Radioactive labeling of α-bungarotoxin with tritium

ANALYTICAL

68, 394-403

BIOCHEMISTRY

Radioactive

Labeling

VESNA A. ETEROVI~,’

Received a-[:‘H] bungarotoxin n-Bgt with tritium. radioactive specifically of cerebral provide tissues

label to the cortex.

a useful with low

( 1975)

of cx-Bungarotoxin RAY G. AUNE,

October

7. 1974;

(Bgt) Specific

May

13.

Tritium

L. BENNETT’”

AND EDWARD

accepted

was prepared activities of

with

1975

by catalytic reduction IO-IS Ci/mmol were

of 1251-labeled attained. The

was found in tyrosine. Tritiated cu-Bgt appears cholinergic receptor of diaphragm and to a similar This specificity and the high specific radioactivity tool for receptor

the study concentration.

of acetylcholine

receptor

to bind component attained

in brain

and

other

The widespread use of cu-bungarotoxin (a-BgtV for the study of cholinergic receptors has prompted the preparation of several radiolabeled toxin analogs. Iodinated cu-Bgt, labeled either with r2r,I or IxlI, has been prepared by the chloramine T method, by the lactoperoxidase method, with [1251]Cl, and by Pressman and Eisen’s method (l-5). Iodinated analogs offer the attractive advantage of high specific activity and relatively simple methods of preparation. They inhibit the physiological response of muscle and nerve cells to acetylcholine and bind specifically to cholinergic receptors as judged from inhibition of binding by other cholinergic agents (see references given above). However. iodinated analogs present the disadvantage of short half-life of the isotopes used: 60 days for IzSI and 7 days for L:{rI. In addition, the higher energies of the electron emissions (p) of ‘-:I than of the p emission from “H would be expected to lead to more self-decomposition of the InsI,labeled a-Bgt than of a-[:‘H]Bgt. While the expected relative radiation damage cannot be accurately calculated since it depends on the toxin concentration and other storage parameters, it can be estimated that the anticipated radiation dose for equal initial specific activities may be as ’ Present Universidad

address: National

z Reprint requests namics. Lawrence 3 Presented

at the

Departamento de Cordoba. should Berkeley VI

de Quimica Biologica, Ciudad Universitaria,

be sent Laboratory, Annual

to

Meeting

Mexico City, March 1975. ’ Abbreviations: a-Bgt, u-bungarotoxin: a-Bgt. iodinated a-Bgt: ““I,-labeled diiodinated cr-Bgt.

E. 1.. Bennett, Laboratory University of California, of the ACh cu-Bgt,

American K,

Society

acetylcholine

monoiodinated

394 Copyright 0 1975 by Academic Press, Inc. All rights of reproduction in any form reserved.

Facultad Cordoba.

de Ciencias Argentina.

of Chemical Berkeley. CA for

Biody94720.

Neurochemistry.

receptor: cu-Bgt:

Quimicas,

““I,-labeled

‘z51-labeled cu-Bgt,

PREPARATION

OF

a-[“H]

BUNGAROTOXIN

395

much as five to ten times larger for the *2Zl-labeled toxin than for the ‘jHlabeled material. This necessitates frequent preparations and prompt use of the labeled toxin. ‘““I-labeled cr-Bgt prepared by the chloramine T method was found unsuitable for studies of ACh R in brain (6). [“Hlacetyl cr-Bgt with 0.4-0.8 acetyl groups per molecule was prepared by Barnard rf trl. (7). Chang et rrl. (8) have separated several acetylated derivatives. finding that N,O-di- and N,N,O-tri[:‘H]acetyl cY-Bgt were specific reagents for the ACh R from diaphragm, while tri-N, tetra, and hexa acetylated products were increasingly less specific. The specific activities obtained were in the order of I Ci/mmol. This makes them unsuitable for biochemical work with tissues of low ACh R concentration. Menez et al. (9) have described the preparation of the cr-[“Hlneurotoxin from N~rjrr nigricdlis through catalytic reduction of the iodinated toxin. This work describes the preparation and some properties of a high specific activity a-ltsH]Bgt using the procedure of Menez et nl. The CY-[:‘H]Bgt binds with high specificity to the ACh R from brain cortex and diaphragm. MATERIALS

