THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.

Vol. 261, No. 12, Issue of April 25, pp. 5314-5319,1986 Prrnted m U.S.A.

Purification of Ricin AI, AZ, and B Chains and Characterization of Their Toxicity”

(Received for publication, December 4, 1985)

R. Jerrold FultonlB, David C. BlakeyB, Philip P. Knowled, Jonathan W. UhrS, Philip E. Thorpeq, and Ellen S. VitettaS From the $Department of Microbiology, Southwestern Medical School, University of T e r n Health Science Center at Dallas, Dallas, Texas 75235 and the TDrug Targeting Laboratory, Imperial Cancer Research Fund Laboratories, London, England, WC2A 3PX

This paper describes a protocol for the preparation of highly purified A (A1 and A,) and B chains of the plant toxin, ricin, and biochemical and biological char- acterization of these proteins. Intact ricin was bound to acid-treated Sepharose 4B and was split on the column into A and B chains with 2-mercaptoethanol. The A chains were eluted with borate buffer containing 2-mercaptoethanol. AI and A2 were then partially sep- arated by cation exchange chromatography and the contaminating B chain was removed by affinity chro- matography on Sepharose-asialofetuin and Sepharose- monoclonal anti-B chain. The B chain was eluted from the Sepharose 4B column by treatment with galactose and was further purified by cation and anion exchange chromatography; contaminating A chains were re- moved by affinity chromatography on Sepharose- monoclonal anti-A chain. The purified A and B chains were active as determined by their ability to inhibit protein synthesis in a cell-free assay and their binding to asialofetuin, respectively. Furthermore, by polyac- rylamide gel electrophoresis, toxicity in mice, and tox- icity on several different cell types, both A and B chains were shown to be minimally cross-contaminated. Fi- nally, it was shown that ammonium chloride signifi- cantly enhanced the nonspecific toxicity of B chains for cells in vitro. In contrast, ammonium chloride did not enhance either the nonspecific toxicity of A chains in vitro or the specific toxicity of A chain-containing immunotoxins prepared with the highly purified A,, A2 chains.

Immunotoxins are cytotoxic agents prepared by linking cell- reactive antibodies to potent toxins or their A chains (1-6). The toxin most widely used is ricin which is a disulfide- bonded heterodimeric glycoprotein consisting of a B chain (which binds to galactose-containing molecules on cell sur- faces) and an A chain (which kills cells by enzymatically inactivating the 60 S ribosomal subunit) (7,8).

The most common way to prepare an immunotoxin is to covalently link the purified A chain of ricin to an antibody. This avoids the problem of nonspecific toxicity seen with immunotoxins prepared with intact ricin. Since ricin immu- notoxins can bind to galactose residues on non-target cells,

* This work was supported by National Institutes of Health Grant CA-28149. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be sent: Dept. of Microbiology, 5323 Harry Hines Blvd., Dallas, TX 75235.

they must be used in the presence of lactose or galactose to maintain specificity. In most laboratories, a single affinity or ion exchange procedure is used to separate the B chain from the A chain, after reduction of the disulfide bond that joins the two chains o f the toxin (1-6). As demonstrated in this report, such A chain preparations are significantly contami- nated with B chains. Since the toxicity of A chain immuno- toxins can be potentiated by free B chain (9, 10) or by B chain immunotoxins (11, 12), the cytotoxic effect of an A chain immunotoxin, contaminated with B chains, could be due to the combined effects of A chain and contaminating B chain. Furthermore, the presence of B chain immunotoxin in A chain immunotoxin preparations could give rise to nonspe- cific cytotoxic effects in vitro and in vivo.

Here we describe a purification procedure for ricin A and B chains which yields fully active materials that are minimally cross-contaminated. These materials have been tested for nonspecific toxicity in mice and on a variety of cells in tissue culture and have been evaluated in cell-free assays and in radioimmunoassays. The results demonstrate that meticulous removal of cross-contaminating A or B chains from the B chain and A chain preparations significantly reduces the in vitro and in vivo cytotoxicity of both chains. Finally, these studies demonstrate that ammonium chloride markedly en- hances the toxicity of ricin B chain, and A chain preparations contaminated with B chain, but does not enhance the cytotox- icity of highly purified A chain or an immunotoxin prepared with the purified A chain.

