You are on page 1of 6


Archives of Biochemistry and Biophysics 410 (2003) 8388

Thiol-activated serine proteinases from nymphal hemolymph of the African migratory locust, Locusta migratoria migratorioides
Jacob Hanzon,a Patricia Smirno,a Shalom W. Applebaum,b Autar K. Mattoo,c and Yehudith Birka,*

Institute of Biochemistry, Food Science and Nutrition, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel b Department of Entomology, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel c Vegetable Laboratory, Henry A. Wallace Beltsville Agricultural Research Center, Building 010A, Beltsville, MD 20705, USA Received 8 July 2002, and in revised form 23 October 2002

Abstract Two unique serine proteinase isoenzymes (LmHP-1 and LmHP-2) were isolated from the hemolymph of African migratory locust (Locusta migratoria migratorioides) nymphs. Both have a molecular mass of about 23 kDa and are activated by thiol-reducing agents. PMSF abolishes enzymes activity only after thiol activation, while the cysteine proteinase inhibitors E-64, iodoacetamide, and heavy metals fail to inhibit the thiol-activated enzymes. The N-terminal sequence was determined for the more-abundant LmHP-2 isoenzyme. It exhibits partial homology to that of other insect serine proteinases and similar substrate specicity and inhibition by the synthetic and protein trypsin inhibitors pABA, TLCK, BBI, and STI. The locust trypsins LmHP-1 and LmHP-2 constitute a new category of serine proteases wherein the active site of the enzyme is exposed by thiol activation without cleavage of peptide bonds. 2002 Elsevier Science (USA). All rights reserved.
Keywords: Insect trypsins; Thiol activation; Trypsin inhibitors; Locusta migratoria migratorioides

Proteinases often occur in the form of catalytically inactive zymogen precursors. Enzyme activity is controlled by the timing of activation and by subsequent inhibition by appropriate proteinase inhibitors. Classical serine proteinase zymogens are activated by proteolysis [1,2], while cysteine proteinase zymogens are activated by proteolysis and thiol disulde exchange [2,3,22]. In insects, serine proteinase zymogens have been found in Drosophila embryos [4], in the cocoonase family of proteinases produced by the Bombyx mori [5], and in the hemolymph of B. mori [6]. Prococoonase from A. polphemus [23] has a molecular mass of 28 kDa, while the active enzyme has a lower molecular mass of 24 kDa. Trypsin-like and chymotrypsin-like enzymes have been isolated from the digestive tract of Locusta migratoria in their active form only [79]. Cathepsin-like enzymes activated in vitro by thiol-reducing agents have

been identied in the developing eggs of L. migratoria [10]. Hemolymph serine proteinases may play vital roles in various physiological processes in insects. A 29-kDa hemocyte proteinase dissociates the fat body at metamorphosis of Sarcophaga peregrina [16]. Hemolymph proteinases are involved in the activation of the prophenoloxidase cascade in Blaberus craniifer [17]. Data collected on hemolymph cysteine proteinase (CP1) from Drosophila melanogaster suggest a role in immunity, participating most likely in the degradation of internalized material in phagocytes [18]. We herein report on the purication and characterization of two novel thiol-activated serine proteinases from L. migratoria hemolymph. Materials and methods Reagents. Solvents were from Merck (Darmstadt, Germany). p-Nitroanilide chromogenic substrates, proteinase inhibitors, Sephadex G-75, and other reagents

Corresponding author. Fax: +972-8-9473654. E-mail address: (Y. Birk).

0003-9861/02/$ - see front matter 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 9 8 6 1 ( 0 2 ) 0 0 6 5 7 - 4