AND METHODS

Crude venom from Br/rzg~wus mu1ticitzctu.s was purchased from Miami Serpentarium and stored at -70°C until used. CM-cellulose (CM-S?) was from Whatman, Viokase from Grand Island Biological Company, and [1251]Cl (50 mCi/mmol) from New England Nuclear Corporation. Carrier-free tritium was supplied by Lawrence Livermore Laboratory. d-Tubocurarine chloride was from Calbiochem, carbamylcholine chloride from Sigma Chemical Company, and nicotine hydrochloride from K and K Laboratories. Palladium, 5% on powdered alumina. was from Matheson, Coleman and Bell. Prrptrr-crtim

of ‘=t-Lcrheled

a-Bgt

cu-Bgt was prepared from crude Bungctrrts tturlticinctlls venom as described previously ( 10. I I). The purity of the cu-Bgt preparations was estimated to be greater than 90% by disk electrophoresis in 10%’ polyacrylamide gels, following the procedure of Laemmli ( 12). The activities of acetylcholinesterase (EC 3.1. I .7). phospholipase A (EC 3. I. I .4), nucleotide pyrophosphatase (EC 3.6. I .9). and the phosphomonoesterase (EC 3.1.3. I) were measured by standard described methods t 13-l 6). These enzymes are usually present as contaminants of cY-Bgt after chromatography on CM-Sephadex. but their activities were found to be negligible in the final preparation. For desalting and concentrating of a-Bgt and its analogs. Amicon UM-7 membranes were used. cu-Bgt was iodinated with [ lZal ] Cl following the procedure of Menez et

396

ETEROVk,

AUNE

AND

BENNETT

al. (9). In the present experiments, [1251]Cl was used at carrier specific activities (10-20 mCi/mmol) in order to monitor the iodination, subsequent separations, and the completeness of the catalytic exchange. After reducing the excess [‘251]Cl with Na,S,@. the reaction mixture was promptly layered over a CM-cellulose column (2.5 x 3 cm) equilibrated with 0.1 M ammonium acetate. pH 5.X. Washing with initial buffer eliminated unbound iodine and the remaining salts. Diiodo, monoiodo, and native cw-Bgt (if any remained as such) were then separated by a linear gradient of 0.1-0.4 M ammonium acetate. pH 5.8 (4). Radioactivity of the effluent was measured in a liquid scintillation spectrometer. [ ‘251]Ci standards were counted simultaneously to correct for the radioactive decay of IZ51. For the determination of specific activities, protein concentration was estimated by multiplying the absorbance at 280 nm by a factor of 0.85 ( IO). Tritirrtion

of’ ‘251,-Lcrheled u-Bgt

The general procedure of Menez et (11. (9) was followed. The tritium system shown in Fig. I is fabricated from stainless steel. glass, and brass. The equipment is enclosed in a glove box, with the mechanical

FIG. I. A schematic diagram of the equipment used for labeling a-bungarotoxin with carrier-free tritium. Abbreviations used: D.P.. diffusion pump; G-l. O-I atm absolute-presI pm to I mm Hastings gauge: L. N., liquid nitrogen: M. P.. mechanical sure gauge: G2. pump: T. M.. tritium monitor.

PREPARATION

OF

a-1

2H]

