TWO POLYPEPTIDES FROM DENDROASPIS ANGUSTICEPS VENOM SELECTIVELY INHIBIT THE BINDING OF CENTRAL MUSCARINIC CHOLINERGIC RECEPTOR LIGANDS
DIANA JERUSALINSKY,*CARLOS CERVEÑANSKY,2 CLARA PEÑA,’ SONIA RASKOVSKY’
and FEDERICO DAJAS2
‘Instituto de Biología Celular e Instituto de Fisicay Quimica Biológicas,Facultad de Medicina,
Universidad de Buenos Aires,Paraguay 2155,1121 Buenos Aires,Argentina
Instituto de Investigaciones Biológicas Clemente Estable,Av. Italia 3318,11600 Montevideo,
Uruguay
(Received 18 October 1990;accepted 12 July 1991)
Abstract–Two new polypeptides were isolated and purified from the venom of the snake Dendroaspis angusticeps,which also contains other neuroactive peptides such as Dendrotoxins and Fasciculins.The amino acid composition of the peptides was determined and the first 10 amino acids from the MTX2 N-terminal fragment were sequenced.The so-called muscarinic toxins (MTX1 and MTX2) have been shown to inhibit the specific binding of ['HJQNB (0.15 nM), ['HJPZ (2.5 nM) and ['HJoxoM (2 nM) to bovine cerebral cortex membranes by 60,88 and 82% respectively.In contrast,they caused only a 30% blockade of the ['H]QNB specific binding to similar membrane preparations from the brainstem.The Hill number for the(‘HJPZ binding inhibition by the putative muscarinic toxin MTX2 was 0.95 suggesting homogeneity in the behaviour of the sites involved. The data from['HJoxoM binding gave a Hill number of 0.83.The decreases in the specific binding involved increases in Kp for the three different ligands (8-fold for ['HJQNB, 4-fold for ['HJPZ and 3.5-fold for ['HJoxoM) without significant changes in B.,except for a slight decrease in the ['HJoxoM binding sites (-19%); such results suggest that there may be a competitive inhibition between the MTXs and these ligands.The Ki for MTX2/['HJPZ was 22.58±3.52 nM;for MTX2/['HJoxoM,144.9±21.07 nM and for MTX2/P'HJQNB,134.98±18.35nM.
The labelling of MTX2 with '2'1 allowed direct demonstration of specific and saturable binding to bovine cerebral cortex synaptosomal membranes.
In conclusion,the results reported in this study strongly support the hypotheses that the two polypeptides isolated from D.angusticeps venom selectively inhibit specific ligand binding to central muscarinic receptors, in a competitive manner at least for the antagonist ['HJPZ and that the MTX2 specifically binds to a central site that is suggested to be a muscarinic receptor of the M,subtype.
Snake venoms are complex mixtures of extremely efficient toxins. Several peptides and proteins from this origin that interact with highly specific targets have proved to be useful for studying the nervous system [see Dolly (1989)].
The venoms from Dendroaspis species contain potent neurotoxins,some of them pharmacologically different from toxins found in other snake venoms. At least three novel effects produced by peptides isolated
*Author to whom all correspondence should be addressed. Abbreviations: AmOAc, ammonium acetate;BSA,bovine serum albumin; MAChR,muscarinic cholinergic recep-tor; MTX, muscarinic toxin;oxoM,oxotremorine-M; PMSF,para-methyl sulfonyl-fluoride;PZ,pirenzepine; QNB,quinuclidinyl benzilate;RP,reverse phase.

from this source have been reported: the anti-cholinesterase action of Fasciculins (Rodriguez-Ithur-ralde et al., 1983; Karlsson et al.,1984),the blocking of K+channels by Dendrotoxins (Harvey and Karls-son, 1980;Halliwell et al., 1986; Anderson and Harvey,1988),and recently,the partial inhibition of binding of the muscarinic antagonist ['H]qui-nuclidinylbenzilate (['HJQNB)by two peptides from Dendroaspis angusticeps venom (Adem et al., 1988).
Since the discovery of the selective antagonist piren-zepine (PZ) (Hammer et al., 1980) it has been possible to distinguish between two main pharmacologically different subclasses of muscarinic cholinergic receptor (MAChR):M1(high affinity sites for PZ) which is located predominantly in the CNS and autonomic ganglia and M2(low affinity sites for PZ) in the peri- 
238

