Biochimica et Biophysica Acta, 405 (1975) 1-10
© Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands BBA 37143 STUDIES ON T H E S T R U C T U R E A N D PROPERTIES OF T H E LECTINS FROM ABRUS PRECATORIUS AND RICINUS COMMUNIS
SJUR OLSNES, KARIN REFSNES, TERJE B. CHRISTENSEN and ALEXANDER PIHL Norsk Hydro's Institutt for Kreftforskning, Forsvarets Mikrobiologiske Laboratorium and Biokjemisk Institutt, Universitetet i Oslo, Oslo (Norway)
(Received March 5th, 1975)
SUMMARY The amino acid composition of the isolated A- and B-chains of the toxic lectins abrin and ricin was determined and c o m p a r e d . Even though the two toxins originate from widely different plants, statistical analysis of the amino acid content indicates extensive homologies in the amino acid sequence of the 4 chains. The intact lectins contain no free SH-groups whereas the isolated A- and B-chains contain close to one free SH-group each. The results indicate that in both toxins the A- and B-chains are connected by a single S-S bond. The B-chains of abrin and ricin contain similar amounts of mannose and glucosamine. The A-chain of ricin also contains some carbohydrate, whereas the A-chain of abrin appears not to be a glycoprotein. The non-toxic abrus and ricinus agglutinins contain more carbohydrate than abrin and ricin. The isoelectric points of the different lectin preparations were measured by isoelectrofocusing. The intact lectins are much more resistant to heat, freezing and chemical treatments than the isolated A- and B-chains. The intact lectins are also very resistant to treatment with proteolytic enzymes, whereas the isolated chains are easily digested. Evidence indicating that the toxins and their chains undergo extensive conformational changes upon reduction of the S-S bond is discussed.
INTRODUCTION The seeds of Abrus precatorius L. and Ricinus communis L. contain the toxic lectins abrin and ricin, respectively, as well as related non-toxic lectins [1-8]. Abrin and ricin inhibit protein synthesis in vivo as well as in cell-free systems [9-15] by inhibiting the ribosome-dependent GTPase activity of the 60 S ribosomal subunit [16, 17]. Both toxins consist of two peptide chains of about equal size, the Aand the B-chain [5, 6]. The inhibitory effect on protein synthesis in cell-free systems is associated with the A-chains of the toxins whereas the B-chain binds to galactosecontaining receptors on the cell surface and is necessary for the toxicity in intact cells. Only intact toxin consisting of both chains is toxic to intact cells and living animals. To elucidate the relationship between the abrus and ricinus lectins we have
studied and compared their chemical properties and their resistance to various chemical and physical treatments. MATERIALS AND METHODS
Lectins Lectins were prepared as earlier described [5-7]. Protein concentrations were measured spectrophotometrically, using the following extinction coefficients (E~n m, 1 cm): abrus agglutinin, 14.6; abrin, 15.9; abrin A-chain, 7.87; abrin B-chain, 18.2; ricinus agglutinin, 11.7; ricin, 11.8; ricin A-chain, 7.65; ricin B-chain, 14.9, as earlier described [7].
Determination of amino acid composition The isolated A- and B-chains of abrin and ricin were precipitated with 10~ (w/v) trichloroacetic acid and the precipitate was washed once with 10 ml.of 5 ~ trichloroacetic acid, and once with 10 ml of ether. The amino acid analysis was carried out in a BioCal BC 200 amino acid analyzer, using the ninhydrin system. Taurine and norleucine were used as internal standards in the analyses [18]. Acid hydrolysis of protein samples was carried out in 6 M HC1 for 24 h. Half-cystine was measured as cysteic acid after performic acid oxidation of the sample as described by Hirs [19]. Statistical analysis of the data (SA Q-test) was carried out as described [20-23].
Determination of sulfhydryl groups Protein samples (200/~1 containing 50-500/zg protein in 50 mM Tris-HC1 (pH 8.0), 50 mM NaC1) were mixed with 10/~1 of a 2.5 mM solution of N-ethyl[~4C]maleimide (Amersham) with spec. act. 2.1 Ci/mol, and incubated at room temperature for 30 min. Then, 25/zl of 2-mercaptoethanol was added, the protein was precipitated with 10~ (w/v) trichloroacetic acid, the precipitate was collected and washed on Gelman glass fiber filters, type A, and the radioactivity was measured by scintillation counting as earlier described [11 ].