MATERIALS AND M E T H O D S ~ ~ ~

RESULTS

Yields o f A and B Chain-Table I summarizes the isolation scheme described under “Materials and Methods,” the yields obtained at each step, and the concentrations required for 50% inhibition of protein synthesis (Em) of the material on Daudi cells in vitro. It should be noted that for the A chain, the largest variation in yield among experiments occurred during and following elution from the Sepharose 4B column. During this step and occasionally during the elution from the

Portions of this paper (including “Materials and Methods” and Footnote 4) are presented in miniprint at the end of this paper. Miniprint is easiIy read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request DOC- ument No. 85M-3949, cite the authors, and include a check or money order for $2.80 per set of photocopies. Full size photocopies are dso included in the microfilm edition of the Journal that is available from Waverly Press.

The abbreviations used in the Miniprint are: HEPES, 442- hydroxyethy1)-1-piperazineethane sulfonic acid; SPDP, succinimidyl 3-(2-pyridyldithio)propionate.

5314

Purification of Ricin A and B Chains 5315

TABLE I Yields and toxicity of A and B chains

The results are from a representative experiment. Sample 9; yield” ICmb (Daudi)

A chain Post-4B Post-CM-cellulose Post-asialofetuin-Sepharose Postmonoclonal anti-B-Sepharose

B chain Post-4B Post-DEAE (2X) Post-CM-cellulose Postmonoclonal anti-A-Sepharose

112 74 38 38

91 85 56 53

d m 1

0.15 0.61

16.7 28.3

0.51 4.1

10.2 14.4

Assuming ricin is 50% A chain and 50% B chain. IC,; concentration of A or B chain resulting in 50% inhibition of

protein synthesis as determined by [3H]leucine incorporation.

FRACTIONS

FIG. 1. Separation of Po, PI, and Pf on CM-cellulose (see “Materials and Methods”). The arrow indicates the elution profile in CM-cellulose elution buffer (see “Materials and Methods”).

45“

28 - FIG. 2. SDS-PAGE of B, AI, and Al under reducing condi-

tions. Lane 1 , ricin; Lane 2, purified B chain; Lane 3, A, (CM- cellulose peak 2); Lane 4, A, + A2; and Lane 5, A2-enriched (CM- cellulose peak 1).

CM-cellulose column, a portion of the material became insol- uble. Nevertheless, despite these variations, the final yields of A and B chains ranged from 38-46% and 46-76%, respec- tively.

Properties of Purified A Chin-Chromatography on CM- cellulose partially separated the A chain into its two forms (Fig. 1) which have been designated A, (MI = 30,000) and A2 ( M , = 32,000) (23). Fig. 1 depicts the separation of the A chain fraction from the Sepharose 4B column on the CM- cellulose column. As shown in the gel depicted in Fig. 2, P2 contained almost exclusively A, chain, whereas P1 was en-

120 I 1

MOLAR CONCENTRATION (LOG)

pm‘

FIG. 3. Top panel, ability of A chain (post-Sepharose-4B), purified A chain (A, + A*), and B chain to inhibit cell-free protein synthesis in the rabbit reticulocyte assay. A, A-post-Sepharose 4B; 0, purified A chain; 0, purified B chain. Bottom panel, binding of A chain, B chain, and ricin to microtiter plate wells coated with 2 pg/well asialofetuin. Bound A, B, or ricin was detected with 1251 affinity- purified rabbit anti-ricin antibody.

100 t

I

-io -9 -8 -7 MOLAR CONCENTRATION (LOG)

FIG. 4. Ability of post-Sepharose 4B (panel A ) and purified A chain (panel B ) to inhibit protein synthesis in Daudi cells. No NH,CI, 20 mM NH4CI, 0. A chains and NRCI were present during 16 h at 37 “C. Cells were then pulsed for 4 h with [‘Hjleucine. Nine wells of untreated cells (control) were included in each treatment group.

riched for AP, but contained approximately 30% A,; the PI and Pz components are, for simplicity, referred to as A2 and A,, respectively. A representative experiment for A chain activity (A, + A2) in the cell-free rabbit reticulocyte assay is shown in the upper panel of Fig. 3. As can be seen, both the post-4B A chain and highly purified A chain (A, + A2) have IC, values of 1 X lo-” to 1 X 10”’ M. In contrast, the IC, values of the post-Sepharose 4B A chain (Fig. 44) versus the highly purified A chain (Fig. 4B) in the Daudi cell assay, show

5316 Purification of Ricin A and B Chains

TABLE I1 Properties of ricin and A chains

The results are from a representative experiment.