DIANA JERUSALINSKY et al.
pheral organs like heart and also in the CNS(Ham-mer and Giachetti, 1982; Melchiorre,1988).
However,it is important to stress that more than two functional subtypes of MAChR exist,since it has been demonstrated that the M2 subtype from rat heart is heterogenous (Peralta et al., 1987).Recently,it has been demonstrated that there are at least five different gene products that behave as muscarinic receptor pro-teins,m,,m2. m,,m. and ms. Although we do not know the functional properties of all these subtypes, their distribution varies among the different CNS regions and in the periphery [Bonner et al. (1987);for review see: Bonner (1989); Hulme et al. (1990)].
Up to now there have been no reports of any selec-tive peptide ligands for MAChR subclasses. The par-tial inhibition of specific binding['HJQNB to cerebral cortex membranes caused by the above mentioned polypeptides from D. angusticeps venom may be the expression of a selective action on specific muscarinic cholinergic receptor subpopulation.The current inter-est in new selective ligands prompted us further to characterize the activity of these peptides,called MTXs.
EXPERIMENTAL PROCEDURES
Isolation and purification of the muscarinic toxins
Lyophilized D.angusticeps snake venom was obtained from Jabria,B.V..Harderwijk,The Netherlands.The general procedure followed to isolate the active muscarinic fractions was described by Karlsson and co-workers (Adem et al.. 1988).The venom dissolved in 0.10 M ammonium acetate (AmOAc),pH 6.9,was gel-filtered on Sephadex G-50(Phar-macia-LKB,Biotechnology,Sweden) in the same buffer,and the protein peak containing polypeptides eluting with molec-ular weights around 7000 Da was pooled and lyophilized.In this fraction are also present some neuroactive polypeptides such as Dendrotoxins (Harvey and Karlsson, 1980) and Fasciculins (Rodriguez-Ithurralde et al., 1983).
The next step was cation-exchange chromatography on Bio-Rex 70,400 mesh(BioRad Laboratories),equilibrated with 0.20 M AmOAc, pH 7.30, and then washed with 100 ml of 0.03 M AmOAc, pH 7.30.The sample was dissolved in the same buffer for application to the column,which was subsequently washed with ca 100 ml of 0.03 M AmOAc buffer until baseline recording was reached before an AmOAc concentration gradient at pH 7.30 was applied for isolation of several other peptides.The MTXs were not bound by the cation-exchanger under these conditions.and so the void fractions were pooled and lyophilized.
The two following cation-exchange chromatography steps were performed according to the original procedure with slight modifications (Adem et al., 1988). The gradients for the low-pressure chromatography steps were formed with an Ultrograd Gradient Mixer (LKB, Bromma, Sweden). The fraction from Bio-Rex 70 was dissolved in 0.05 M AmOAc, pH 5.20,and applied to a SP Sephadex C-25 column equi-librated with the same buffer.The elution was performed

with a linear gradient of 0.05 AmOAc,pH 5.20 to 1.00 M AmOAc,pH 6.5, as described in Fig. 1.The fractions from the protein peak showing muscarinic activity were pooled and adjusted to pH 7.4 with 1 M NH,OH,and then lyophilized.
This sample was submitted to a second step on SP Sephadex C-25 equilibrated with 0.01 M AmOAc,pH 7.00,and eluted with a linear gradient of 0.01 M AmOAc, pH 7.00, to 0.5 M AmOAc.pH 7.00,as described in Fig. 2. Two protein peaks corresponding to the putative MTXs and designated MTXI and MTX2 according to the order of elution,showed the muscarinic activity further described in this paper.
The purity of the isolated fractions was tested by reverse phase (RP) HPLC using a Bakerbond WP-Butyl(C4) column,4.6x50mm(J.T.Baker Co.,U.S.A.)using a linear gradient of 0.1 M AmOAc (Buffer A) to 0.1 M AmOAc/40% 2-propanol (Buffer B)for elution of the samples,as fully described in Fig.3.
Aminoacid determination
Samples of MTXI and MTX2 were hydrolysed with 6 N HCI containing phenol solution at 110℃ for 24 h.The hydrolysates were processed in a Beckman autoanalyser with measurement of the absorbance at 440 and 520 nm to deter-mine the aminoacid profile.
The number of residues was estimated by integrating the areas under the recorded u.v.profile to calculate the molar ratios and by rounding to nearest integer.
N-terminal sequence analysis of MTX2 was performed using an automatic gas phase sequencer(Applied Biosystems Protein Sequencer 477).
Synaptosomal membranes preparation
Crude synaptosomal membranes either from bovine cer-ebral cortex and brainstem or from rat cerebral cortex were

Fraction No.
Fig. 1. First SP Sephadex C-25 ion exchange step of the sample from Bio-Rex 70 (starting from 1.4 g of venom). Column dimensions : 45 x20 mm;fraction size:6.4 ml;flow rate:25.6 ml/h.The gel was equilibrated in 0.05 M AmOAc. pH 5.20 and the lyophilized sample was dissolved in 15 ml of the same buffer to be applied to the column,which was then eluted with a linear gradient (starting at the arrow,and indicated by a dotted line) performed with an Ultrograd gradient mixer from 0.05 M AmOAc,pH 5.20 (Buffer A) and 1.00 M AmOAc,pH 6.50 (Buffer B).Muscarinic activity eluted in the main peak, and the fractions were pooled as indicated by the horizontal bar. 
Snake venom polypeptides selective for muscarinic receptors

