The Partial Amino Acid Sequence of Trypsin Inhibitor II from Garden Bean, PhaseoZus uulgaris, with Location of the Trypsin and Elastase-reactive Sites*

The amino acid sequences of garden bean (Phaseolus uulgaris) trypsin inhibitor II and a related molecular species II’ have been examined. The entire sequence of II’ has been determined with the exception of five internal residues, positions 24 to 28. Inhibitor II’ appears to be derived from II by loss of eight NH,-terminal residues. The garden bean inhibitors are highly homologous to the Bowman-Birk soybean inhibitor and lima bean trypsin inhibitor IV. The trypsin-reactive site has been located in the second half of the molecule, while the first reactive site has been found to be directed against elastase. Garden bean inhibitor II (and II’) is thus a double-headed inhibitor, simultaneously inhibiting 1 molecule of trypsin and 1 of elastase. The isolation of three isoinhibitors of trypsin from the garden bean, Phaseolus vulgaris, var. great northern, has been previously reported These inhibitors closely resemble other low molecular weight inhibitors such beans mung navy the Bowman-Birk inhibitor They low

From the Laboratory of Enzymology, Roswell Park Memorial Institute and the Department of Biochemistry, Roswell Park Division, State University of New York at Buffalo, Buffalo, New York 14263 The amino acid sequences of garden bean (Phaseolus uulgaris) trypsin inhibitor II and a related molecular species II' have been examined. The entire sequence of II' has been determined with the exception of five internal residues, positions 24 to 28. Inhibitor II' appears to be derived from II by loss of eight NH,-terminal residues. The garden bean inhibitors are highly homologous to the Bowman-Birk soybean inhibitor and lima bean trypsin inhibitor IV. The trypsin-reactive site has been located in the second half of the molecule, while the first reactive site has been found to be directed against elastase. Garden bean inhibitor II (and II') is thus a double-headed inhibitor, simultaneously inhibiting 1 molecule of trypsin and 1 of elastase.
The isolation of three isoinhibitors of trypsin from the garden bean, Phaseolus vulgaris, var. great northern, has been previously reported (1). These inhibitors closely resemble other low molecular weight inhibitors from legumes such as lima beans (2,3), mung bean (4), navy and kidney beans (5,6), and the Bowman-Birk inhibitor of soybean (7). They have in common low molecular weights (approximately 8,000) and amino acid compositions characterized by high contents of serine, aspartic acid, and half-cystine, and the absence of tryptophan and carbohydrate.
While all three inhibitors from Phaseolus uulgaris strongly inhibit bovine trypsin, they vary greatly in their inhibition of bovine a-chymotrypsin.
Inhibitor IIIb simultaneously and independently inhibits both trypsin and chymotrypsin. It therefore resembles the double-headed inhibitors typified by the Bowman-Birk soybean and lima bean IV inhibitors (3,(8)(9)(10)(11). However, both inhibitors I and II are essentially inactive toward chymotrypsin and inhibit only 1 mol of trypsin per mol of inhibitor.
This suggests that the second reactive site, normally directed against chymotrypsin, may have been changed during the course of evolution to a nonreactive form.
In the hope of elucidating the structural changes responsible * This research was supported by Grant HL-15892 from the National Heart and Lung Institute and Grant PRP-30 from the American Cancer Society. A portion of this work was taken from a thesis by K for the loss of activity at this reactive site, we have determined the amino acid sequence of one of the inhibitors.
Inhibitor II was chosen for several reasons. It is a major, if not the major, inhibitor species in the bean. Second, its action against trypsin varies significantly from that of the other two inhibitors. As previously noted, the dissociation constant of II with bovine & trypsin is two to three orders of magnitude higher than those of inhibitors I or IIIb (1). We thus also hoped to obtain from the primary structure some indication of the basis for this significant difference in reactivity.
This paper presents a partial sequence of isoinhibitor II' and most of the sequence of isoinhibitor II, and describes the location of the two independent reactive sites directed against elastase and trypsin. The Edman degradation of T-III proceeded until the glutamine residue at position 7 was exposed. Cyclization of this residue to pyrrolidonecarboxylic acid hampered further degra- nal sequence of T-S was unambiguously determined to be Ser-Met-Pro-Gly-Lys-by degradation. The peptides T-VIII, IX, and X are established in that order by overlaps supplied by chymotryptic peptides C-IV, V, and VI. This and C-I order T-VI and T-VII. Therefore, the order of the tryptic peptides in T-S is:

(T-VI)-(T-VII)-(T-VIII)-(T-IX)-(T-X).