Assay Target

LD50 (md" Mice (25-27 g) IC50 (cells) (pg/ml)" Daudi

AKR/A Lipopolysaccharide blasts Nonparenchymal cells

IC," (Ddml)" Cell-free

Ricin Post-4B A

7.6 X lo4 0.11 1.2 X 10-~ 0.30 1.5 X 10-3 N.D.b 0.9 X 10-3 3.5

0.03 X 10-3 N.D. N.D. 304

Postmonoclonal anti-B preparations

PAAd PI(&)

1.1 0.76 1.5 21.3 18.2 27.4 4.4 3.6 7.5

0.32 0.6 0.15

AI + A1

>15 >15 >15

152 N.D. N.D. Molar values can be calculated using molecular weights of 62,000 for ricin, 31,600 for A chain, and 31,400 for -

B chain. N.D., not determined.

TABLE I11 Properties of purified B chains

The results are averages of four experiments. Assay -NH&I +NH4Cl

LD50 (mice) (mg)" 0.56 & 0.1 N.D." 'IC, (Daudi) (pglml) 12.8 & 3.2 0.066 2 0.008 IC, (cell-free) (mg/ml) > O B N.D.

mg/25-g mice by intraperitoneal injection. N.D., not determined.

that the former is 100-fold more toxic to cells., Furthermore, the toxicity of the post-4B A chain was enhanced 10-fold by NH4C1 (Fig. 4A), whereas the toxicity of the purified A chain was enhanced only 2-fold (Fig. 4B). As summarized in Table 11, the postdB A chain was 10-fold more toxic to mice and 70-fold more toxic to Daudi cells than the purified A chain (A, + Az). Table I1 summarizes the toxicity of ricin and the purified A chains on Daudi cells, AKR-A cells, mouse lipo- polysaccharide blasts, and rat nonparenchymal cells and in- dicates that different cell lines differ in their sensitivity to the purified A chains. The nonparenchymal cells were 4 times more sensitive to A, than A,, whereas the other three cell types were slightly more sensitive to A, than to A,.

Properties of Purified B Chin-The biological activity of B chain was assessed by a radioimmunoassay which measured the ability of the purified B chain to bind to asialofetuin- coated wells of a microtiter plate. As shown in the bottom panel of Fig. 3, purified B chain binds to asialofetuin almost as well as native ricin. The half-maximal binding of 5 different batches of B chain to wells coated with 2 pg of asialofetuin averaged 1.5 pg/ml compared with 0.9 pg/ml for native ricin, representing a 3-fold difference on a molar basis. Ricin A chain does not bind to the asialofetuin-coated plates except at high concentrations.

The B chain was slightly more toxic to mice and as toxic to Daudi cells as the mixture of A, and A2 chains (Tables I and 111). Since the B chain had an IC,, in the rabbit reticu- locyte assay of greater than 1 mg/ml (1 X lo-, M) (Fig. 3, top), its toxicity to mice is probably not due to contaminating A chain (which had an IC,, of 100 pg/ml (1 X 10-l' M)), but rather to some toxic property of the B chain itself. Thus, by the criterion of enzymatic activity, the B chain preparations contained less than 1 part per ten million of contaminating A chain. In contrast to the lack of potentiation of A chain toxicity by NH4CI, the toxicity of the purified B chain for Daudi cells was increased by approximately 200-fold when it was added to the cells in the presence of 20 mM ammonium chloride (Fig. 5 and Table 111).

Characteristics of an Immunotoxin Prepared with Purified A Chain-Table IV summarizes the toxicity of rabbit anti-

-9 -6 -7 MOLAR CONCENTRATION (LOG)

FIG. 5. Ability of B chain to inhibit protein synthesis in Daudi cells in the presence (0) or absence (m) of 20 mM NH4Cl. The cells were incubated as described in the legend to Fig. 4.

TABLE IV The toxicitv of A chain immunotoxins on Daudi cells in vitro " .