Carbohydrate determination Isolated A- and B-chains were dialyzed against 10 mM Tris-HC1 (pH 7.7) overnight and then precipitated with l0 ~ (w/v) trichloroacetic acid to denature the proteins and remove non-covalently bound carbohydrates. The precipitate was collected by low-speed centrifugation, washed twice with ether and dried in a desiccator. The carbohydrate analysis was carried out by gas chromatography of the trimethylsilyl derivatives according to Bolton et al. [24! with some minor modifications. The procedure involves release of monosaccharides as the methyl glycosides by treatment with 1 M methanol/HC1 for 3 h at 95 °C. Reacetylation of amino sugars was then carried out by neutralizing with silver carbonate, addition of 2 0 ~ (w/v) acetic anhydride and heating at 95 °C for 2 h. The insoluble material was removed by centrifugation and the supernatant evaporated. Then 50/~l silyl reagent (pyridine/ trimethylsilylchloride/hexamethyldisilazane,5:1:1) was added and 2-/~1 samples were analyzed by gas chromatography at 180 and 220 °C using Helium as a carrier gas in an F. & M. Biomedical Gas Chromatograph, model 400, with a flame ionization detector.
Isoelectrofocusing Measurements of the isoelectric points of the lectins and of isolated chains were carried out in an LKB-7900 Uniphor apparatus (Bromma, Sweden) in the following way: A dense electrode solution containing 0.7 ml concentrated sulphuric acid in 60 ~ sucrose (w/v) was added to the anode at the bottom of the column. The sample (5-8 mg of protein) was mixed into a 230 ml sucrose gradient (0--40 ~ w/v) containing 1 ~ (w/v) of carrier ampholytes (Ampholine 4 0 ~ , pH 3-10, LKB-produkter AB, Bromma, Sweden) and 1 ~ Brij-58. After the column was filled, the light electrode solution (1 ~o w/v NaOH) was added to the cathode at the top of the column, and a potential of 400 V was applied to the column. Focusing was carried out at 2 °C for 75 h and then the column was emptied by the lower exit at a flow rate of about 1-2 ml/min. Fractions of 1 ml were collected and the pH and absorbance at 280 nm of each sample were measured.
Measurement of biological activities Toxicity of abrin and ricin expressed as LDso was measured after intraperitoneal injections into groups of mice [5, 6]. Activity of A-chains was measured in a cell-free system from rabbit reticulocytes as earlier described [11]. Various amounts of toxins or their A-chains were added and the amount necessary to give 90~o inhibition of protein synthesis was measured. The intact toxins were first treated with 1 ~ 2-mercaptoethanol at room temperature for one hour. In the case of toxins pretreated with proteolytic enzymes, the treatment with 2-mercaptoethanol was carried out in a somewhat different way. 100 #g of treated toxins were diluted in 1 ml of 50 mM Tris-HC1 (pH 7.7) containing 1 mg/ml bovine serum albumin. The solution was kept on ice and 2-mercaptoethanol was added to a final concentration of 50 mM, the samples were incubated at 0 °C for 1 h and their ability to inhibit protein synthesis was measured. In every case the controls were treated with 2-mercaptoethanol under exactly the same conditions. Activity of agglutinins was measured as ability to induce direct hemagglutination and the activity of B-chains as ability to induce indirect hemagglutination [7]. RESULTS AND DISCUSSION
Amino acid composition The amino acid composition of the isolated A- and B-chains of the toxins is shown in Table I. Our results for the ricin chains are in general agreement with those reported by previous authors [25, 26] whereas the amino acid composition of the abrin chains has not been determined earlier. Our results (Table I) indicate some similarity in amino acid composition between all four peptide chains. Statistical examination of the data using the SAQ test [20-23] indicated that abrin A-chain is closely related to ricin A-chain (SAQ ~ 24), and to abrin B-chain (SAQ ---- 17). Ricin A-chain appears to be closely relatedto abrin B-chain (SA Q = 29). Ricin B-chain is less closely related to the three other chains (SAQ ---- 56-85). The number of sulfhydryl groups in the different proteins was measured by their ability to bind N-ethyl[14C]maleimide. The results in Table II show that the intact toxins and agglutinins contain no reactive sulfhydryl groups, either in the absence or presence of urea in accordance with the data obtained by other authors
TABLE I AMINO ACID COMPOSITION OF THE CONSTITUENT PEPTIDE CHAINS OF ABRIN AND RICIN The values are given as residues per chain and are not corrected for loss during hydrolysis. The values in parentheses denote the range between data obtained with vacuum dialysis and performic acid oxidation. Half-cystine was calculated as cysteic acid after performic acid oxidation. N.D., not determined. Amino acid