RUNGAROTOXIN

397

and diffusion pumps located on the outside. Both pumps are vented through a tritium oxidizer which, in turn, brings the exhaust back into the glove box. The main box exhaust is filtered through a number of silica gel filters ( 17). The exhaust line is also monitored before and after the filters. The two absolute-pressure gauges used are both of the Wallace and Tiernan type. Model 62-075. To check the system for the absence of leaks, a Hastings DV 3M gauge was used. All the valves. except for the main pumpout valve and the one closest to the uranium tritide bed. are the Nupro “H” series bellows valves, Model B4-H. The other two are the Veeco forged brass angle valve. Model SL-38-S. and Hoke. Model 413506Y-3 16-55. respectively. The reaction flask had a small rotatory spoon to contain the catalyst during initial flushing with tritium (9): its total internal volume was IO ml. Desalted and lyophilized ‘““I,-labeled a-Bgt in 0.5 ml of 4 rnM sodium phosphate buffer, pH 7.4. was frozen in the reaction flask and connected to the tritiation system. Ten milligrams of palladium 5% on alumina were placed into the upper compartment of the reaction flask and flushed for 30 min with carrier-free :iH, from which any contaminating helium had been removed. The actual tritiation of the toxin took place at 400 mm of :‘H, for 30 min. Tritium remaining in the system beyond the reaction flask was returned to the uranium trap. When the reaction was finished. excess ::H, was flushed with an N, stream and collected into evacuated cylinders. The catalyst was separated from the solution by centrifugation and washed once with I ml of distilled water. The combined supernatant fluids were lyophilized twice. At this stage most adsorbed “H, and [“H,]O were eliminated and the solution was removed from the glove box. The reaction products were adsorbed on a C‘Mcellulose column similar to the one used after the iodination step and washed with initial buffer until all [“HZ]0 was eliminated. tu-[“H]Bgt was then separated from unreacted ‘“:I,-labeled tr-Bgt and partially reacted iZ51,-labeled a-Bgt by gradient elution. I251 was determined in a y-ray Nal well counter (Nuclear Chicago) to avoid interference from :‘H which had a specific activity several orders of magnitude higher. Onemilliliter aliquots of tv-[‘;H] Bgt (-30 pg) in 0.2 M ammonium acetate, pH 5.8, were frozen in liquid nitrogen and stored at ~20°C until used.

CY-(:$H] Bgt (0.7 pg) plus 500 pg of carrier cY-Bgt were hydrolyzed with a mixture of pancreatic hydrolytic enzymes (Viokase. 500 pg) for 40 hr at 37°C (3) in sealed tubes. The hydrolysate was diluted to 3.5 ml and filtered through a UM-2 membrane. The ultrafiltrate was lyophilized and analyzed in a Beckman automatic amino acid analyzer, Model l20C. in which the effluent of ion-exchange columns was diverted to a fraction

398

ETEROVi,

AUNE

AND

BENNETT

collector to allow measurement of radioactivity in individual amino acids. Lethalities to mice of native and labeled toxins were estimated by time to death after intraperitoneal injections ( IO). Twenty- to thirty-gram male CD-I Swiss mice were used. For brain and diaphragm experiments, male Sprague-Dawley rats weighing about 200 g were used. Binding to diaphragm was assayed following the technique of Berg et ~11.( 1). except that the tissue was burned in a Packard sample oxidizer. Model 306, prior to scintillation counting of bound tritium. Binding to homogenates of cerebra1 cortex was measured as described previously (6), except for the fact that cortices were homogenized directly in Ringer solution instead of 0.32 M sucrose. RESULTS AND DISCUSSION lodirmted

a-Bgt

Under the conditions described in Materials and Methods, the CMcellulose column is able to separate native a-Bgt from monoiodo- and diiodotoxin (Fig. 3). However, with the proportions of [1Z51]Cl to toxin used here (5 : I or 10: 1) no unlabeled a-Bgt was detected. In several preparations, diiodotoxin was 80-100% of the recovered material, and

0.06

1200

P 2 0.02

400

0 120

140 FRACTION

160

8 N . t 2

0

NUMBER

FIG. 2. Separation of cu.Bgt from monoand diiodinated derivatives. a-Bgt, I .X my, plus tu-Bgt (3.7 x IO” dpm) in 0. I M 150 pg of a mixture of ““I,-labeled a-Bgt and “‘1 ,-labeled ammonium acetate, pH 5.8, were separated by a column of CM-cellulose (see Materials and Methods). Fractions of 2.4 ml were collected. Peak I contained no radioactivity. Ninety percent of the radioactivity in peaks II and trichloroacetic acid. In other chromatographic separations activities of peaks 11 and III were determined. showing. of ““1 per molecule of toxin.

III was precipitable by 10% with bigger samples, the specific respectively, one and two atoms

PREPARATION

OF

(Y-[:‘H]

399

BUNGAROTOXIN

the rest was monoiodotoxin. The recovery of protein was 7.5%, and of radioactivity about 50%. The absorption spectrum of 12”I,-labeled a-Bgt was shifted slightly toward longer wavelengths, but otherwise was not significantly different from the spectrum of native cr-Bgt (Fig. 3A). This red shift was observed previously with proteins iodinated on tyrosine

l-’ 280

,

1

300

320

X inm)

X

(nm)

Absorption spectra of native (-) I 3. (A) and diiodinated t-----J Fluorescence spectra of native (-) and diiodinated I-----I toxin. Absorption wer-e obtained on a Gary I I8C spectrophotometer. and have been normalized FIG

optical Elmer

densities MPF 2A

tiplier. The A,,,, ot’ both

at 2X0 recording

excitation solutions

nm. Fluorescence fluorescence

hand was

was I .2h.

at 180

spectra spectrometer nm:

excitation

were obtained with with a Hamamatsu and

emission

toxin.