IC&

-NH4C1 +NH4CI Immunotoxin Experiment

RAHIg-A1 1 2

1.5 1.4

Average 1.5 (6.8 X lo-' M)"

RAHIg-A2 1 2.2 2 2.8 Average 2.5

(1.1 X lo-* M) RAHIg-A,+ A2 1 2.2

2 1.7 Average 1.9

(8.7 x 10-~ M)

1.4 1.6 1.5

(6.8 X lo-' M) 2.1 1.8 1.9

(8.7 x 10-9 M) 1.7 2.1 1.9

(8.7 x 10-9 M) "The IC, values are in pg/ml total protein and the molarity has

been calculated assuming a molecular weight of 180,000 for the immunotoxin.

human Ig coupled to A,, A,, and a mixture of A, and A2 on Daudi cells. The toxicities to Daudi cells of all these immu- notoxin preparations were similar with IC50 values ranging between 1.6 and 2.5 pg/ml. The toxicity was not enhanced by the addition of ammonium chloride.

DISCUSSION

This paper describes a new protocol for purifying the A and B chains of the plant toxin, ricin, with minimal cross-contam- ination. A number of modifications of the original purification procedure described by Olsnes and Pihl (7) were made to obtain A and B chains with minimal nonspecific toxicity in vitro and in vivo and maximal functional activity. The A chain was purified by cation exchange chromatography, affinity

Purification of Ricin A and B Chains 5317

chromatography on Sepharose 4B (a gel matrix containing galactose residues) and asialofetuin (a galactose-containing glycoprotein that binds ricin with high affinity), and finally by filtration through a monoclonal anti-B chain antibody column. The A chain was partially separated by CM-cellulose chromatography into the AI form which has a molecular weight of 30,000 and the Az form which has a molecular weight of 32,000 (24). The higher molecular weight of the A2 chain has been attributed to the presence of an extra oligo- saccharide side chain: the AI chain has a single complex oligosaccharide whereas the Az chain has a high mannose type oligosaccharide in addition to the complex unit (24). The highly purified A chain was 10-fold less toxic to mice and 70- 180-fold less toxic to Daudi cells than A chain purified only by elution from the Sepharose 4B column.

The B chain was purified by anion and cation exchange chromatography, affinity chromatography on Sepharose, and by filtration through a monoclonal anti-A chain antibody column. The Sepharose-coupled monoclonal antibodies used to form the columns were selected from panels of anti-A chain and anti-B chain antibodies for maximal ability to remove contaminating A and B chains, respectively. The hybridomas will be made available to interested individuals on request and will be banked with the ATCC.

The toxicity of the purified A chain (i.e. a mixture of A, and Az chains) to different cell types in vitro varied greatly: rat nonparenchymal cells > AKR-A cells > Daudi cells > lipopolysaccharide blasts. The higher sensitivity of the non- parenchymal cells may be due to the presence of receptors for mannose-terminating oligosaccharides (25). Such cells may bind and endocytose the A chain more efficiently? The non- parenchymal cells were about 4-fold more sensitive to the Az chain than to the A1 chain, possibly because the two oligosac- charides on the Az chain cause it to bind to the cells with higher avidity. In the other two cell types, the toxicity of the AI- and A2-enriched preparations was similar. The toxicity of purified Az chain to mice was about half that of A1 chain; it is possible that the high mannose type oligosaccharide on the AB chain promotes the clearance of this chain from the blood- stream and thereby reduces the damage inflicted upon those unidentified tissues in the mouse which determine lethality.

The purified B chain was as toxic to Daudi cells as the purified A chain, but was slightly more toxic to mice. In either instance, the toxicity of the B chain was probably not due to contamination with A chain because the B chain preparation had an IC6, in the rabbit reticulocyte assay of greater than 1 mg/ml as opposed to 100 pg/ml for the A chain ( ie . a differ- ence of 1 X 107-fold). It is probable that the B chain is toxic because it binds to, and interferes with, the function of galactose-containing molecules on the cell surface which are needed for nutrient transport, control of ion balance, and other functions necessary for cell survival. Alternatively, the B chain may insert into membranes and thereby damage the cells (26). The toxicity of the B chain has not been previously emphasized probably because of the uncertainty of whether or not it was pure.

Immunotoxins prepared by linking rabbit anti-human im- munoglobulin to purified AI, Az, or to a mixture of the two were equally toxic to Daudi cells in vitro. It would, therefore, not seem reasonable at this time to restrict the preparation of immunotoxins for therapy to only one of the A chain forms. The final decision must await more information on the blood clearance, in vivo effectiveness, and nonspecific toxicity of immunotoxins prepared with the individual chains.