Abrin
Ricin
A-chain Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-cystine Tryptophan Valine Methionine Isoleucine Leucine Tyrosine Phenyialanine Histidine Lysine Arginine
31.0 ( ± 19.8 (423.4 (433.3 (414.6 (416.4 (417.9 (42.0 N.D. 15.0 (42.6 (413.8 (422.4 (410.5 (412.1 (45.8 (44.6 (418.4 (4-
0.3) 0.1) 0.4) 0.1) 0.1) 0.5) 1.0) 0.6) 0.4) 0.8) 3.1) 0.5) 2.0) 0.8) 0.4) 1.4)
B-chain
A-chain
B-chain
27.9 (-4- 0.1) 21.8 (4- 0.6) 27.8 (4- 0.1) 31.4 ( i 0.1) 15.7 (4- 0.8) 18.3 (4- 0.4) 22.6 (4- 0.2) 2.2 N.D. 18.5 (4- 0.3) 4.1 (4- 0.2) 12.6 (4- 0.3) 27.2 (4- 0.1) 16.0 (4- 0.4) 11.8 (4- 0.1) 3.9 (4- 0.1) 7.8 (4- 0.0) 16.6 (4- 0.2)
27.3 ( t 0.3) 18.1 (-¢- 0.2) 20.0 (-4- 0.9) 31.2 (4- 0.6) 16.9 (4- 0.7) 18.5 (4- 0.6) 25.4 (4- 0.2) 1.2 N.D. 13.7 (4- 0.4) 2.2 (4- 0.5) 19.4 (4- 0.1) 23.1 (4- 0.2) 14.9 (4- 0.9) 14.3 (4- 0.1) 3.6 (4- 0.6) 1.9 (4- 0.7) 20.4 (4- 0.9)
43.6 (423.4 (421.6 ( + 23.9 (416.0 (423.0 (417.2 (46.0 N.D. 16.9 (42.9 (416.3 (426.9 (412.5 (44.8 (42.3 (47.9 (414.8 (4-
0.2) 0.2) 0.3) 0.3) 0.2) 0.3) 0.1) 0.6) 0.3) 0.1) 0.3) 0.5) 0.3) 0.1) 0.3) 0.5)
[25, 27, 29]. On the other hand, the free A- and B-chains were found to contain close to one reactive SH-group both in the presence and absence of urea (Table II). This supports the suggestion [29] that the two chains are connected by a single disulfide bond. Since no free SH-group could be detected in the intact toxins, the number of half-cystine residues in excess of one, found in each chain must be present as intrachain disulfides. In view of this it is possible that the half-cystine determinations in Table I are too low and that the correct values should be 3 residues of half-cystine in each abrin chain, 1 residue in ricin A-chain and 7 in ricin B-chain. TABLE II SULFHYDRYL GROUPS IN ABRIN, RICIN AND THEIR CONSTITUENT PEPTIDE CHAINS The buffer used was 0.05 M sodium phosphate (pH 8.0) containing 0.1 M NaC1. Incubation medium
No. of sulfhydryl groups per molecule Abrin
Buffer 10 M urea in buffer
Intact
A-chain
B-chain
0.02 0.03
0.73 0.70
0.62 0.64
Abrus agglutinin
Ricin Intact
A-chain B-chain
0.3 0.2
0.02 0.02
0.92 0.93
0.82 0.82
Ricinus agglutinin 0.3 0.3
Carbohydrate content Abrin and ricin are known to be glycoproteins but the published results show considerable differences as to the amount and type of carbohydrates present [3, 30, 31 ]. Our own results indicate (Table III) that both toxins contain mannose and glucosamine (probably N-acetylglucosamine) and that most of the carbohydrates are present in the B-chains. The amount of mannose and glucosamine (probably Nacetylglucosamine) present in the B-chains was almost the same in the two toxins. TABLE III CARBOHYDRATE CONTENT IN ABRUS AND RICINUS LECTINS Protein
Number of sugar residues per molecule of protein or polypeptide Mannose .Glucose Glucosamine
Abrin 10.0 Abrin A-chain 0 Abrin B-chain 9.7 Ricin 15.6 Ricin A-chain 4.3 Ricin B-chain 10.7 Abrus agglutinin 55.8 Ricinus agglutinin 25.4
1.6 0 0 2.3 0 0 2.4 4.4
1.6 0 1.9 2.4 0.4 1.33 6.8 4.9
Also various amounts of galactose were found in the isolated abrin B-chain. Since we did not find galactose in the intact abrin, we believe that the galactose in abrin B-chain represents noncovalently bound galactose trapped during the preparation of this rather unstable peptide chain. The small amounts of glucose found in intact abrin and ricin, but not in either of their isolated peptide chains, may similarly represent artifacts. Our data indicate that abrin A-chain is not a glycoprotein, in contrast to the findings in a previous report [32]. The two agglutinins also contain mannose and glucosamine, although in different amounts compared with the toxins. It is interesting that mannose and glucosamine have been found in a variety of other lectins as well [33].
lsoelectrofocusing of the leetins The data in Fig. 1 show that the p I of abrin (6.1) was only moderately lower than that of ricin (7.1). There was, however, a striking difference in the charge o f the constituent peptide chains of the toxins. Thus, the A-chain of abrin and the B-chain of ricin were rather acidic, whereas the B-chain of abrin and the A-chain of ricin were almost neutral. It is thus clear that both toxins consist of an acidic and a neutral polypeptide chain. The results in Fig. 2 show that abrus agglutinin is more acidic than abrin and that ricinus agglutinin is more basic than ricin. Funatsu et al. [34] found p I values of 7.34, 7.42 and 5.17 for intact ricin, ricin A-chain (their Ile-chain) and ricin B-chain (their Ala-chain), respectively, in general agreement with our own data. Isoelectrofocusing of abrus agglutinin carried out by Wei et al. [28] indicated a p I of 4.7 which is somewhat lower than the value (pl = 5.2) obtained by us.
i
i
i
[~
0.8
i .~.~.x'~
•
p I7 A
~.~~
/4
0.~
C
"6 0.8
y-~........
pIT,.s ......
