1.5 nmol/ventricle) studied led to significant reductions in the number of ChAT cell bodies in the dorsal (51%) and midline (35%) subgroups. Irn~~~tly, AF64A failed to affect the number of cholinergic perikarya in the ventral subdivision. In the present study we examined the effects of AF64A on the number of ChAT-IR and parvalbumin-IR cell bodies in the three subdivisions of the septal complex as detailed by Kiss et al. {IO). Accordingly, the dorsal and midline ChAT subgroups of Amaral and Kurz (2) would be found within the

medial septum (MS) division of Kiss et al. (10). AF64A (1.5 nmollventricle) produced a 28% loss of ChAT-IR in the MS (Table l), consistent with our previous findings (15). The dorsomedial component of the intermediate ChAT subgroup (2) overlaps with the vertical limb of the diagonal band of Broca (vDB) as described by Kiss et al. (10). AF64A produced a 30% reduction in vDB ChAT neurons, again consistent with our earlier findings (15). The horizontal limb of the diagonal band of Broca (hDB) according to Kiss et al. (10) would include the ventrolateral component of the intermediate subgroup as well as the ventral subdivision of ChAT-IR neurons (2). Because we previously failed to discern any effects of ICV AF64A on ventral subgroup ChAT-IR cell bodies, it, thus, is not surprising that the effects of AF64A on hDB cholinergic cell number (20%) was less than observed on vDB (30%) and MS (28%) ChAT-IR neurons. The question may be asked whether the AF64A-induced loss of septal ChAT-IR perikarya is due to neuronal degeneration or to a reduction in ChAT activity and antigenicity without cell loss, such that the immunohist~he~cal method employed would not be able to detect the lower levels of ChAT in otherwise healthy

220

DONG,

septal cholinergic neurons (I 2,16). To address this question, we recently have examined the effects of AF64A ( I .5 nmollventriclef on the number of septal neurons that are labeled by ChAT antibodies and by the retrogradely transported dye, Fluoro-GoldR (FG), following its injection into the caudal hippocampus 2 weeks after the administration of AF64A. We found that AF64A produced an equivalent reduction in the the number of FG labeled and of double labeled (FG + ChAT) neurons in the dorsal and

HANIN

AND LORENS

intermediate subgroups (7). Whether or not these effects of AF64A are due to ceil death has yet to be determined. Nevertheless, the present results are consistent with the view that ICV AF64A damages septal cholinergic neurons that project to the hippocampus in a selective manner. ACKNOWLEDGEMENT

This research was supported in part by USPHS Grant No. MH-42572.

REFERENCES 1. Alonso. 3. R.: Covenas. R.; Lara, J.; Aijon, J. Distribution of parvalbumin immunoreactivity in the rat septal area. Brain Res. Bull. 24:41-48; 1990. 2. Amaral, D. G.; Kurz, J. An analysis of the origins of the cholinergic and nonchofinergic septal projections to the hippocampal formation of the rat. J. Comp. Neural. 140:37-59: 1985. 3. Baimbridge, K. G.; Miller, J. J.; Parkes, C. 0. Calcium-binding protein distribution in the rat brain. Brain Res. 239519-525; 1982. 4. Brashear, H. R.; Zaborszky, L.; Heimer, L. Distribution of GABAergic and cholinergic neurons in the rat diagonal band. Neuroscience 17(2):439-45 I ; 1986. 5. Celia, M. R. Calbindin D-28k and parvalbumin in the rat central nervous system. Neuroscience 35375-475; 1990. 6. Chrobak, J. J.; Hanin, I.; Schmechel, D. E.; Walsh, T. J. AF64Ainduced working memory impairment: Behavioral. neurochemical and histological correlates. Brain Res. 463: 107- 117; 1988. 7. Dong, X. W.; Hanin, 1.; Lorens, S. A. AF64A reduces the number of septal choline acetylcholinester~e (ChAT) but not septal parvalbumin (PARV) or raphe serotonin (SHT) immunoreactive (IR) cell bodies. FASEB J. 6:A1275; 1992. 8. Freund, T. F. GABAergic septohippocampal neurons contain parvalbumin. Brain Res. 478:375-381; 1989. 9. Hanin, I. AF64A-induced central cholinergic hypofunction. In: Aquilonius, S.-M.; Gilberg, P.-G., eds. Progress in brain research. vol. 84. Amsterdam: Elsevier; 1990:289-299. IO. Kiss. J.; Patel, A. J.; Baimbridge. K. 0.; Freund, T. F. Topographical

I I.

12.

13.

14.

15.

16.

17.

localization of neurons containing pa~aIbumin and choline acetyltransferase in the medial septum-diagonal band region of the rat. Neuroscience 36:61-72; 1990. KShler, C.; Chan-Palay, V.; Wu, J.-Y. Septal neurons containing glutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain. Anat. Embryo]. 169:41-44; 1984. Lams, B. E.; Isaacson. 0.; Sofroniew, M. V. Loss of transmitterassociated enzyme staining following axotomy does not indicate death of brainstem cholinergic neurons. Brain Res. 475:401-406; 1988. Leventer, S. M.; McKeag, D.; Clancy, M.; Wulfert, E.; Hanin, 1. lntracerebroventricular administration of ethylcholine mustard aziridinium ion (AF64A) reduces release of acetylcholine from hippocampal slices. Neuropharmacology 24:453-459; 1985. Leventer, S. M.: Wulfert, E.; Hanin, I. Time course of ethylcholine aziridinium ion (AF~A)-induced cholinotoxicity in viva. Neuropharmacology 26:361-365; 1987. Lorens, S. A.; Kindel, G.; Dong, X. W.: Lee, J. M.; Hanin, 1. Septal choline acetyltransferase immunoreactive neurons: Dose-dependent effects of AF64A. Brain Res. Bull. 26:965-971; 1991. Naumann, T.; Peterson, G. M.; Frotscher, M. Fine structure of rat septohippocampal neurons: il. A time course analysis following axotomy. J. Camp. Neurol. 325:219-242; 1992. Wainer, B. H.; Levey, A. 1.; Rye, D. B.; Mesulam, M. M.; Mufson, E. J. Cholinergic and noncholinergic septohippocampal pathways. Neurosci. Lett. 54:45-52; 1985.