239

Fraction No.
Fig.2.SP Sephadex C-25 ion exchange chromatography step of the main peak from (Fig. I). Column dimensions:50x20 mm;fractions size 6.9 ml;flow rate:27.6 ml/h.The gel was equilibrated in 0.01 M AmOAc,pH 7.00,and the lyophilized sample was dissolved in 12 ml of the same buffer and applied to the column,which was then eluted with a linear gradient (starting at the arrow and indicated by a dotted line) per-formed with an Ultrograd gradient mixer from 0.01 M AmOAc,pH 7.00 (Buffer A), and 0.50 M AmOAc, pH 7.00 (Buffer B). Two protein peaks with muscarinic activities (referred as MTXI and MTX2),were eluted and pooled as indicated by horizontal bars.
prepared following the original procedure of De Robertis and co-workers (De Robertis et al., 1966),with slight modi-fications.The tissue was homogenized in 20 vol.of 0.32 M sucrose,submitted to centrifugation at 1000 g,for 10 min; the pellet was washed twice and discarded;then, the super-natants were mixed and centrifuged at 40,000 g for 30 min. The pellet was resuspended in 20 vol. of Tris-HCI 10 mM and centrifuged again at 100,000 g for 45 min.
The pellet was resuspended to reach a protein con-centration of 5 mg/ml,determined by the method of Lowry et al.(1951).The membranes were stored at-70℃ until used.
Radioligand binding assays
(a) Inhibition experiments. To investigate the effect of MTXs on the ['H]QNB specific binding,the experiments were carried out in triplicate using aliquots containing 0.1 mg protein/ml in 50 mM PO-Na+,K+buffer pH 7.4 with MTX1 or MTX2 from 0.01 to 50 μM peptide concentration. The peptide concentrations were estimated from the dilution of a stock solution (1 mg/ml)prepared by weighing the dried peptides;then,the absorbance at 278 nm wave length was recorded (1.4 for MTX2). The ['H]QNB (43.9 Ci/mmol, NEN), a classical muscarinic antagonist (Yamamura and Snyder, 1974) was added (0.15 nM final concentration) and the aliquots were incubated at 37℃ for 90 min.All the experiments were done in the presence of 2 μM PMSF (SIGMA) and 0.01% BSA. The assays with ['HJPZ (87 Ci/ mmol,NEN), a selective M,antagonist, were carried out in a similar way but aliquots containing 0.5 mg protein/ml were incubated with a final concentration of 2.5 nM['HJPZ; and with MTXs concentrations from 0.5 nM to 5μM.
To study the effect on the binding of ['HJoxotremorine-M (I'HJoxoM)(84.9 Ci/mmol,NEN), a muscarinic agonist,a

Min
Fig.3.Reverse-phase HPLC of the muscarinic toxins isolated from D.angusticeps snake venom. The chromatographic conditions were: column: BAKERBOND WP-Butyl (4.6x50 mm);eluant: 0.1 M AmOAc, pH 6.9 (Buffer A), and 0.1 M AmOAc,pH 6.9/40% 2-propanol (Buffer B); gradient: 10% B for 2 min, 10 to 75% B in 30 min,75% B for 3 min, and 10% B for 5 min (re-equilibration step); flow rate: 0.8 ml/min; detection: u.v. at 280 nm: injection volume: 50 μl.(A) MTX1, ca 15 μg; (B) MTX2,60 μg.
similar protocoI was used; aliquots containing 0.5 mg protein/ml in 20 mM Tris-HCI,were incubated at 4±1℃for 3 h with ['HJoxoM,2 nM final concentration.The reac-tion was terminated by filtering through Whatman GF/B filters that were pretreated with 0.01% polyethylenimine,as previously described (Raskovsky et al.,1988).To determine the non-specific binding, similar aliquots with 5 μM atropine were incubated simultaneously.The filters were washed three 
240