D
.x- ~
-
pl &6] ,,'*'*/~ X I
E
I o.8
P~~ 5 ' " " :
F
I
~"
x.XF
x~. x"
~........
.E
~IL,.8 ~ .~
x,x" ×,
0.~
N'i I
I
~0
80
~0
80
Fraction number Fig. 1. Isoclcctrofocusing of abrin, ricin and their constituent pcptide chains. A) abrin; B) ricin; C) abrin A-chain; D) ricin A-chain; E) abrin B-chain; F) ricin B-chain. (O) Absorbancc at 280 nm; ( × ) pH of fractions.
2.4
12
A
E e- 2.0
/
e~ 1.6 ~ 1.2
x.x" pi-s.2 ~
x.X"
x.X.x/ x"x.X-
pi-7.8 ~ ,x.X.X"x'x'x x,ex"
x x "x
10 8
6Z
x.xX "x'x/x'x~x" Ii ) x.X-x'x ~~!
4 x
x.x"
! T
J
X
2
O •
20
40
60
~J II , , 80
100 Fraction
20
40
60
80
100
number
Fig. 2. Isoelectrofocusing o f abrus agglutinin (A) and ricinus agglutinin (B). (O) Absorbance at 280 n m ; ( × ) p H o f fractions.
Resistance of lectins to various physical and chemical treatments The intact lectins could be stored in the frozen state without loss of activity, and also in the absence of galaetose. Repeated freezing and thawing of the solution had little influence on their toxicity. In the presence of 0.1 M galactose, the toxins could be stored in the refrigerator for several months without loss of activity. The toxins were inactivated by boiling and abrin is largely inactivated by incubation at 60°Cfor lh. The isolated chains are much less stable than the intact toxins. Ricin B-chain could be stored at --20 °C in the absence of glycerol and abrin A- and B-chain could be sorted if 10% glycerol was present during the freezing. So far we have not been able to store ricin A-chain in the frozen state under any conditions. Similarly, the isolated chains are less resistant than the intact toxins to moderate heating. The intact lectins are stable over a wide pH range. However, the isolated chains of the toxins are much less stable, at least at acid pH (Table IV). Thus, treatment with 0.1 M acetic acid removed most of the biological activity of the isolated chains. Both intact toxins and isolated A- or B-chains could be incubated overnight in Tris buffer at pH 10.0 or in acetate buffer at pH 4.0 without loss of activity. Treatment with EDTA did not change the properties of the lectins. Treatment with sodium dodecyl sulphate, which inactivates most proteins, has no effect on the biological activity of the A-chains of either toxins (Table IV). The B-chains on the other hand lost at least 90 % of their ability to bind erythrocytes after this treatment. Due to the fact that trace amounts of sodium dodecyl sulphate induce hemolysis, the exact residual activity of the B-chains could not be measured in this system. The B-chains (especially that of ricin) are also more sensitive than the A-chains to treatment with chloramine-T. Since it is known that chloramine-T oxidizes methionine [35] this suggests that a methionine residue is located in close proximity to the binding sites for galactose. Interestingly, chloramine-T appears to form polymers of the B-chains which may result in precipitation of the Protein. Also other oxidizing agents may inactivate ricin [36]. The data in Table V demonstrate that intact abrin and ricin are extremely resistant to treatment with proteolytic enzymes, as are abrus and ricinus agglutinins (not demonstrated). In contrast, the isolated peptide chains of abrin and ricin are sensitive to proteolytic enzymes, the A-chains being somewhat more sensitive than the B-chains. As expected DNAase, RNAase and neuraminidase did not have any significant effect on the different lectin preparations.
Conformational changes of the toxins T h e interchain disulfide bond in abrin and ricin is easily reduced with 2mercaptoethanol even in the native proteins, whereas the intrachain disulfide bonds are more difficult to reduce. Only after denaturation of the toxins with high concentrations of urea could these S-S bonds be reduced (data not shown). Our data indicate that abrin A- and B-chain have each one internal disulfide bond, whereas ricin B-chain has three. We have previously produced immunological evidence that the ricin chains undergo conformational changes upon combining to form intact ricin [37]. Also, the isolated B-chain of abrin is immunologically different from the B-chain present in
100 100 -100 100 100 --
--
--
--
(c)
50 100 1 1 10 0
Abrin (b)
--
--
50 100 --30 10
(d)
20
60
1 5 0 100 80 80
Abrin A-chain (c)
.