J. Hanzon et al. / Archives of Biochemistry and Biophysics 410 (2003) 8388

were from Sigma (St. Louis, MO). The ion exchangers were from Whatman (England). Insects. The African migratory locust (L. migratoria migratorioides) was reared in gregarious phase on grass and aked oats and at temperature of 2832 C. Fifth (last) instar nymphs were used in this study. Enzyme assays. Specic proteinase activity was assayed with various peptidyl-p-nitroanilide (pNA)1 substrates. After a series of initial experiments to determine suitable pH and constituents for optimal activity, assays were routinely performed on the serine protease chromogenic substrate Bz-Phe-Val-Arg-pNA. This was dissolved in dimethyl formamide (DMF) and diluted to a nal concentration of 2.5 mM in the reaction mixture, which was buered with 100 mM TrisHCl, pH 8.0, and supplemented with 2 mM dithiothreitol (DTT) and 2% butanol. Proteolytic activity was assayed after preincubation of the sample in assay buer for 20 min. After 30 min incubation at 37 C, proteolysis was terminated with 30% acetic acid, and activity measured at 410 nm as the release of pNA from Bz-Phe-Val-Arg-pNA. One unit of activity was dened as A410 nm=30 min . Inhibition of proteinase activity was assayed by preincubation of the enzymes for 30 min at 37 C in 2 mM DTT, which was later removed by dialysis, followed by a second preincubation step of 15 min at 37 C with various inhibitors. Residual enzyme activity was assayed with Bz-Phe-Val-Arg-pNA as detailed above. BowmanBirk inhibitor (BBI), soybean trypsin inhibitor (STI), phenylbutylamine (PBA), and p-chloromercuribenzoic acid (pCMB) were dissolved in buer. Phenylmethysulfonyl uoride (PMSF) was dissolved in iso-propanol. N-a-tosyl-L -lysine chloromethyl ketone (TLCK), N-a-tosyl-L -phenlalanine chloromethyl ketone (TPCK), p-aminobenzamidine (pABA), and HgCl2 were dissolved in ethanol. trans-Epoxy succinyl-L -leucylamido-(4-guanidino)butane (E-64), was dissolved in dimethyl sulfoxide (DMSO). Polyacrylamide gel electrophoresis. PAGE was performed according to Laemmli [19] in the presence of SDS. Gels were run in a Hoeer mini-gel apparatus at 120 V and 40 mA. Staining was performed with Coomassie blue R 250.

Proteolytic activity was detected in gels using 10% polyacrylamide containing 0.1% gelatin as substrate. Electrophoresis was performed according to Laemmli [19] except that SDS was omitted. The gels were incubated for 1 h at 37 C in 100 mM TrisHCl, pH 8.0, supplemented with 10 mM DTT, and then stained with Coomassie blue R 250. Protein content. The protein content was measured by the method of Bradford [20], using bovine serum albumin (BSA) as the standard protein, or estimated by absorbance at 280 nm. Sequence determination. The N-terminal amino acid sequence was determined by Edman degradation using an automated sequencer after reduction and alkylation of the peptides. The respective PTH-amino acid derivatives were identied by reverse-phase HPLC analysis.

Results Purication of trypsin-like hemolymph proteinases The sequence of steps leading to the purication of the two L. migratoria trypsins (LmHP-1 and LmHP-2) is presented in the form of a owchart in Fig. 1. Step 1. A total of 5 ml hemolymph was collected into 25 ml of 10 mM ammonium acetate, pH 6.5, from about 300 fth instar locust nymphs. The dorsal integumental membrane of each nymph, between the head capsule and the pronotum, was delicately punctured and hemolymph was transferred by capillary immediately into ice-cold buer. No darkening of hemolymph samples was observed during this process. Diluted hemolymph was centrifuged at 4 C for 10 min at 15,000g.

1 Abbreviations used: LmHP-1 and LmHP-2, Locusta migratoria hemolymph proteases 1 and 2; PMSF, phenylmethylsulfonyl uoride; pABA, p-aminobenzamidine; TLCK, N-a-tosyl-L -lysine chloromethyl ketone; BBI, BowmanBirk soybean inhibitor; STI, soybean trypsin inhibitor; pNA, p-nitroanilide; DMF, dimethyl formamide; DTT, dithiothreitol; PBA, phenylbutylamine; pCMB, p-chloromercuribenzoic acid; TPCK, N-a-tosyl-L -phenylalanine chloromethyl ketone; E64, trans-epoxy succinyl-L -leucylamido-(4-guanidino)butane; DMSO, dimethyl sulfoxide; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; PTH, phenylthiohydantoin; HPLC, highperformance liquid chromatography; EDTA, ethylenediaminetetraacetic acid.

Fig. 1. Flow chart of purication protocol for hemolymph trypsins from L. migratoria.

