BRAIN RESEARCH ELSEVIER Brain Research 717 (1996) 135-146 Research report Polyamine amides are neuroprotective in cerebellar granule cell cultures challenged with excitatory amino acids A.C. Green a,*, K. Nakanishi b, P.N.R. Usherwood a " Department of Life Science, UniL'ersiO, of Nottingham, Universi~ Park, Nottingham NG7 2RD, UK b Department of Chemistry, Columbia University, New York, NY 10027, USA Accepted 27 December 1995 Abstract Primary cultures of rat cerebellar granule cells have been used to assess the potential neuroprotective effects of philanthotoxins and argiotoxin-636 (ArgTX-636). These polyamine amides are potent antagonists of ionotropic L-glutamate (L-Glu) receptors. In granule cells loaded with fluo-3, ArgTX-636 and philanthotoxin-343 (PhTX-343) antagonised increases of intracellular free calcium concentration ([Ca2+]i) that were stimulated by N-methyl-D-aspartate (NMDA). The antagonism was use-dependent. Antagonism by PhTX-343 was fully reversible, but recovery following antagonism by ArgTX-636 was slow and only partial during the time-course of an experiment. Neither compound inhibited K+-induced increases in [Ca 2+ ]i" In excitotoxicity studies with cerebellar granule cells, the release of lactate dehydrogenase (LDH) and morphological observations were used to assess cell death. A 20-30 min exposure to 500 /xM NMDA, 100 /xM L-Glu or 500 /zM kainate was sufficient to kill > 90% of the cells after 18-20 h. When added 5 rain prior to, and during agonist exposure, PhTX-343 and ArgTX-636 provided total neuroprotection. ArgTX-636 was about 20-30 fold more potent than PhTX-343 against NMDA, but was approximately equipotent with PhTX-343 against a kainate challenge. Neither of the toxins showed any inherent toxicity even at 400 #M and 100 /xM respectively. Some analogues of PhTX-343 are more potent, both in terms of antagonism of NMDA-stimulated increases of [Ca2+] i and neuroprotection, than PhTX-343 and ArgTX-636. Keywords: Polyamine amide toxin; Excitotoxicity; Neuroprotection; Cerebellar granule cell; Excitatory amino acid receptor 1. Introduction [4,5,13,23,27,38,39,53,62,63]. In addition to their effects on glutamate receptors, polyamine amide toxins antagonise Polyamine amide toxins are low molecular weight com- nicotinic acetylcholine receptors [7,61] and, at higher con- pounds that have been isolated from the venoms of a wide centrations, voltage-sensitive calcium channels (VSCCs) variety of spiders (e.g. argiotoxin-636; ArgTX-636), and [40,51,64,65,67]. from the Egyptian digger wasp Philanthus triangulum Sustained agonist stimulation of the ionotropic gluta- (e.g. philanthotoxin-433; PhTX-433) [10,35,69,70]. They mate receptors of excitable cells can lead to a form degen- share a similar structural design, consisting of an aromatic eration of these cells termed excitotoxicity [18,48], i.e. the chromophore, a polyamine backbone, one or more amino agonist is excitotoxic. In humans, excitotoxic cell death acid residues and a terminal amino or guanidino group [70] may contribute to the neuronal injury that follows acute (Fig. 1). These natural products contribute to the paralysing CNS insults such as ischaemia, seizures and trauma and effects of spider and wasp venoms by non-competitively may also be involved in the development of certain chronic antagonising ionotropic glutamate receptors of prey such neurodegenerative diseases [18,48]. Data from in vitro as insects [20,42]. They are also potent, non-competitive models suggest that the entry of Ca 2+ plays a pivotal role antagonists of ionotropic glutamate receptors of mam- in the initiation of a cascade of changes that leads eventu- malian central nervous systems (CNS), where both N- ally to cell death [17,19,25,28,29,60]. Prevention of Ca 2+ methyl-D-aspartate (NMDA) and non-NMDA L-glutamate entry, particularly through channels gated by NMDA re- (L-G lu) receptors are antagonised ceptors [57,68] and VSCCs [6,71] may be an efficacious neuroprotective strategy, which might be achieved using either competitive or non-competitive antagonists of these * Corresponding author. Fax: (44) (115) 951-3251. signalling proteins. However, some subtypes of non- 0006-8993/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PI1 S0006- 8993(96)00042-X 136 A.C. Green et al. / Brain Research 717 (1996) 135-146 NMDA receptor types are also permeable to Ca 2+ [12]. 