3
10 10 0
© Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands BBA 37143 STUDIES ON T H E S T R U C T U R E A N D PROPERTIES OF T H E LECTINS FROM ABRUS PRECATORIUS AND RICINUS COMMUNIS
SJUR OLSNES, KARIN REFSNES, TERJE B. CHRISTENSEN and ALEXANDER PIHL Norsk Hydro's Institutt for Kreftforskning, Forsvarets Mikrobiologiske Laboratorium and Biokjemisk Institutt, Universitetet i Oslo, Oslo (Norway)
(Received March 5th, 1975)
SUMMARY The amino acid composition of the isolated A- and B-chains of the toxic lectins abrin and ricin was determined and c o m p a r e d . Even though the two toxins originate from widely different plants, statistical analysis of the amino acid content indicates extensive homologies in the amino acid sequence of the 4 chains. The intact lectins contain no free SH-groups whereas the isolated A- and B-chains contain close to one free SH-group each. The results indicate that in both toxins the A- and B-chains are connected by a single S-S bond. The B-chains of abrin and ricin contain similar amounts of mannose and glucosamine. The A-chain of ricin also contains some carbohydrate, whereas the A-chain of abrin appears not to be a glycoprotein. The non-toxic abrus and ricinus agglutinins contain more carbohydrate than abrin and ricin. The isoelectric points of the different lectin preparations were measured by isoelectrofocusing. The intact lectins are much more resistant to heat, freezing and chemical treatments than the isolated A- and B-chains. The intact lectins are also very resistant to treatment with proteolytic enzymes, whereas the isolated chains are easily digested. Evidence indicating that the toxins and their chains undergo extensive conformational changes upon reduction of the S-S bond is discussed.
INTRODUCTION The seeds of Abrus precatorius L. and Ricinus communis L. contain the toxic lectins abrin and ricin, respectively, as well as related non-toxic lectins [1-8]. Abrin and ricin inhibit protein synthesis in vivo as well as in cell-free systems [9-15] by inhibiting the ribosome-dependent GTPase activity of the 60 S ribosomal subunit [16, 17]. Both toxins consist of two peptide chains of about equal size, the Aand the B-chain [5, 6]. The inhibitory effect on protein synthesis in cell-free systems is associated with the A-chains of the toxins whereas the B-chain binds to galactosecontaining receptors on the cell surface and is necessary for the toxicity in intact cells. Only intact toxin consisting of both chains is toxic to intact cells and living animals. To elucidate the relationship between the abrus and ricinus lectins we have
studied and compared their chemical properties and their resistance to various chemical and physical treatments. MATERIALS AND METHODS
Lectins Lectins were prepared as earlier described [5-7]. Protein concentrations were measured spectrophotometrically, using the following extinction coefficients (E~n m, 1 cm): abrus agglutinin, 14.6; abrin, 15.9; abrin A-chain, 7.87; abrin B-chain, 18.2; ricinus agglutinin, 11.7; ricin, 11.8; ricin A-chain, 7.65; ricin B-chain, 14.9, as earlier described [7].
Determination of amino acid composition The isolated A- and B-chains of abrin and ricin were precipitated with 10~ (w/v) trichloroacetic acid and the precipitate was washed once with 10 ml.of 5 ~ trichloroacetic acid, and once with 10 ml of ether. The amino acid analysis was carried out in a BioCal BC 200 amino acid analyzer, using the ninhydrin system. Taurine and norleucine were used as internal standards in the analyses [18]. Acid hydrolysis of protein samples was carried out in 6 M HC1 for 24 h. Half-cystine was measured as cysteic acid after performic acid oxidation of the sample as described by Hirs [19]. Statistical analysis of the data (SA Q-test) was carried out as described [20-23].
Determination of sulfhydryl groups Protein samples (200/~1 containing 50-500/zg protein in 50 mM Tris-HC1 (pH 8.0), 50 mM NaC1) were mixed with 10/~1 of a 2.5 mM solution of N-ethyl[~4C]maleimide (Amersham) with spec. act. 2.1 Ci/mol, and incubated at room temperature for 30 min. Then, 25/zl of 2-mercaptoethanol was added, the protein was precipitated with 10~ (w/v) trichloroacetic acid, the precipitate was collected and washed on Gelman glass fiber filters, type A, and the radioactivity was measured by scintillation counting as earlier described [11 ].