J. Hanzon et al. / Archives of Biochemistry and Biophysics 410 (2003) 8388


Fig. 2. Enzymes activity in situ in gels detected as digestion of 0.1% gelatin in native PAGE. Lane 1: Hemolymph proteolytic prole. Lane 2: Void volume fraction from DEAE-cellulose column. After electrophoresis the gel was incubated at 37 C for 30 min with TrisHCl, pH 8.0, containing 10 mM DTT and then was stained with Coomassie brilliant blue.

of 2.5 ml each were collected. Both enzymes migrated as a single peak (elution prole not shown). The peak fractions containing the proteinases were pooled and lyophilized. Specic activity increased 1550-fold with a yield of 53%. Step 4. The proteolytic activity that emerged from the Sephadex G-75 column was applied to a cation-exchange Mono-S column (Fig. 3A). HPLC was performed for 15 min with 10 mM ammonium acetate, pH 4.0, at a ow rate of 0.5 ml/min, followed by stepwise elution with increasing buer concentration after 15, 45, and 75 min to 20, 50, and 100 mM, respectively. The active fractions were pooled and lyophilized. This procedure separated the two isoenzymes, designated LmHP-1 and LmHP-2. LmHP-1 and LmHP-2 were homogeneous as shown by SDSPAGE in the presence of b-mercaptoethanol (Fig. 3B) with a recovery of 4.2 and 14.5%, respectively, and 3600-fold purication. Characteristics of LmHP-1 and LmHP-2 The molecular mass of both LmHP-1 and LmHP-2 is 23 kDa as indicated by SDSPAGE (Fig. 3B). Both are activated by the thiol-reducing agents DTT, b-mercaptoethanol, cysteine, and sulte ions (Fig. 4). DTT was routinely used to activate the two enzymes and subsequently removed by dialysis. Removal of the reducing agent after activation did not aect enzyme activity. The eect of proteinaceous and synthetic inhibitors on the activity of LmHP-1 and LmHP-2 on Bz-Phe-Val-ArgpNA is given in Tables 2 and 3. The serine protease inhibitor PMSF irreversibly inhibited the activated LmHP-1 and LmHP-2 but did not inhibit if added before the DTT activation. All of the four inhibitors of trypsin-like enzymes, TLCK, pABA, STI, and BBI, effectively inhibited the activation of LmHP-1 and LmHP-2. These results indicate the presence of essential serine and histidine residues at the active site. The specic cysteine proteinase inhibitor E-64 or the sulfydryl reagents pCMB, HgCl2 ; AgNO3 , and iodoacetamide, had no eect on LmHP-1 and LmHP-2 activity, either prior or subsequent to thiol activation. These results exclude the possibility of thiol being involved in the active site of the enzymes. The metal chelators EDTA

Step 2. The supernatant from step 1 was loaded onto a DE-52 anion-exchange column (1:4 25 cm) and equilibrated with 10 mM ammonium acetate, pH 6.5. The column was eluted at 4 C with three stepwise concentrations of NaCl (50, 200, and 500 mM) in the same buer at a ow rate of 20 ml/h. Four milliliter fractions were collected (elution prole not shown). Two proteinases eluted with the void volume while a minor peak of proteolytic activity appeared after elution with 0.5 M NaCl. The void volume was collected, pooled, and concentrated to a volume of 3 ml with an ultraltration cell (Amicon 8010) MWCO-3000. This step separates the two proteolytic activities, designated LmHP-1 and LmHP-2, from other hemolymph proteinases, as shown by gelatin-PAGE (Fig. 2). The major proteolytic activity was found to be puried 17-fold with a yield of 55% (Table 1). Step 3. The void volume eluted in step 2 was subjected to gel ltration on a Sephadex G-75 column (1:4 110 cm), equilibrated with 50 mM ammonium acetate buer, pH 6.5, at a ow rate of 12 ml/h Fractions

Table 1 Purication of LmHP-1 and LmHP-2 from nymphal hemolymph of Locusta migratoria Step Hemolymph DEAE-cellulose Sephadex G-75 Mono-S HP-1 HP-2 Volume (ml) 33 42 35 2 2 E280 nm 517 16.72 0.18 0.006 0.021 Total activity (U) 6000 3302 3150 252 870 Specic activity (U/E280 nm ) 11.6 197 18,010 42,000 41,430 Yield (%) 100 55 53 4.2 14.5 Purication factor 1 17 1550 3620 3571

One unit of activity was dened as A410 nm=30 min .