(BI spectra to the same

a Hitachi-PerkinR 106 photomul-

\litwidtha

wele

4 nm.

400

ETEROVIk,

AUNE

I

AND

I 0.4

0

I 0.8

C9 /g FIG.

1. Lethality

to

mice

of Lu-Bgt

and

Preparation

1.2

MOUSE

its radioactive

group were used. Mortality was 100% in all cases. (see Materials and Methods). t@) native u-Bgt: from

BENNETT

analogs.

Three

to five

mice

per

Potency is estimated from time to death a-Bgt: (A) w[“H]Bgt (m) “‘I,-labeled

1.

and is expected since the absorption maximum of tyrosine shifts from 375 to 387 nm upon diiodination (18). In addition, the pK,, of the phenolic hydroxyl is lowered to near neutral pH (I 8,19), and partial ionization could further increase the red shift (to 3 I 1 nm) as well as the extinction coefficient (62.50 cm-’ . M-I). Fluorescence of ‘““I,-labeled a-Bgt was quenched to 20% of that of native toxin, and the emission maximum is shifted from 342 to 347 nm (Fig. 3B). The quenching of tryptophan fluorescence could arise from a direct interaction of the iodine atom with the indole side chain (70) or via resonance transfer of the excitation energy from tryptophan to an iodotyrosine residue with a lower fluorescence intensity. Lethality to mice of 12”l,-labeled a-Bgt was similar to that of the native toxin (Fig. 4). Vogel et ~1. (3) have also reported that the lethality of ““I.,-labeled a-Bgt was similar to that of the native toxin. Tritiuted

a-Bgt

rZ51,-labeled Lu-Bgt was catalytically reduced with tritium, as described in Materials and Methods. The reaction mixture was then separated into CY-[:jH] Bgt and the remaining iodinated toxin analogs by a CM-cellulose column. Under the conditions used, 50-70% of protein was recovered from the catalyst: 75% of this latter fraction was recovered from the column. Only 75% of the recovered material was cu-[“H]Bgt, which contained less than the equivalent of 0.2 ‘?“I atoms per molecule in the first preparation, and less than the equivalent of 0.4 rZ51 atoms per molecule in the second.

PREPARATION

OF

cw-[“H]

401

BUNGAROTOXIN

The nature of this rZZl contamination in the purified ~y-[:jH] Bgt is not known since the unreacted ‘“ZI,-labeled cw-Bgt or ‘““I,-labeled cu-Bgt was separated by chromatography on a CM-cellulose column. The specific activity of the purified cu-[:‘H] Bgt was 0.36-0.49 tritium atoms per molecule-that is. 10.7-14.5 Ci/mmol. This specific activity is about 70% of the theoretical maximum of two atoms of tritium per molecule. Although dilution of tritium with hydrogen atoms from water has obviously occurred, the specific activity attained makes this derivative valuable for studies of ACh R in very low concentrations. In order to determine the position of tritium, a-Bgt was hydrolyzed enzymatically and the hydrolysate subjected to amino acid analysis. As shown by ultrafiltration through a UM-2 membrane, only 40% of the toxin was hydrolyzed to species of molecular weight less than 1000. after 40 hr of incubation. While complete hydrolysis of the cr-Bgt was not obtained, the proportions of the amino acids obtained were analytically generally similar to those observed after acid hydrolysis. Eighty percent of the radioactivity in the ultrafiltrate was recovered in the effluent of the ion-exchange columns: 4% in glycine and 76% in tyrosine. It was concluded that [:‘H]tyrosine is the principal labeled amino acid in ru-]:‘H]Bgt, and that tritiation occurs by displacement of iodine I

I

B

LI 1

2 pg

4

a [3H]-Bgt

FIG. 5. Binding of a-[:‘H] Bgt to with radial sections of rat diaphragm. described in Materials and Methods. binding to the area without end tubocurarine competitors the incubation

and carbamyl for I hr. u-[.IH] was continued

and Methods. (W) a-[:‘H] carbamyl choline chloride: tuhocurarine.