Carbohydrate determination Isolated A- and B-chains were dialyzed against 10 mM Tris-HC1 (pH 7.7) overnight and then precipitated with l0 ~ (w/v) trichloroacetic acid to denature the proteins and remove non-covalently bound carbohydrates. The precipitate was collected by low-speed centrifugation, washed twice with ether and dried in a desiccator. The carbohydrate analysis was carried out by gas chromatography of the trimethylsilyl derivatives according to Bolton et al. [24! with some minor modifications. The procedure involves release of monosaccharides as the methyl glycosides by treatment with 1 M methanol/HC1 for 3 h at 95 °C. Reacetylation of amino sugars was then carried out by neutralizing with silver carbonate, addition of 2 0 ~ (w/v) acetic anhydride and heating at 95 °C for 2 h. The insoluble material was removed by centrifugation and the supernatant evaporated. Then 50/~l silyl reagent (pyridine/ trimethylsilylchloride/hexamethyldisilazane,5:1:1) was added and 2-/~1 samples were analyzed by gas chromatography at 180 and 220 °C using Helium as a carrier gas in an F. & M. Biomedical Gas Chromatograph, model 400, with a flame ionization detector.
Isoelectrofocusing Measurements of the isoelectric points of the lectins and of isolated chains were carried out in an LKB-7900 Uniphor apparatus (Bromma, Sweden) in the following way: A dense electrode solution containing 0.7 ml concentrated sulphuric acid in 60 ~ sucrose (w/v) was added to the anode at the bottom of the column. The sample (5-8 mg of protein) was mixed into a 230 ml sucrose gradient (0--40 ~ w/v) containing 1 ~ (w/v) of carrier ampholytes (Ampholine 4 0 ~ , pH 3-10, LKB-produkter AB, Bromma, Sweden) and 1 ~ Brij-58. After the column was filled, the light electrode solution (1 ~o w/v NaOH) was added to the cathode at the top of the column, and a potential of 400 V was applied to the column. Focusing was carried out at 2 °C for 75 h and then the column was emptied by the lower exit at a flow rate of about 1-2 ml/min. Fractions of 1 ml were collected and the pH and absorbance at 280 nm of each sample were measured.
Measurement of biological activities Toxicity of abrin and ricin expressed as LDso was measured after intraperitoneal injections into groups of mice [5, 6]. Activity of A-chains was measured in a cell-free system from rabbit reticulocytes as earlier described [11]. Various amounts of toxins or their A-chains were added and the amount necessary to give 90~o inhibition of protein synthesis was measured. The intact toxins were first treated with 1 ~ 2-mercaptoethanol at room temperature for one hour. In the case of toxins pretreated with proteolytic enzymes, the treatment with 2-mercaptoethanol was carried out in a somewhat different way. 100 #g of treated toxins were diluted in 1 ml of 50 mM Tris-HC1 (pH 7.7) containing 1 mg/ml bovine serum albumin. The solution was kept on ice and 2-mercaptoethanol was added to a final concentration of 50 mM, the samples were incubated at 0 °C for 1 h and their ability to inhibit protein synthesis was measured. In every case the controls were treated with 2-mercaptoethanol under exactly the same conditions. Activity of agglutinins was measured as ability to induce direct hemagglutination and the activity of B-chains as ability to induce indirect hemagglutination [7]. RESULTS AND DISCUSSION
Amino acid composition The amino acid composition of the isolated A- and B-chains of the toxins is shown in Table I. Our results for the ricin chains are in general agreement with those reported by previous authors [25, 26] whereas the amino acid composition of the abrin chains has not been determined earlier. Our results (Table I) indicate some similarity in amino acid composition between all four peptide chains. Statistical examination of the data using the SAQ test [20-23] indicated that abrin A-chain is closely related to ricin A-chain (SAQ ~ 24), and to abrin B-chain (SAQ ---- 17). Ricin A-chain appears to be closely relatedto abrin B-chain (SA Q = 29). Ricin B-chain is less closely related to the three other chains (SAQ ---- 56-85). The number of sulfhydryl groups in the different proteins was measured by their ability to bind N-ethyl[14C]maleimide. The results in Table II show that the intact toxins and agglutinins contain no reactive sulfhydryl groups, either in the absence or presence of urea in accordance with the data obtained by other authors
TABLE I AMINO ACID COMPOSITION OF THE CONSTITUENT PEPTIDE CHAINS OF ABRIN AND RICIN The values are given as residues per chain and are not corrected for loss during hydrolysis. The values in parentheses denote the range between data obtained with vacuum dialysis and performic acid oxidation. Half-cystine was calculated as cysteic acid after performic acid oxidation. N.D., not determined. Amino acid