DIANA JERUSALINSKY et al.
times with 5 ml buffer,and then dried at 90C and put into a scintillation vial containing 3 ml of 2.5-diphenyloxa-zole (PPO)/xilene (5 g/l) as scintillation fluid.The specific binding was obtained by substracting the non-specific from total binding.The non-specifically bound radioactivity amounted to 10,15 and 25%,respectively,for ['HJQNB. ['HJPZ and ['HJoxoM.
These experiments were done with crude synaptosomal membranes from cerebral cortex and brainstem from bovine or rat brain, but only the results concerning the bovine brain are shown. Using the same preparation of cortical membranes, similar experiments were carried out with ['H]prazosin (NEN) and ['HJflunitrazepam (Amer-sham).α-adrenergic and benzodiazepinic receptor specific ligands respectively.
(b) Saturation curves in the presence of the toxins.Classi-cal saturation curves for the muscarinic ligands were per-formed in the absence and presence of different con-centrations of MTX2 (10-%,10-'.10-M)and from 0.025 to 2 nM of ['HJQNB,from 0.5 to 20 nM ['HJPZ,or from 0.1 to 20 nM ['HJoxoM [for methods,see Jerusalinsky et al. (1981); Raskovsky et al. (1988)]. Synaptosomal membranes from bovine cerebral cortex were incubated in duplicate with MTX2 and the appropriated radioligand,at 37C±1,in the presence of 2 μM PMSF and 0.01% BSA.The incubation was carried out for 90 min. To reach the steady state for ['H]oxoM binding the samples were incubated in an ice-water bath for 3 h.The non-specific binding was determined in duplicate under the same conditions but in the presence of 5 μM atropine.The samples were filtered and counted as in(a).
(c)Radioiodination of MTX2 and hinding assays. MTX2 was labelled with INa (NEN) following the method of Greenwood et al. (1963). The ['2'I]MTX2 was purified by RP-HPLC and three different peaks were obtained.The most homogeneous peak,with the highest inhibitory potency and retention time nearest to the peptide position was selected for the experiments after the three peaks were assayed for their inhibitory effect. Specific activities between 59 and 76 μCi/μg(from 406 to 520 Ci/mmol)were obtained.The radio-activity was recorded in a Gamma counter,with an efficiency of 60%.
To test the['2*I]MTX2 binding capacity,either crude synaptosomal membranes fraction or purified synaptosomal membranes from cerebral cortex were used;these purified membranes were obtained by submitting the crude fractions to discontinuous sucrose gradients (De Robertis et al., 1966). Aliquots of membranes containing 0.5 mg protein/ml were incubated at 4°C for 2 h to reach steady state. Saturation assays were performed using a range from 0.8 to 80 nM MTX2(estimated concentrations).
Non-specific binding was determined by adding 10 μM non-labelled MTX2,to equivalent samples. Similar aliquots were incubated with 10 μM atropine. The assays were ter-minated by filtering through Whatman GF/B filters which were previously incubated in 0.02% polyethylenimine for I h at 4C to reduce non-specific binding to them.
RESULTS
Isolation and prification of the MTXs
A combination of chromatographic methods based on different principles was employed to isolate MTXI

and MTX2 peptides and to assess their purity.Gel filtration on Sephadex G-50 and ion exchange chro-matography on the cation exchanger Bio-Rex 70(data not shown) are steps in common to the procedure for isolation of Dendrotoxin and Fasciculinfrom D. angusticeps venom (Harvey and Karlsson,1980; Rodriguez-Ithurralde et al.,1983).The unretarded peak from the Bio-Rex 70 run containing the MTXs was submitted to two consecutive chromatographic steps on the cation exchanger SP-Sephadex C25 under conditions fully described in Figs 1 and 2. Shallower gradients were employed in this study compared to those used in the original procedure (Adem et al.. 1988).Enhanced separation of MTXI from MTX2 and also higher purity of the peptides as revealed by the RP-HPLC analysis shown in Fig.3,were obtained. Under the chromatographic conditions employed using the Butyl (C4)RP column,the retention times were 27.3 min for MTX1 and 12.5 min for MTX2.
About 10 mg of MTXs were recovered per g of dry venom,70% of the total being MTX2.For this reason most of the assays were performed with MTX2, par-ticularly those for which considerable amounts of pep-tide were required.
In Table I is shown the probable amino acid com-position of MTXI and MTX2 compared with those determined by Adem et al. (1988). As can be seen, the
Table 1.Amino acid composition of the two polypeptides,MTXI and MTX2,from the 7000 Da fraction of D.angusticeps venom
Amino acid sequence of MTX2 N-terminal: LEU-THR-(CYS)-VAL-THR-THR-LYS-SER-ILE-GLY 10
A M MTX2
acid a b c 1 b c
Asp 7.39 7 7 11.09 11 10
Thr 8.7 9 8 6.76 7 7
Ser 2.62 3 3 1.94 2 2
Glu 5.5 5 5 2.25 2 2
Gly 3.32 3 3 4.9 5 5
Ala 1.72 2 2 2.47 2 3
Val 3.04 3 3 5.17 5 5
lle 5.0 5 4 3.2 3 3
Leu 2.14 2 2 1.19 1 1
Tyr 2.66 3 3 1.94 2 2
Phe 1.69 2 2 0.88 1 1
His (0.28) 0 0 1.0 1 1
Lys 4.75 5 5 5.03 5 6
Arg 3.04 3 3 1.82 2 2
Trp ND 2 ND - 1
Half-cys ND 8 ND - 8
Pro 3.11 3 4 2.96 3 4
Total residu 55 64 52 63
a.Molar ratio(n=3).
b.Amino acid residues number(preliminary results).
c.Amino acid residues number obtained by Adem et al.(1988).
ND.not determined. 
Snake venom polypeptides selective for muscarinic receptors