D. N. Skilleter and P. E. Thorpe, unpublished results.

The specific cytotoxic effects of A chain immunotoxins have been reported by several groups to be enhanced by ammonium chloride (27-29). This effect has been attributed to the ability of ammonium chloride to elevate endosomal and lysosomal pH and thereby inhibit proteolytic digestion of the immunotoxin. However, in the present studies, the addition of 20 mM ammonium chloride only marginally enhanced the toxicity of purified A chains to Daudi cells in vitro and failed to enhance the toxicity of the A chain immunotoxins. In contrast, the in vitro toxicity to Daudi cells of purified B chains was enhanced nearly 200-fold by ammonium chloride. It is, therefore, possible that some of the potentiation observed when ammonium chloride is used with A chain immunotoxins in other test systems is due not to an effect upon the A chain immunotoxin itself, but to an effect on contaminating B chain. Paradoxically, immunotoxins containing A chains of the same purity as that described here can be enhanced 1000- fold in the presence of ammonium chloride in other test systems (30). The variables that determine this major differ- ence in potentiation of different A chain immunotoxins have not been identified. The possible variables include the density and nature of the target antigen, the specific antibody (affin- ity, isotype), the nature of the target cell, and the route of entry of the immunotoxin.

In summary, although the procedures for A and B chain purification described in this paper appear lengthy, they can be completed within 1 week. As far as we can determine by cell free assay, radioimmunoassay, toxicity to intact cells, and toxicity in vitro, there is little or no cross-contamination of the purified chains. Obviously, comparison of nonspecific toxicity with A chains synthesized by recombinant DNA technology is necessary to quantify B chain contamination should it exist. The extensive purification described is neces- sary both to minimize the nonspecific toxicity of A chain immunotoxins for therapy in uivo and to distinguish the cytotoxic effects of A chain and B chain immunotoxins in in vitro assays.

Acknowledgments-We thank Y. Chinn, K. Frazer, R. Baylis, L. Trahan, S. Gorman, B. Smith, C . Higgins, F. LaMontagne, R. Knyba, and A. N. F. Brown for technical assistance and G. A. Cheek for secretarial assistance.

1.

2.

3. 4.

5.

6.

7. 8. 9.

10.

11.

12.

13.

REFERENCES

Vitetta, E. S., Krolick, K. A., and Uhr, J. W . (1982) Immunol. Reu. 62.159-183

MiyazakL'H., Beppu, M., Terao, T., and Osawa, T. (1980) Gann ~~

71,766-774 Raso, V. (1982) Immunol. Reu. 62,93-117 Jansen, F. K., Blythman, H. E., Carriere, D., Casellas, P., Gros,

O., Gros, P., Laurent, J. C., Paolucci, F., Pau, B., Poncelet, P., Richer, G., Vidal, H., and Voisin, G. A. (1982) Immunol. Rev.

Neville, D. J., Jr., and Youle, R. J. (1982) Immunol. Rev. 62, 75-

Thorpe, P. E., and Ross, W . C. J. (1982) Immunol. Rev. 62,119-

Olsnes, S., and Pihl, A. (1973) Biochemistry 12,3121-3126 Olsnes, S., Refsnes, K., and Pihl, A. (1974) Nature 249,627-631 Youle, R. J., and Neville, D. M., Jr. (1982) J. Biol. Chem. 257,

McIntosh, D. P., Edwards, D. C., Cumber, A. J., Parnell, G. D., Dean, C. J., Ross, W. C. J., and Forrester, J. A. (1983) FEBS

Vitetta, E. S., Cushley, W., and Uhr, J. W . (1983) Proc. Natl.

Vitetta, E. S., Fulton, R. J., and Uhr, J. W . (1984) J. Exp. Med.

62,185-216

91

158

1598-1606

Lett. 164,17-20

Acad. Sci. U. S. A. 80,6332-6335

160,341-346 Nicolson, G. L., and Blaustein, J. (1972) Biochim. Biophys. Acta

266,543-547

5318 Purification of Ricin A and B Chains 14. Kennett, R. H. (1978) Methods Enzymol. 58, 345-359 23. Ramakrishnan, .S., Eagle, M. R., and Houston, L. L. (1982) 15. Fraker, P. J., and Speck, J. C. (1978) Biochem. Biophys. Res. Biochim. Biophys. Acta 719, 341-348

16. Olsnes, S., Refsnes, K., Christensen, T. B., and Pihl, A. (1975) 24. Foxwell, B. M. J., Donovan, T. A., Thorpe, P. E., and Wilson, G.