++-+

6

COMPETlTOR

/ ml

rat

choline: Radial Bgt was then for 20 min. Bgt binding (0) binding

[M]

diaphragm. (A) (x-[:!H] Bgt wab incubated for I which were then washed, dissected and counted (U) Binding to the area containing end plates: (. plates. (BI Inhibition of ~[,‘HlBgt binding by

hr as I) d-

sections of diaphragm were incubated with the added at a final concentration of 0.7 &ml. and Further procedure wa:, as described in Materials

to the area containing to the area with

end

end plates in the presence ol plate\ in the presence of (i-

402

ETEROVli..

AUNE

AND

BENNETT

from iodotyrosine as expected. No :lH radioactivity was detected at the elution position of monoiodotyrosine. Due to the limited amount of cr-[:‘H] Bgt available, accurate determination of its lethality in mice was not possible. However, the limited testing indicated that the lethality of the a-[“H]Bgt was at least 50% that of the native Lu-Bgt (Fig. 4). This lowered lethality may be due to partial reduction of disulfide bridges during the catalytic exchange. In vitro binding of tritiated toxin to rat diaphragm (Fig. ZA) was saturable and showed the characteristic preference for the end plate region ( 12 122). Inhibition of toxin binding by cholinergic agonists and antagonists in the diaphragm (Fig. 5B) also was as expected for cw-Bgt t I ). It should be pointed out that some loss of lethality with a concomitant retention of in rlitvo specificity for ACh R was observed previously for other toxin analogs (8.23). Binding of a-[:‘H] Bgt to brain cortical homogenate was saturable at the expected low toxin concentrations and strongly inhibited by the cholinergic agonist nicotine (Fig. 6). It was previously shown that cr-[:‘H]Bgt binds to a cortical crude mitochondrial fraction which contains the nerve endings, with all the characteristics of a nicotinic antagonist (6). Furthermore, the tritium label in LY-[~H] Bgt was found to be very stable: only 6% of radioactivity was exchanged with the solvent in 6 mo. Binding to brain and diaphragm of a LY-[:(H] Bgt preparation maintained frozen for 1 yr was very similar to that observed with freshly tri-

NlCOTlNE FIG.

(50

6. Binding of u-]:rH] mg wet weight/ml) was

Bgt to brain cortex homogenate. incubated with U-[:‘H] Bgt for

Further processes were as described in Materials of homogenate (50 mg/ml) were incubated with added in 0.5 ml of Ringer at a final concentration and further processed as described in Materials

[Ml

(A) Brain cortex homogenate I hr. in a total volume of Z ml.

and Methods. t B) One-milliliter aliquots ml of nicotine for I hr; toxin was then of 6.25 x IO-! M. incubated for 20 min and Methods.

0.5

PREPARATION

OF

LY-

[:‘H]

403

BUNGAROTOXIN

tiated a-B@. This behavior contrasts sharply with that observed for ‘““Ilabeled cu-Bgt, for which specific binding to diaphragm was drastically decreased, while the nonspecific binding to brain was increased within a few months after iodination (6). ACKNOWLEDGMENTS This tion

work

and

hy

through Fromageot

was the

supported Consejo

a postdoctoral for supplying

to Hiromi Morimoto. tance with the design also due to Dr. Larry Nrjfr rated tion.

by the

trdtlrcl

Energy

Resources

and

Development

de lnveatigaciones

fellowship additional

to one of the author\ (V..L\.E.). detail\ on tritiation of an a-toxin.

Wallace of the Vichery

i/l ~~ro(!/:

U.S.

National

Recent

Cientificas

Admini\tra-

y Tecnicas

of

Argentina

We thanh We are alho

Prof. indehted

P.

Erwin. Vashen H. l‘ashinian. and Marie S. Hehert fur assistritiation apparatus and the chemical procedures. Thanks are for helpful suggestions in the pr.eparation of the manuscript. worh

suggests

into hixtidine which i\ not completely the most rcccnt preparation ~I’cY-[

that

a limited

amount

of “‘I

may

bc incorpo-

replaced during the catalytic tritiation. In addiBgt exhibited lethality equivalent to the native

‘H]

Ck-Bgt.