241
probable number of residues is rather similar in both cases.Moreover sequence analysis of the first ten amino acids from the MTX2 N-terminal portion (see legend to Table 1)showed unique amino acid per position and the sequence was exactly as one described by Ducancel et al. (1991) and Karlsson et al. (1991).
Binding experiments
(a) Inhibition experiments. MTX1 and MTX2 inhibited up to 60% of the specific binding of ['H}-QNB to cerebral cortex membranes [Fig.4(A)],while inhibition only reached 34% or less in brainstem mem-branes [Fig.4(B)]. Apparent inhibition constants(K) were calculated by applying the Cheng-Prusoff cor-rection for the concentration of radioligand which assumes a competitive interaction(Cheng and Prusoff, 1973). The apparent K, for MTX2/['H]QNB was 134.98±18.35 nM(mean±SEM) in cerebral cortex membranes.
['HJPZ specific sites were about 50% of those for ['H]QNB in cerebral cortex membranes.Maximum inhibition by MTXI was 84%,and by MTX2 88% [Fig. 5(A)]; the apparent K, for MTX2 was 22.58±3.52 nM,calculated using the correction of Cheng and Prusoff.The Hill plot has a slope of 0.95 [Fig.5(B)].
When the specific binding of ['HJoxoM to cerebral

cortex membranes was analysed (Fig.6), an 80 to 82% maximal inhibition was observed in the presence of MTX1 or MTX2. The apparent K; for inhibition of ['H]oxoM binding was 144.9±21.07 nM (after cor-rection for the ligand concentration).
Both ['H]prazosin and ['H]flunitrazepam specific binding to cerebral cortex membranes were not sig-nificantly changed by 1 or 5 μM MTXs (data not shown).
(b) Saturation curves in the presence of the toxins. The apparent dissociation constant (Kp)and the maximal number of specific sites (Bmax)for each radio-ligand were estimated from the saturation curve data by using the computer program LIGAND (Munsons, 1987). In the case of ['HJPZ, a Scatchard plot of the data was adequately fitted by a straight line suggesting that a single population of binding sites might be involved.Saturation assays with ['H]PZ in the pres-ence of 1 μM MTX2 were consistent with 4-fold increase in Kp,without significant change in the Bmax (Fig.7),although the latter was not well-defined.
As can be seen from Table 2,incubation with 1 μM MTX2, led to an 8-fold increase in the Kp for ['H]QNB with little effect on the Bmax,in saturation experiments,in cerebral cortex membranes. In ['HJoxoM saturation experiments there was a slight decrease in the maximal number of sites (Bmax) and

Log [MTX]
Fig.4.Inhibition of ['HJQNB specific binding to bovine cerebral membranes by MTXs. (A) Inhibition curves of a representative assay(from eight) carried out in cerebral cortex membranes with 0.15 nM ['H]QNB, incubated with 10-8 to 2×10-3 M concentrations of either MTXI or MTX2. The ['H]QNB bound was expressed as the percentage of the specific binding (100%=0.63 pmol/mg protein) and the MTXs concentration was expressed as the logarithm of molarity. (B) Inhibition of ['H]QNB specific binding to bovine brainstem membranes by MTX2.Data of a representative assay (from five) carried out with brainstem membranes incubated with 0.15 nM ['H]QNB in the presence of MTX2 [the incubation
conditions were the same as in(A)]. 
242

DIANA JERUSALINSKY et al.

Log[MTX21M
Fig.5.Inhibiton of ['HJPZ specific binding to bovine cerebral cortex membranes by MTX2.(A)Inhibition curve of a representative assay (from 7)carried out in cerebral cortex membranes with 2.5 nM ['H]PZ incubated with 0.5 nM to 5 μM concentrations of MTX2.The ['HJPZ bound was expressed as the percentage of the specific binding (100%=0.186 pmol/mg protein) and the MTX2 concentrations were expressed as the logarithm of molarity. (B) Hill plot for the MTX2 inhibition data from(A).P.% bound. nH=0.95.
3.5-fold increase in Kp in the presence of MTX2 (Table 2).
(c) Radioiodination of MTX2 and binding assays. MTX2 was radioiodinated and the estimated specific activities varied from 59 to 76 μCi/μg.
There was reproducible and saturable binding of ['2I]MTX2 to cerebral cortex membranes.In Fig. 8(A) a saturation curve(selected from 12)for the specific binding of ['2I]MTX2 to purified synap-tosomal membranes from bovine cerebral cortex is shown.When the binding was performed to the same crude synaptosomal membrane fraction used for the other assays,the non-specific binding was 50% of the total bound radioactivity, while it was 35% in the purified synaptosomal membranes.
Samples incubated with 10 μM atropine showed a similar pattern of “non-specific” binding.
There was a good correlation between the specific binding and the protein concentration,being linear from 0.08 to 1 mg/ml (c.c.=0.94).
Saturation of binding was achieved with 25 nM of ['25I]MTX2(estimated concentration) and the bind-ing reaction reached equilibrium in 120 min at 4°C.
The concentrations of the toxin were estimated

from the initial dry weight of the peptide, the absorb-ance of the stock dilution at 278 nm and from the absorbance and radioactivity of the peak after the RP-HPLC run:however,they might not represent the actual concentration of active peptide since degra-dation and non-specific binding to the tubes and to the membranes may occur.
Scatchard transformation of the saturation iso-therms [Fig. 8(B)] gave a good fit to a straight line with a Kp of around 17 nM and a Bmax of 0.87±0.11 pmol/mg protein in the purified synaptosomal membranes.
DISCUSSION
Considering that the purification procedure is based on principles other than hydrophobicity(i.e.mainly on size and charge differences),the narrow and sym-metric peaks in which the MTXs eluted in the RP-HPLC suggests homogeneity(Fig.3).The repro-ducibility of the amino acid determination data also points to the high purity of the peptides employed in the present study.Furthermore the sequencing MTX2 
Snake venom polypeptides selective for muscarinic receptors