Abrin
Ricin
A-chain Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-cystine Tryptophan Valine Methionine Isoleucine Leucine Tyrosine Phenyialanine Histidine Lysine Arginine
31.0 ( ± 19.8 (423.4 (433.3 (414.6 (416.4 (417.9 (42.0 N.D. 15.0 (42.6 (413.8 (422.4 (410.5 (412.1 (45.8 (44.6 (418.4 (4-
0.3) 0.1) 0.4) 0.1) 0.1) 0.5) 1.0) 0.6) 0.4) 0.8) 3.1) 0.5) 2.0) 0.8) 0.4) 1.4)
B-chain
A-chain
B-chain
27.9 (-4- 0.1) 21.8 (4- 0.6) 27.8 (4- 0.1) 31.4 ( i 0.1) 15.7 (4- 0.8) 18.3 (4- 0.4) 22.6 (4- 0.2) 2.2 N.D. 18.5 (4- 0.3) 4.1 (4- 0.2) 12.6 (4- 0.3) 27.2 (4- 0.1) 16.0 (4- 0.4) 11.8 (4- 0.1) 3.9 (4- 0.1) 7.8 (4- 0.0) 16.6 (4- 0.2)
27.3 ( t 0.3) 18.1 (-¢- 0.2) 20.0 (-4- 0.9) 31.2 (4- 0.6) 16.9 (4- 0.7) 18.5 (4- 0.6) 25.4 (4- 0.2) 1.2 N.D. 13.7 (4- 0.4) 2.2 (4- 0.5) 19.4 (4- 0.1) 23.1 (4- 0.2) 14.9 (4- 0.9) 14.3 (4- 0.1) 3.6 (4- 0.6) 1.9 (4- 0.7) 20.4 (4- 0.9)
43.6 (423.4 (421.6 ( + 23.9 (416.0 (423.0 (417.2 (46.0 N.D. 16.9 (42.9 (416.3 (426.9 (412.5 (44.8 (42.3 (47.9 (414.8 (4-
0.2) 0.2) 0.3) 0.3) 0.2) 0.3) 0.1) 0.6) 0.3) 0.1) 0.3) 0.5) 0.3) 0.1) 0.3) 0.5)
[25, 27, 29]. On the other hand, the free A- and B-chains were found to contain close to one reactive SH-group both in the presence and absence of urea (Table II). This supports the suggestion [29] that the two chains are connected by a single disulfide bond. Since no free SH-group could be detected in the intact toxins, the number of half-cystine residues in excess of one, found in each chain must be present as intrachain disulfides. In view of this it is possible that the half-cystine determinations in Table I are too low and that the correct values should be 3 residues of half-cystine in each abrin chain, 1 residue in ricin A-chain and 7 in ricin B-chain. TABLE II SULFHYDRYL GROUPS IN ABRIN, RICIN AND THEIR CONSTITUENT PEPTIDE CHAINS The buffer used was 0.05 M sodium phosphate (pH 8.0) containing 0.1 M NaC1. Incubation medium
No. of sulfhydryl groups per molecule Abrin
Buffer 10 M urea in buffer
Intact
A-chain
B-chain
0.02 0.03
0.73 0.70
0.62 0.64
Abrus agglutinin
Ricin Intact
A-chain B-chain
0.3 0.2
0.02 0.02
0.92 0.93
0.82 0.82
Ricinus agglutinin 0.3 0.3
Carbohydrate content Abrin and ricin are known to be glycoproteins but the published results show considerable differences as to the amount and type of carbohydrates present [3, 30, 31 ]. Our own results indicate (Table III) that both toxins contain mannose and glucosamine (probably N-acetylglucosamine) and that most of the carbohydrates are present in the B-chains. The amount of mannose and glucosamine (probably Nacetylglucosamine) present in the B-chains was almost the same in the two toxins. TABLE III CARBOHYDRATE CONTENT IN ABRUS AND RICINUS LECTINS Protein
Number of sugar residues per molecule of protein or polypeptide Mannose .Glucose Glucosamine
Abrin 10.0 Abrin A-chain 0 Abrin B-chain 9.7 Ricin 15.6 Ricin A-chain 4.3 Ricin B-chain 10.7 Abrus agglutinin 55.8 Ricinus agglutinin 25.4
1.6 0 0 2.3 0 0 2.4 4.4
1.6 0 1.9 2.4 0.4 1.33 6.8 4.9
Also various amounts of galactose were found in the isolated abrin B-chain. Since we did not find galactose in the intact abrin, we believe that the galactose in abrin B-chain represents noncovalently bound galactose trapped during the preparation of this rather unstable peptide chain. The small amounts of glucose found in intact abrin and ricin, but not in either of their isolated peptide chains, may similarly represent artifacts. Our data indicate that abrin A-chain is not a glycoprotein, in contrast to the findings in a previous report [32]. The two agglutinins also contain mannose and glucosamine, although in different amounts compared with the toxins. It is interesting that mannose and glucosamine have been found in a variety of other lectins as well [33].
lsoelectrofocusing of the leetins The data in Fig. 1 show that the p I of abrin (6.1) was only moderately lower than that of ricin (7.1). There was, however, a striking difference in the charge o f the constituent peptide chains of the toxins. Thus, the A-chain of abrin and the B-chain of ricin were rather acidic, whereas the B-chain of abrin and the A-chain of ricin were almost neutral. It is thus clear that both toxins consist of an acidic and a neutral polypeptide chain. The results in Fig. 2 show that abrus agglutinin is more acidic than abrin and that ricinus agglutinin is more basic than ricin. Funatsu et al. [34] found p I values of 7.34, 7.42 and 5.17 for intact ricin, ricin A-chain (their Ile-chain) and ricin B-chain (their Ala-chain), respectively, in general agreement with our own data. Isoelectrofocusing of abrus agglutinin carried out by Wei et al. [28] indicated a p I of 4.7 which is somewhat lower than the value (pl = 5.2) obtained by us.
i
i
i
[~
0.8
i .~.~.x'~
•
p I7 A
~.~~
/4
0.~
C
"6 0.8
y-~........
pIT,.s ......