T-L was found to contain the limit tryptic peptides T-II, III, IV, and V only. This is consistent with the amino acid composition of T-L (Table IV), assuming approximately 10% contamination, presumably with intact inhibitor. T-II and T-III have already been ordered by T-I, and may be placed at the NH,-terminus of T-L due to the blocked NH, terminus of T-II. Carboxypeptidase B digestion of T-L reveals a COOHterminal arginine. This is consistent with the previously observed presence of arginine at the reactive site of inhibitor II (1). Carboxypeptidase A digestion following carboxypeptidase B treatment rapidly released threonine, S-carboxamidomethylcysteine, and methionine, indicating T-V to be COOH-terminal in T-L. The limit tryptic peptides in T-L may thus be ordered: (T-II was carried out and the peptides fractionated in a manner identical with that with RCAM-II.
Peptides identical with T-III, IV, V, VI, VII, VIII, IX, and X were found in RCAM-II'.
Peptides T-I' and T-II', compared to the analogous peptides in RCAM-II, lacked the same 8 amino acid residues (2 histidyl, 3 aspartyl, 2 seryl, and 1 glutamyl) lacking in intact RCAM-II' (Table V). Unlike RCAM-II, the NH, terminus of RCAM-II' was reactive in the Edman reaction. Degradation in this manner directly places T-I' and T-II' at the NH,-terminal position (Table VI).
To elucidate the structure of T-II', the peptide was further fragmented with thermolysin. T-II', 167 nmol, was incubated with 13 Kg of thermolysin in 0.01 M Tris-Cl + 0.002 M CaCl,, pH 8.0, at 40". After 4 hours the reaction was stopped by freezing.
Preparative high voltage paper electrophoresis yielded two major components, Th-A (mobility 0.02 compared with the mobility of 0.27 of T-II') and Th-B (mobility 0.26). Together they account for the composition of T-II' (Table V). Th-B was found by dansylation to have both Ile-and Val-as NH, termini, but neither free Ile nor Val. It was therefore concluded that Th-B consists of an approximately equimolar mixture of Ile-Cys and Val-Cys (insufficient material was available for further fractionation).
The complete structure of Th-A was deduced as shown in Table VI Fig. 1. Electrophoresis was at pH 3.6 at 2 kV for 100 min as described in the text. The capital letters near the origin correspond to the pools in Fig. 1. Roman numerals to the right of each peptide indicate the corresponding tryptic peptide. The major staining peptides in each pool are cross-hatched; minor peptides are unhatched. Mobilities are relative to arginine. Pool A contained only a weakly staining peptide at the origin and is not shown. the resulting fractions by high voltage paper electrophoresis revealed over 40 distinct peptides, most in low yields. This apparently was caused by the susceptibility of bonds involving S-carboxamidomethylcysteinyl residues to chymotryptic cleavage. Several useful peptides were obtained, as presented in Tables II and III. Attention is at this point drawn to peptides C-II and C-III, which establish the sequence -Thr-Thr-Asx-Tyr-in T-VIII, and C-V and C-VI which establish the sequence -Ser-Asx-Ser-Gly-in T-X. Ordering of Tryptic Peptides of RCAM-II-Reduction and carboxamidomethylation of trypsin-modified RCAM-II resulted in two peptides which were separable on Sephadex G-50 (Fig. 4). Rechromatography under the same conditions resulted in peptides T-L and T-S. T-S was found to contain the tryptic peptides T-VI, VII, VIII, IX, and X only upon analytical tryptic digestion and paper electrophoresis.
The amino acid composition of T-S (Table IV) is consistent with this finding, assuming a 10 to 15% contamination with T-L. The NH,-termi-   ;.jzj 0.21) 1:9(z) :: succinate at pH 5.0. After 48 hours at room temperature, the modified inhibitor was subjected to Edman degradation. The absence of a reactive NH, terminus in intact inhibitor II allowed the unambiguous identification of the new terminus as Ser-Ile-Pro-.
The position of the elastase reactive site was thus established. DISCUSSION The ordering of the tryptic peptides in inhibitor II was accomplished by use of several chymotryptic peptides and the isolation of the two fragments from the trypsin modified inhibitor.