J. Hanzon et al. / Archives of Biochemistry and Biophysics 410 (2003) 8388

Fig. 4. Eect of various thiols on LmHP-2 activity. Puried proteinase was assayed at 37 C in 100 mM TrisHCl, pH 8.0, 2% butanol, and a thiol-reducing agent. Activities are expressed as A410 nm per 30 min. DTT s, cysteine D, Na2 SO4 d, b-mercaptoethanol .

Table 2 Activation mechanism of LmHP-1 and LmHP-2 by thiol reducing agent (DTT) Preincubation with Enzyme activity In the absence of DTT In the presence of DTT LmHP-1 Control (none) 0 DTT 2 mM 107 PMSF 10 mM 0 DTT 2 mM + PMSF 7 10 mM LmHP-2 0 108 0 4 LmHP-1 100 101 100 6 LmHP-2 100 98 102 5

LmHP-1 and LmHP-2 were preincubated in 100 mM TrisHCl, pH 8.0, at 37 C for 20 min, with dierent combinations of 10 mM PMSF and 2 mM DTT. After the preicubation step, the enzymes were dialyzed at 4 C and assayed in 100 mM TrisHCl, pH 8.0, containing 2% butanol, with or without 2 mM DTT at 37 C for 30 min. These values are expressed as percentages relative to control and are given as means (N 3).

Fig. 3. (A) Cation-exchange HPLC of the proteolytic peak from the Sephadex G-75 column on a column of Mono-S. Elution was performed for 15 min with 10 mM ammonium acetate, pH 4.0, at a ow rate of 0.5 ml/min, followed by stepwise elution with 20, 50, and 100 mM ammonium acetate, pH 4.0, after 15, 45, and 75 min, respectively. The upper graph is the protein elution prole monitored at 280 nm. The lower graph represents the plotted levels of pNA released by proteolytic activity of each individual fraction, as described under Materials and methods. (B) SDSPAGE of the puried proteinases LmHP-1 and LmHP-2 obtained after Mono-S column chromatography. Lane 1: puried LmHP-2. Lane 2: puried LmHP-1. Lane 3: Molecular weight markers.

and 1,10-phenanthroline did not aect the two enzymes, indicating that the enzymes are not metalloproteinases. The optimal pH of the proteinases for the hydrolysis of Bz-Phe-Val-Arg-pNA was in the range of 7.58.5 (data not shown). The enzymes were not inuenced by either of the cations, tested Na ; K ; Mg2 , or Ca2 . Short aliphatic alcohols, on the other hand, increased the activity; 70% activation was obtained by the addition of 2% butanol to the reaction mixture. Substrate specicity The specicity of puried LmHP-1 and LmHP-2 was investigated using a number of pNA substrates. The preferred substrate was Bz-Val-Leu-Arg-pNA (Table 4). Both enzymes preferred substrates with arginine at the P1 position. Substituting lysine for arginine at P1 reduced activity by 49 and 39% for LmHP-1 and LmHP-2,

J. Hanzon et al. / Archives of Biochemistry and Biophysics 410 (2003) 8388 Table 3 Eect of proteinase inhibitors on amidolytic activity of puried LmHP-1 and LmHP-2 Inhibitor Concentration % Activity remaining HP-2 None Serine proteinase inhibitors PMSF 10 mM TLCK 1 mM TPCK 1 mM PABA 10 mM PBA 10 mM STI 2 lM BBI 2 lM Cysteine proteinase inhibitors E-64 0.2 mM Iodoacetamide 100 mM 1 mM HgCl2 AgNO3 1 mM pCMB 1 mM Metalloproteinase inhibitors 1,10-Phenantroline 10 mM EDTA 10 mM 100 4 50 91 2 77 30 4 100 100 114 100 100 101 100 HP-1 100 5 33 90 3 71 6 5 101 100 110 100 102 100 102


LmHP-1 and LmHP-2, respectively, compared to -ValLeu-. Substituting Gly for Leu at the P2 position reduced activity by 60 and 40% for LmHP-1 and LmHP-2, respectively. No activity was found when substrates for chymotrypsin and elastase were used.