(1985) Biochim. Bwphys. Acta 840,193-203

17. Laemmli, U. K. (1970) Nature 227,680-685 25. Stahl, P. D., Rodman, J. S., Miller, M. J., and Schlesinger, P. H.

Merril, " R.* D*7 Sedman9 s. A'7 and Eberty M' 26. Ischida, B., Cawley, D. B., Reue, K., and Wisnieski, B. J. (1983) (1978) Proc. Natl. Acad. Sci. U. S. A. 75, 1399-1403

19. Klein, E., Klein, G., Nadkami, J. S., Nadkarni, J. J., Wigzell, H., 27. Casellas, p., B. J. p., Gros, p., and Jansen, F. K. (1984) J. Biol. Chem. 258,5933-5937

20. Thorpe, P. E., Ross, W. C. J., Brown, A. N. F., Myers, C. D., J. Bwl. Chem. 259,9359-9364

Commun. 80,849-857

Biochim. Biophys. Acta 405, 1-10

(1981) Science 211,1437-1438

and Clifford, P. (1968) Cancer Res. 28, 1300-1310

Cumber, A. J., Foxwell, B. M. J., and Forrester, J. T. (1984) 28. RamakfiShnan, s., and Houston, L. L. (1984) Science 2239 58- Eur. J. Bwchem. 140,63-71 61

21. Seglen, P. O., and Reith, A. (1976) Exp. Cell. Res. 100, 280-284 29. Raso, v., and Lawrence, J. (1984) J. Exp. Med. 160,1234-1240 22. Skilleter, D. N., Paine, A. J., and Stirpe, F. (1981) Biochim. 30. Martin, P., Hansen, J. A., and Vitetta, E. S. (1985) Blood 66,

Biophys. Acta 677,495-500 908-912

supp1arenta1 Elateria1 to: 3. Projvction of Pmuxloml Anti-A and Anti-B Chain.

MRTwIAIsANDMnncm

1. "tS and Buffers.

Buffers.

a. Borate Buffer (sephamse-4B -ding Buffer) . m i s buffer mntains 0.05~ bmic acid, 0.24MNaCl and InM EDTA, and is adjusted to pH 9 with 1ON NaOH and stored a t 4°C.

b. Borate Buffer Containing 2"ercaptcethanol (seiharose Buffer). Borate buffer prepared as described b e is adjwted to 5% Ivlvl with 2-rremapto-

48 A Chain Elution

ethanol and adjusted to pH 9 with lON NaM.

C. PIE-Galactose Buffer (-se-4B B Chain Elution Buffer). This buffer contains 0.01~ N"H2m4, 0.25 M galactose and o.001~ dithiothreitol u i r r 1 , PH 8.0.

o . 1 ~ galactose, adjusted u) PH 8.0 with ION N~CH.

o . 1 ~ galaceose, adjusted t o PH 8.0 with ION km.

d. D m start inq Buffer. This buffer contains 0.0lM NaH2m4, 0.00lM DIT and

e. DFAE Elution Buffer. This buffer contains 0.24 NaH2p04, 0.OOlM DlT, and

bf PFS mntaining 0.M galactose.

4 . Se-tim of A and B Chain

he extinction mefficients (E &I used to wi lda te p ro t e in concentrations far

purification Of A and B chains was -id out at 4OC. A Sepharase4B m l m was washed with ricin, ricin A chain and B chain were 11.8, 7.65 and 14.9, respectively (16) . The entire

applied to the Sepharosp4B ml- a d allared to f la r in to the gel matrix. The m l m *BS washed with 1.5 bed wl-s of the borate NaCl buffer. The effluent was collected and the

was determined (a correctly-prepared m l m -ld bind greater than 95% of a pure

was nm into the m lm. The outlet was then closed and the m1m was incubated f m 3 hrs. F&&im of ricin]. (hie bed wl- of m l d b r a t e buffer mnta- 5% 2-mrcaptOethanal

a t 4 T . The A chain was eluted with 1.5-2 bed vollmps of bsrate buffer mntaining 5% (v/vl

metrically. It should be noted that, with the recormended bed v o l m Of sepharose-4B/mg of 2-~rcaptoethanol and the protein -tent of the eluate was determined spectreo-