2. Materials and methods For this reason polyamine amides might be ideally suited as neuroprotectants because they antagonise both NMDA and non-NMDA classes of L-Glu receptor as well as 2.1. Granule cell culture VSCCs [40,65,67,70]. In this study, polyamine amide- mediated antagonism of NMDA receptors and VSCCs in primary cultures of rat cerebellar granule cells has been Primary cultures of cerebellar granule cells were grown examined by monitoring changes of intracellular free Ca 2+ by a modification of published procedures [49]. The cere- ([ Ca2+ ]i)- In addition, these cells undergo a marked excito- bella from 7-8 day old rats (Wistar) were pooled and toxic degeneration in response to challenge with NMDA, finely chopped. The slices were collected in a modified L-Glu or kainate [41,46,47] thus allowing polyamine Kreb's medium supplemented with trypsin and DNAse amide-mediated neuroprotection to be investigated quanti- (0.25% and 0.01% w/v, respectively). After a 15 min tatively. incubation at 25°C, the trypsin solution was removed and A -Trptophan- ~ NAI'~NH3o + -Gycine (-Trp-) ~ H (-Gly) o I N~NH3 + -Lysine tt ~l~i3 (-Lys) Diiodo- o~ ~H (Iz -) i O NIt 2 N~N.~2 -Arginine . ~i 3 I4 (-Arg) O Hexadiene- ~'v +NH2,.~-..~~'~" -343- O H2 H H 2 H 2 + Ph- N~,~N~N~.I~N~ -3343- CY' H2 O C 7- ............) O Cm- B ArgTX-636 OH h O 0 ,,NH2 N + + + 1~ q X /H H2 H2 H -~ H HO~f"~s~ u "~CONIt,2 21"13 C .~.f.on C4 -Tyr-Asn-533-Arg H O ,NH2 N + + .~+ Fig. 1. Structures of the polyamine amides used in these studies. A: structure of the wasp toxin PhTX-433, and substitutions. B: structure of the spider toxin ArgTX-636. C: structure of a hybrid philanthotoxin-argiotoxin molecule C4-Tyr-Asn-533-Arg. Note that the nmnbers in PhTX-433 and analogues and in the hybrid molecule refer to the number of methylenes between the amino groups. In ArgTX-636 they represent the molecular weight of the toxin. A.C. Green et al./ Brain Research 717 (1996) 135-146 137 replaced with culture medium (Minimum Essential Medium reproducible responses could be routinely recorded from (MEM) with Earle's salts, supplemented with 10% foetal successive stimulations of the same group of cells (see calf serum, 2 mM glutamine, 5 /xg/ml insulin, 3.5 g/1 Results). Calibration of these data using ionomycin and D-glucose, KC1 to a final concentration of 25 raM, peni- Mn 2+ quenching proved unreliable, with fuorescence in- cillin 10 IU/ml and streptomycin 10 /xg/ml) supple- creases in the presence of ionomycin/Ca 2+ being unstable mented with 0.01% DNAse. The tissue was then dissoci- and not well-maintained. Consequently, these results have ated by gentle trituration through a graded series of fire- been expressed as a change in the fluorescence intensity polished Pasteur pipettes. The dissociated cells were col- relative to baseline (i.e. AF/F where AF is the change in lected by a brief centrifugation (1000 × g max 5 min) and fluo-3 fluorescence intensity and F is the baseline fluores- resuspended in fresh culture medium. The cell suspension cence measured after subtraction of the background signal was passed through a 77 ~m steel mesh to remove undis- recorded following lysis of cells by Triton-X100). Fluo-3 sociated tissue and cell clumps and diluted to ~ 1.5 × 10 6 was used in these studies because of constraints imposed cells/ml. Cells were plated in 24-well plates (0.5 ml/well) by the available single wavelength microfluorimetry equip- coated with poly-D-lysine for toxicity studies or on to 22 ment. Subsequently however we have obtained preliminary mm square coverslips (No. 1) coated with poly-D-lysine data from granule cells loaded with the ratiometric Ca R+ and placed in 35 mm Petri dishes for Ca R+ studies. The sensitive dye fura-2. Measurements of [Ca 2+ ]i obtained cultures were incubated at 37°C in humidified 95% air/5% with a duel wavelength fluorescence imaging system and CO 2. After 18-24 h, the culture medium was replaced IonVision software (Image processing and Vision Com- with fresh medium containing 10 /xM cytosine-/3-D- pany Ltd.) have been fully supportive of the data presented arabinofuranoside to prevent the replication of non-neuro- in this report and will be the subject of future publications. nal cells. The medium was not changed again during the culture period. Cells were used after 9-14 days in vitro for 2.3. Toxicity studies Ca R+ studies and at 12-14 days for toxicity studies. Acute exposures to excitotoxic agents (500/xM NMDA, 2.2. Calcium studies 100 /xM L-Glu and 500 /xM kainate) were conducted at 22-25°C in HEPES-buffered Locke's solution. For expo- Measurements of changes of [Ca2+] i were made using sures to NMDA and L-Glu, Mg z+ was omitted. The the cae+-sensitive fluorescent dye fluo-3 essentially as cultures were washed twice in this solution before the test described by Richardson et al. [59]. The dye was intro- agents were added. Antagonists, when present, were added duced into the granule cells by a 30-40 min incubation 5 min prior to an agonist and were present throughout the with 2.5 /xM fluo-3 AM in serum-free culture medium period of exposure to agonist.