REFERENCES I.

Berg. D. K.. Kelly, K. B.. Sargent. h'trr. A(~trt/. .Sc,i. USA 69. 147.

2. Eldefl-awi. 3. Vogel. Z.. 69, -1. Clark,

h4. E.. and Sytkow\hi.

31x1). D.
Fertuch, H. C‘. ( 1971) A. .I.. and Nirenberg.

Macmurchie.

tery. 5. Miledi. h. EtcroviC.

M. A. ( lY72) R.. Molinoff. V. A.. and

7. Barnard. X. (‘hang.

C.

I’. B.. Williamson.

I>. D..

Elliott.

F..

12. I?.

Laemmli. Fllman,

U. G.

Rb~c~lrc~m. 13. 15. lh. 17.

Hehert,

K. ( I Y70) I... Courtney. Pl7trrllltrc~ol.

Saito. I;.. and Kornherg. A.. Lowry. 0. H. eds.). Vol. Aune, R. G., and Carter.

R. C;..

P.. Ronseray.

M. S.. Hanley.

,%‘trrrf/.c~ (Lrj/rtlor~) K. 0.. Andre\.

I 15~721

.-l~,trc/.

I.andel.

A. M..

M. R.. and

227, V..

Bennett.

6X0. .lr..

and

Boquet. E. I

A. hl..

and

Rnf-

233. 37.

207. 147.

554. .JJh.

P.. and (IY75)

Featherstone.

(‘hangeux. 7,~.ric~~,rr

R.

M.

Edelhoch. H. (I 962) ./. Hiol. Chew. 237. 277X. Woltf. .I.. and Covelli. I. ( I Y)hh) Biochcmi.str\ 5. X67. Lehrer. S. S. ( 147 I ) Biochcntistry IO. 3754.

7 I. 22.

Lee. C. \‘.. Tseng. I.. F.. and Chiu. T. Porter. C‘. W.. Chit]. T. H.. Wieckt>w\ki. Hb~l. 241. 3.

23.

C‘uopel-.

Reich.

13.

I IYhI

7. xx.

Hanakan. D. .I. ( lY6Z) Bir~~~/~r,,~i.srj:\ 1. 52 I. and Pricer. W. E. ( 1950) J. Bicd. ~‘/rc~~~~. 182, 763. ( IY57) i/l Method5 in Enzymology ((‘olowich. S. I’.. and Kaplan. -1. pp. 37 l-.772. .4cadcmic PI-~\. New b’orh. C‘nnteluu. H. P.. and Boltin. R. IL. ( I Y7 I ) irr rritium (Moghibsi. M. W.. eds.). p. 73Y. Me\\ensrr Graphics. 1 as Vega\. NV.

and

/‘~cK.

SI i. I’S.4

T. H. ( 1971) ,Ylrrrrrc C/.O!IC/OII) S. ‘I‘. (lY7i) Hrit. ./. I’htrr,>rtrc.,~/.

IX. IY. 20.

D..

%. W.

.-1//o/. Bi~~/rc~!rr. 5X. 63. hl. W. ( lY72) P~CC ;\:tr/.

E.. Wolcott.

.I.. and Chiu. and Chuang.

Y. Menel. A.. Morgat. .I. L.. Fromageot. .r. I’. ClY7l) /.-EBS I>cjrt. 17, 333. V. A..

Hall.

Ri~~c~/~c~nri~rr)~ 11. 166.7. P.. and Potter. 1.. T. ( I Y7 I ) Ntr//,rr (LO/IC/CJ,I) 229. Bennett. E. I.. (lY73) ~!Iioc./~i/,r. Hio/‘/~y.s. /l~,rcl 362.

F. A., Wiechowski. c’.. Chen, T.

111. Eterovi?. 37.

I’.. and

E. ( I Y721 ./. Nisi.

H. ( lY671 .l.. and C‘/~C,HI.

!\‘trrrlrc Barnard. 247.

300X.

(L.~~/rt/~~/r) 215. E. A. I 1973)

I 177. Ntrr/tr~~

N.

0..

A.

A..

.b’nt,

)