243
(A)

(B)

Log[MTX]
Fig.6.Inhibition of ['HJoxoM specific binding to bovine cerebral cortex membranes by MTXs.(A) Inhibition curve of a representative experiment taken out from ten.The membranes were incubated with 2 nM['HJoxoM in the presence of either MTX1 or MTX2 from 10-*to 2x10-'M.The['HJoxoM bound was expressed as the percentage of the radioligand specific binding (100%=0.124 pmol/mg protein).(B) Hill plot for the MTX2 inhibition data from (A).P,% bound.nH=0.83.
N-terminal confirms the degree of purity of this peptide.
There is strong evidence to show that the muscarinic antagonist ['H]QNB binds in a monophasic way to a single class of non-interacting receptor sites (Yama-mura and Snyder, 1974; Birdsall et al., 1978).
The fact that MTXs inhibited up to 60% of the ['H]QNB specific binding in cerebral cortex mem-branes [Fig. 4(A)], while not affecting the binding of α-adrenergic and benzodiazepine ligands, suggest that MTXs interact specifically with MAChRs, though perhaps not equally with all subtypes.
Using pirenzepine,ratio of M,/M2 was found to be higher in the cerebral cortex than in the brainstem (Cortés and Palacios, 1986). The Bmax for ['H]PZ is about 50-60% of that for['H]QNBin cerebral cortex; these differences may account for the markedly lower inhibition of the ['H]QNB specific binding in the brainstem than in the cerebral cortex and may provide an “a priori” explanation for the results of the ['H]QNB partial inhibition by the toxins (Fig. 4). It seemed reasonable to suppose that MTXs would preferentially affect some sub-population which is dis-tinguishable from the whole population. The 88% inhibition of the ['H]PZ specific binding to cerebral cortex membranes by MTX2(Fig. 5)gave support to

this hypothesis.Furthermore,the Scatchard and Hill transformation of the data from the ['HJPZ binding suggested an homogeneous behaviour of the involved sites,reinforcing the possibility that they belong to the M,subtype.
Agonists displacement of the ['H]QNB binding, had revealed a complex kinetic of inhibition,which is compatible with the hypotheses of multiple agonist binding sites (Birdsall et al., 1978; Birdsall and Hulme,1983).This hypotheses has been corroborated by direct agonist binding in the CNS (Raskovsky et al.,1988)and more recently,with the cloning of different MAChR molecules (Kubo et al., 1986; Bonner et al., 1987; Bonner, 1989; Hulme et al., 1990).
As it was already mentioned,the muscarinic agonist ['HJoxoM preferentially binds to a receptor subtype/s with high affinity for agonists, in the experimental conditions used (Raskovsky et al., 1988).The Bmxfor this agonist was about 20% of that for ['H]QNB in cerebral cortex.Therefore,the 82% inhibition of the ['H]oxoM specific binding in the cerebral cortex mem-branes (Fig. 6) suggested that MTXs could be related to these central receptor subtype; however,the Hill number of 0.83 and the decrease in Bmax(-19%) could suggest another kind of interaction,e.g.with 
244

DIANA JERUSALINSKY et al.
an heterogeneous population of sites or an allosteric effect.
It is necessary to point out that we were unable to perform enough good assays of specific binding of ['H]oxoM to brainstem membranes in order to study the possible inhibition by MTX also in that brain area.There could be a reasonable explanation for that:this may be due to the small amount of MAChR in the brainstem and the small proportion of the involved putative muscarinic subtype.
Preliminary assays with atrial homogenates from rat heart have shown a 23% inhibition of the ['H]QNB specific binding and a 70% inhibition of the ['HJPZ specific binding (it should be considered that under the experimental conditions used['HJPZ binding sites are only one third of those for ['HJQNB). Such inhi-bition of the PZ binding seems to be non-competitive since there was a decrease in the Bm,without sig-nificant changes in the Kp.
From the calculated parameters from the saturation curves for ['HIJPZ binding (Fig. 7) and from the Hill number of 0.95(Fig. 5) it could be suggested that a single type of non-interacting sites was involved in the action of the MTX2.The Scatchard analysis of the saturation assays showed changes of the Kp (apparent dissociation constant) but there were not significant changes in the Bx(maximal concentration of recep-tor sites) determined in absence or presence of MTX2 for the antagonists (Table 2 and Fig.7).Although this pattern of results indicates that the toxins com-petitively inhibited the ['H]PZ specific binding to the central muscarinic cholinergic receptors,an allosteric interaction can not be ruled out, considering that the change in Kp for ['HJPZ binding in the presence of MTX2 is much lower than the expected according to the displacement studies.
Preliminary experiments showed that there was a specific binding of ['25I]MTX2,which is saturable at about 25 nM total concentration, having an estimated Kp of 17 nM,and a Bmax of about 0.87 pmol/mg protein in purified synaptosomal membranes from bovine cerebral cortex (Fig. 8). Further investigations are needed in order to assure the pharmacological specificity of [125I]MTX2 binding.