ricin. the p?ak eluted w i t h bo ra t e NaC1-2~cap tae thaml my "in A chain a t a -en-

higher, it my precipitate. Therefore, an e s t i w t e s k d d be mde of the anount of m t e r i a l tration of greater than 3 mJ/ml. m this buffer, i f A chain is a t t h i s concentration or

l ikely to elute 125-508 of applied protein1 . me eluted protein can be mllected in bulk i n a M e r mta- a wl- of H 0 which will yield a final concentration of appracimattely 1.5 mgs/ml of A chain. After the 6 l U t i o n Of the A chain, the m l m was washed W i t h 1.5-2.0 bed vel- of P/E buffer. lRe B chain was then eluted f m the ml- with Dm-Sephacel

a t least 10 bed vollnoes Of tOrateNac1 buffer. Ricin, a t 10-20 rq/ml in the Jam3 l f f e r , was

starting buffer.

bOay were -led t o each ml of Sepharose-48 as abrnre (bl and the gel was stored a t 4-c. d. I+~To~oM~ anti-Ficin A Chain sepharose. 2-5 mgs of affinity-purified anti-

MiECdlaIleoUS

a. All dialysis tubing was boiled in d is t i l l ed water in the pre-e of PDTA and stored in 10% ethanol in water a t 4'C.

b. m. Glycycerol was steri l ized by autalaving or by fi l tration.

c. 1 M qalactase. 2 liters of a IM solution of D-galactose (Si-) in water was stirred with 10 grm? of activated charma1 a t ram temperature for 1 hr to rermye nmtami- nants with abso- a t 280 m. The -nsion was fi l tered t h r ~ u g h Wham 54 f i l t e r paper & then thrmqh a 0.45 miwon nal- f i l t e r to rerove the charcoal. The sterile solution was stored a t 4-c.

e. Glycine-El. This contained 0.M NaCl in 0.lM glyEineHC1 buffer, pH 2.8.

Laboratmies inilots. It had an In5o in mlB/c mice Of 500 ng/259 boay might by intrapritcmeal (i.p.1 injeiw. 21 Ficm was pified fmn African castor beans ( C d a Premier O i l s , Ltd., Hull, U.K.) 113). It had an Inso in mice of 110 ng/25g baay weight by i.p. i n j a t i o n .

f. Ricin. nu " of ricin here used: 11 Ficin was p d s e d frcm vector

a. Acid-treatedSe-45. 2rm. b. Rsialofetuin-Sepharose. 1 ml. d. Klnrlonal anti-A Chain Sechxos c. mnocl-1 anti-Ricin B Chain S%hrose. 2 ml.

e. 2 mL.

9. Ol-Cellulose for B Chain. 1 ml. f . CM-Cellulose for A Chain. 2 ml. e. DEAE-svhacel. 1 ml.

21 50% Iethal Dsse (DA in Mice. Grarps of 259 w / c mice were injected i.p. with 0.1, 0.2, 0.5, 1.0, a n d Z . 0 ng of puifid A or B chains. Mice were w e i m daily and atKerved for one week. with 2 uq/well asialofehlin (Signrl) . The plates were washed extensively in d is t i l l ed water. 3) B i d i c q of B Chain to Asialofetuin. Wells Of a 96 Well microtiter plate were mateed

D i l u t i o n s of r icin, A chain or B chain w e added and the plates yere incubated for 2 9 a t 2 5 T d then for 16 hrs a t 4-c. The ~lates were washed again in d is t i l l ed water. 10 CFM

Purification of Ricin A and B Chains 5319

chased fran Fre Biotec a d s tored a t -70% in 1.0 ml slim. a. nabbit Fetidme ~ysate. untreated rabbit retimlycyte lysate was pur

0. W. Press, E. S. Vitetta, and P. J. Martin, manuscript submitted for publication.

glyml. 90 mls of KC1 and 605.7 ngs of his were b m g h t to a to ta l vo lm of 25 mls with b. Hemin stock. 16.3 mgs of bovine heart hemin (Si-), 22.5 mls of ethylene

glass-distilled water, adjusted t o Q PH of 8.2 with concentrated Hc1 and stored at -2Ooc.

t o a final wlm of 100 ials in glass-distilled water and stored a t 4-C. e. IC" Mgc1,.2M Xcl. 0.203 g ~ " Eq2l2.65O. 14.92 gr- Of XCI, adjusted

f. . that 20

as described4