D
.x- ~
-
pl &6] ,,'*'*/~ X I
E
I o.8
P~~ 5 ' " " :
F
I
~"
x.XF
x~. x"
~........
.E
~IL,.8 ~ .~
x,x" ×,
0.~
N'i I
I
~0
80
~0
80
Fraction number Fig. 1. Isoclcctrofocusing of abrin, ricin and their constituent pcptide chains. A) abrin; B) ricin; C) abrin A-chain; D) ricin A-chain; E) abrin B-chain; F) ricin B-chain. (O) Absorbancc at 280 nm; ( × ) pH of fractions.
2.4
12
A
E e- 2.0
/
e~ 1.6 ~ 1.2
x.x" pi-s.2 ~
x.X"
x.X.x/ x"x.X-
pi-7.8 ~ ,x.X.X"x'x'x x,ex"
x x "x
10 8
6Z
x.xX "x'x/x'x~x" Ii ) x.X-x'x ~~!
4 x
x.x"
! T
J
X
2
O •
20
40
60
~J II , , 80
100 Fraction
20
40
60
80
100
number
Fig. 2. Isoelectrofocusing o f abrus agglutinin (A) and ricinus agglutinin (B). (O) Absorbance at 280 n m ; ( × ) p H o f fractions.
Resistance of lectins to various physical and chemical treatments The intact lectins could be stored in the frozen state without loss of activity, and also in the absence of galaetose. Repeated freezing and thawing of the solution had little influence on their toxicity. In the presence of 0.1 M galactose, the toxins could be stored in the refrigerator for several months without loss of activity. The toxins were inactivated by boiling and abrin is largely inactivated by incubation at 60°Cfor lh. The isolated chains are much less stable than the intact toxins. Ricin B-chain could be stored at --20 °C in the absence of glycerol and abrin A- and B-chain could be sorted if 10% glycerol was present during the freezing. So far we have not been able to store ricin A-chain in the frozen state under any conditions. Similarly, the isolated chains are less resistant than the intact toxins to moderate heating. The intact lectins are stable over a wide pH range. However, the isolated chains of the toxins are much less stable, at least at acid pH (Table IV). Thus, treatment with 0.1 M acetic acid removed most of the biological activity of the isolated chains. Both intact toxins and isolated A- or B-chains could be incubated overnight in Tris buffer at pH 10.0 or in acetate buffer at pH 4.0 without loss of activity. Treatment with EDTA did not change the properties of the lectins. Treatment with sodium dodecyl sulphate, which inactivates most proteins, has no effect on the biological activity of the A-chains of either toxins (Table IV). The B-chains on the other hand lost at least 90 % of their ability to bind erythrocytes after this treatment. Due to the fact that trace amounts of sodium dodecyl sulphate induce hemolysis, the exact residual activity of the B-chains could not be measured in this system. The B-chains (especially that of ricin) are also more sensitive than the A-chains to treatment with chloramine-T. Since it is known that chloramine-T oxidizes methionine [35] this suggests that a methionine residue is located in close proximity to the binding sites for galactose. Interestingly, chloramine-T appears to form polymers of the B-chains which may result in precipitation of the Protein. Also other oxidizing agents may inactivate ricin [36]. The data in Table V demonstrate that intact abrin and ricin are extremely resistant to treatment with proteolytic enzymes, as are abrus and ricinus agglutinins (not demonstrated). In contrast, the isolated peptide chains of abrin and ricin are sensitive to proteolytic enzymes, the A-chains being somewhat more sensitive than the B-chains. As expected DNAase, RNAase and neuraminidase did not have any significant effect on the different lectin preparations.
Conformational changes of the toxins T h e interchain disulfide bond in abrin and ricin is easily reduced with 2mercaptoethanol even in the native proteins, whereas the intrachain disulfide bonds are more difficult to reduce. Only after denaturation of the toxins with high concentrations of urea could these S-S bonds be reduced (data not shown). Our data indicate that abrin A- and B-chain have each one internal disulfide bond, whereas ricin B-chain has three. We have previously produced immunological evidence that the ricin chains undergo conformational changes upon combining to form intact ricin [37]. Also, the isolated B-chain of abrin is immunologically different from the B-chain present in
100 100 -100 100 100 --
--
--
--
(c)
50 100 1 1 10 0
Abrin (b)
--
--
50 100 --30 10
(d)
20
60
1 5 0 100 80 80
Abrin A-chain (c)
.
3
10 10 0