[3H-PZ](nM)

Bound(pmol/mg)
Fig.7.['HJPZ saturation curve (from 6) in the presence of MTX2 (1 μM). (A)Aliquots of purified synaptosomal mem-branes (0.5 mg/ml) from rat cerebral cortex were incubated at 37°℃ for I h with varying concentrations of ['HJPZ (from 0.5 to 20 nM).(B)Scatchard plot from the same data in (A). Bmux=0.965 (control) and 0.880 (MTX2,1 μM) pmol/mg protein; Kp = 2.37 (control) and 8.69 (MTX2, 1 μM) nM. For n=6,Bmx=1.087±0.09 (control) and 0.969±0.06 (MTX2,I μM)pmol/mg protein;Kp=2.25±0.72(con-trol) and 9.58±0.59 (MTX2,I μM) nM (expressed as media±SEM).
Table 2.Effects of MTX2 on the maximal number of sites(B.)and apparent dissociation constant
(Kp)of the {'HJQNB and['HJoxoM specifically bound to cerebral cortex membranes
(Kp)of the {'HJQNB and['HJoxoM specifically bound to cerebral cortex membranes
MTX2 MTX2
Ligand ['HJQNB (1μM) ['HjoxoM (1μM)
Bmux(pmol/mg) 1.740±0.120 1.560±0.170 0.320+0.025 0.260±0.018
Kp(nM) 0.129±0.052 0.998±0.136 1.895±0.540 6.525±1.082
Bmax,expressed as mean,in picomol per mg of membrane protein±SD.Kp.mean,innM±SD,n=6. 
Snake venom polypeptides selective for muscarinic receptors

245
(A)

Fig.8.(A)['2'I]MTX2 saturation curve (from 12) in bovine cerebral cortex synaptosomal membranes.Aliquots of puri-fied synaptosomal membranes (0.5 mg/ml) were incubated at 4°C for 2 h with a range of ['2'IJMTX2 estimated con-centrations from 8x10-1° M to 7.5x10-*M.The con-centration of ['2'I]MTX2 was expressed as total ligand in nM,estimated from the measured radioactivity (sp.act., 70 μCi/μg).(B)Scatchard plot from the same data in (A).
Bmax=0.92 pmol/mg protein;Kp=14 nM.
In conclusion,the results reported in this study strongly support the hypotheses that the two poly-peptides isolated from the D.angusticeps venom may selectively inhibit the binding of muscarinic ligands to central muscarinic receptors suggesting that they could be competitive inhibitors for the M, phar-macological subtype.
Acknowledgements-To Dr E. Karlsson for his helpful suggestions and advice;to J.T.Baker Chem.Co.for the kind donation of protein HPLC columns; to Dr J. Medina and Dr C.Wolfman for their suggestions and collaboration.
This work was partially supported by the International Program in the Chemical Sciences (IPICS,Uppsala Uni-versity),and the International Foundation for Science(IFS Grants No.F/994-2,F/1385/1 and F/1691/1) from Sweden, and by the University of Buenos Aires and the National Council for Science and Technology(CONICET),from Argentina.
REFERENCES
Adem A.,Asblom A.,Johansson G.,Mbugua P.M.and Karlsson E. (1988)Toxins from the venom of the green mamba Dendroaspis angusticeps that inhbit the binding