While these two fragments were not obtained in pure form, we are confident that the ordering data obtained are correct. The degree of contamination present would not be sufficient to invalidate the results of the analytical digests of T-L and T-S.
The complete amino acid sequence of inhibitor II' had been determined with the exception of 5 internal residues, 24 to 28, which have been placed by homology to the other two legume inhibitors.
The sequence of II is much less complete, with residues 1 to 21 unsequenced in addition. This difficulty arises  Fig. 3. *Purification steps: (a) electrophoresis at pH 3.6; (b) paper chromatography. due to the poor yields of the peptides T-I and T-II and the apparently blocked NH, terminus in the intact inhibitor (and these two peptides).
It seems likely, however, that at least residues 12 to 19 in II are identical with residues 4 to 11 in II'. A peptic digest' of T-II yielded two electrophoretic components, one corresponding in composition to residues 4 to 11 in inhibitor II'. The other component is equivalent in composition to the remainder of T-II and is presumably a mixture of two (or more) peptides corresponding to residues 1 to 11 and 20 to 23 (isoinhibitor II). If for the moment residues 9 to 21 of inhibitors II and 1 to 13 of II' are assumed to be identical, it is apparent that II' could arise from II by limited proteolysis with loss of the first 8 residues.
Comparison of the sequences of Bowman-Birk soybean inhibitor (8), lima bean inhibitor IV (19,20), and the garden bean isoinhibitors II and II' is made in Fig. 5. It is apparent that a high degree of homology exists among the four inhibitors. This is especially striking if both the NH,-and COOHterminal regions, which appear to be highly variable, are   tions predicted  for T-L containing  tryptic  peptides  T-II, III, IV and V,  and T-S containing  T-VI, VII, VIII, IX and X.    (including 4 half-cystine residues) and 5 are related by a single base change in their codons.
While the garden bean inhibitor in many ways resembles the soybean and lima bean inhibitors, it does differ in several quite significant respects. As noted in the introduction, this investigation was undertaken with the hope of elucidating the structure of the second, apparently inactive, reactive site of an inhibitor that was presumed should be double-headed.
In fact, the sequence suggested, and experiment has shown, that inhibitors II and II' are double-headed toward trypsin and elastase. While elastase inhibition by legume extracts and inhibitor preparations has been noted (21,22), this is believed to be the first instance in which such an inhibitor has been isolated and characterized. This is also believed to be the first elastase inhibitor reactive site to be sequenced. While inhibitors II and II' have now been eliminated as having a nonfunctional reactive site, the same can not be said of inhibitor I. Of the proteases tested, this species inhibits only trypsin significantly, and indeed either has an inactive reactive site or possesses activity toward an as yet unidentified protease.
A second significant difference between the garden bean and the lima and soybean inhibitors is the location of the trypsinreactive site. As previously noted, in both the lima and soybean inhibitors the first reactive site is directed against trypsin, while the second is reactive to chymotrypsin (8)(9)(10)(11). In garden bean inhibitors II and II' it is the second site that is directed against trypsin.
The following speculative explanation is proposed. These inhibitors presumably initially arose as a single-headed inhibitor which through gene doubling and condensation gave rise to a double-headed inhibitor.
The first double-headed inhibitor would presumably have both sites directed against the same enzyme. For the sake of simplicity, that enzyme can be assumed to be trypsin. At the present time, there are examples of both of these archetypes, the peanut (23) and mung bean (24) inhibitors, respectively. With the presence of two equivalent reaction sites, one site is free to further evolve without loss of the molecule's original function. If it is the second site that changes e.g. to a chymotrypsin reactive site, one obtains the configuration present in the lima and soybean inhibitors.
If, on the other hand, it is the first reactive site that evolves, e.g. to an elastase reactive site, one obtains the configuration in garden bean inhibitors II and II'. Finally, the question of the lower affinity (pKd 8.5) of inhibitors II and II' for trypsin as compared to the other garden bean inhibitors remains. The sequence around the trypsin reactive site of II' is quite similar to that of the Bowman-Birk inhibitor (pKd 10, Ref. 25). It is not immediately obvious how the exchange of methionine for alanine or asparagine could account for the large difference in complex stability. The change from a lysine-type to an arginine-type reactive site also seems unlikely to cause the large change in behavior. The

K A Wilson and M Laskowski, Sr
Phaseolus vulgaris, with location of the trypsin and elastase-reactive sites.
The partial amino acid sequence of trypsin inhibitor II from garden bean,