of quinuclidinyl benzilate to muscarinic acetylcholine receptors.Biochem.biophys.Acta 968,340-345.
Anderson A.J.and Harvey A. L.(1988)Effects of the po-tassium dendrotoxins on acetylcholine release and motor nerve terminal activity. Br.J. Pharmac.93,215-221.
Birdsall N.J.M.and Hulme E.C.(1983) Muscarinic receptor subtypes.Trends Pharmac.Sci.51,459 463.
Birdsall N.J.M.,Burgen A.S. V. and Hulme E.C.(1978) The binding of agonists to brain muscarinic receptors. Molec. Pharmac.14,723-736.
Bonner T.I.(1989)The molecular basis of muscarinic recep-tor diversity.Trends Neurosci.12,148-151.
Bonner T.I.,Buckley N.J.,Young A.C.and Brann M.R. (1987) Identification of a family of muscarinic acethyl-choline receptor genes.Science 237,527-532.
Cheng Y.and Prusoff W.(1973)Relationship between the inhibition constant(K)and the concentration of inhibitor which causes 50 per cent inhibition (ICso) of an enzimatic reaction.Biochem.Pharmac.22,3099-3108.
Cortés R.and Palacios J. M. (1986)Muscarinic cholinergic receptor subtypes in the rat brain. I. Quantitative auto-radiographic studies. Brain Res.362,227-238.
De Robertis E., Alberici M., Rodriguez de Lorez Arnaiz G. and Azcurra J.M.(1966) Isolation of different types of synaptic membranes from the brain cortex.Life Sci.5, 577-582.
Dolly O.(1989)Neurotoxins as Tools in Neurochemistry (Dolly O.,ed.).Ellis Horwood,England.
Ducancel F.,Rowan E., Cassar E., Harvey A., Ménez A. and Boulain J.(1991)Amino acid sequence of a muscarinic toxin deduced from the cDNA nucleotide sequence. Toxicon 29,516-520.
Greenwood F.C.,Hunter W.M. and Glover J. S. (1963) The preparation of ’2′I-labelled human growth hormone of high specific radio-activity. Biochem. J.89,114-123.
Halliwell J.V.,Othman I.B.,Pelchen-Mathews A.and Dolly J.O.(1986)Central action of dendrotoxin: selective reduction of a transient K+conductance in hippocampus and binding to localized acceptors.Proc.natn.Acad.Sci. U.S.A.83,493-497.
Hammer R.and Giachetti A. (1982) Muscarinic receptor subtypes:M,and M2biochemical and functional char-acterization. Life Sci.31,2991-2998.
Hammer R.,Berrie C.P.,Birdsall N.J.M.,Burgen A.S.V. and Hulme E.C.(1980)Pirenzepine distinguishes between different subclasses of muscarinic receptors.Nature 283, 90-92.
Harvey A.L. and Karlsson E. (1980)Dendrotoxin from the venom of the Green Mamba, Dendroaspis angusticeps. Naunyn-Schmiedeberg’s,Archs Pharmac.312,1-6.
Hulme E. C., Birdsall N. J.M.and Buckley N.J.(1990) Muscarinic receptor subtypes.A.Rev.Pharmac. Toxic. 30,633-673.
Jerusalinsky D., Aguilar J. S., Brusco A. and De Robertis E. (1981) Ontogenesis of muscarinic receptors and acetyl-cholinesterase activity in various areas of chick brain. J.Neurochem.37,1517-1522.
Karlsson E.,Mbugua P.and Rodriguez-Ithurralde D.(1984) Fasciculins,anticholinesterase toxins from the venom of the green mamba Dendroaspis angusticeps. J.Physiol. Paris 79,232-240.
Karlsson E.,Risinger C.,Jolkkonen M.,Wernstedt C.and Adem A. (1991) Amino acid sequence of a snake venom toxin that binds to the muscarinic acetylcholine receptor. Toxicon 29,521-526. 
246

DIANA JERUSALINSKY et al.
Kubo T.,Fukuda K.,Mikami A.,Maeda A.,Takashashi H.,Mishina M.,Haya T.,Haga K.,Ichigama A.,Kan-gawa K.,Kojima M.,Matsuo H., Hirose T.and Numa S. (1986)Cloning,sequencing and expression of comp-lementary DNA encoding the muscarinic acetylcholine receptor. Nature 323,411-416.
Lowry O.H.,Rosebrough N.J.,Farr A.L.and Randall R.L. (1951)Protein measurement with the Folin phenol reagent.J.biol.Chem.193,265-275.
Melchiorre C.(1988)Polymethylene tetramines:a new gen-eration of selective muscarinic antagonists. Trends Phar-mac. Sci.9,216-220.
Munsons P. J. (1987) A user’s guide to LIGAND (Coburn J.M.,ed.).Systex Inc.,5020 Sunnyside Ave,Beltsville, U.S.A.
Peralta E.G..Winslow J.W..Peterson G. L.Smith. D.H..

Ashkenazi A.,Ramachandran J.,Schimerlik M.I.and Capon D. J.(1987) Primary structure and biochemical properties of an M2muscarinic receptor.Science 236,600 605.
Raskovsky S.,Aguilar J. S., Jerusalinsky D. A. and De Robertis E. (1988) A ‘H-oxotremorine binding method reveals regulatory changes by guanine nucleotides in chol-inergic muscarinic receptors of cerebral cortex. Neurochem. Res. 13,525-530.
Rodriguez-Ithurralde D.,Silveira R.,Barbeito L.and Dajas F.(1983)Fasciculin,a powerful anticholinesterase poly-peptide from Dendroaspis angusticeps venom. Neurochem. Int.5,267-274.
Yamamura H.I. and Snyder S.H.(1974) Muscarinic chol-inergic binding in rat brain.Proc.natn.Acad.Sci.U.S.A. 71,1725-1729. MTX-211