Discussion Two L. migratoria larval hemolymph isoenzyme proteinases, Lm HP-1 and LmHP-2, have been puried to homogeneity and characterized. Both enzymes are activated by thiol or sulte ions, which is characteristic of cysteine proteinases [3] but are not inhibited by E-64, a specic inhibitor of cysteine proteinases, nor by the sulphydryl reagents iodoacetamide, PCMB, HgCl2 , or AgNO3 . LmHP-1 and LmHP-2 are unaected by PMSF prior to thiol activation, but are completely inhibited by PMSF after thiol activation; PMSF inhibition of papain is reversed by reducing agents [21], but the inhibition of LmHP-1 and Lm HP-2 by PMSF is not reversed by DTT. These results suggest the presence of a serine residue at the reactive site as the nucleophilic agent and exclude the possibility of a cysteine residue being the nucleophilic agent. Classication of the enzymes as serine type proteinases is also supported by the similarity of LmHP-2 N-terminal sequence to those of other trypsin-like and chymotrypsin-like enzymes (Fig. 5). From the eect of the inhibitors pABA, SBTI, TLCK, and BBI (Table 3), and substrate specicity (Table 4), both LmHP-1 and LmHP-2 can be classied as trypsin-like enzymes. We conclude that these enzymes are serine proteinase zymogens activated by the opening of disulde bridge(s). Thiol-activated LmHP-1 and LmHP-2 are unrelated to the clip-domain family of serine proteinases [22]. Dierent from the hemolymph proteinases LmHP-1 and LmHP-2, no inactive zymogen have been identied for the trypsin- and chymotrypsin-like enzymes puried from L. migratoria digestive tract [7,8]. Trypsin zymogens in the digestive tract of insects have been previously

The puried enzymes were preincubated in 100 mM TrisHCl, pH 8.0, containing 2 mM DTT, and 2% butanol for 20 min at 37 C (enzyme activation). After dialysis to remove DTT, a second incubation step was performed with the inhibitor for 15 min. The activity was measured at 37 C with Bz-Phe-Val-Arg-pNA as substrate in 100 mM TrisHCl, pH 8.0, containing 2% butanol. These values are expressed as percentages relative to control and are given as means SD, N 3.

Table 4 Substrate specicity of puried LmHP-1 and LmHP-2 Substrate Relative activity (%) HP-1 Val-Leu-Arg-pNA Val-Leu-Lys-pNA Bz-Phe-Val-Arg-pNA Bz-Val-Gly-Arg-pNA Boc-Leu-Ser-Thr-Arg-pNA Bz-Pro-Phe-Arg-pNA Bz-Arg-pNA Ac-Ala-Ala-Ala-pNA Bz-Glu-Phe-Leu-pNA Ac-Tyr-pNA 100 51 75 40 50 39 3 0 0 0 HP-2 100 61 87 60 64 46 4 0 0 0

LmHP-1 and LmHP-2 were assayed at 37 C in 100 mM TrisHCl, pH 8.0, 2% butanol, and 2 mM DTT. Activities are expressed as a percentages of the maximal activity. Values are given as means (N 3). Abbreviations: Bz, N-a-benzoyl; Boc, N-tert-butoxy-carbonyl; Ac, acetyl.

respectively. The minimal substrate, Bz-Arg-pNA, was hydrolyzed at a very slow rate. The secondary specicity of the enzymes was examined by substituting groups at P2 ; P3 , and P4 . Replacement of the sequence -Val-Leuwith -Phe-Val-, -Pro-Phe-, or -Leu-Ser-Thr- resulted in reduced activity. The least favored substitution -ProPhe- resulted in 61 and 54% activity reduction for

Fig. 5. N-terminal sequence of LmHP-2, and its comparison with N-terminal sequences of other trypsin-like and chymotrypsin-like enzymes from insects.


J. Hanzon et al. / Archives of Biochemistry and Biophysics 410 (2003) 8388 [8] E. Sakal, S.W. Applebaum, Y. Birk, Int. J. Pept. Protein Res. 34 (1989) 98505. [9] W. Lam, G.M. Coast, R.C. Rayne, Insect Biochem. Mol. Biol. 27 (1999) 653660. [10] S. Kuk-Meiri, N. Lichtenstein, A. Shulov, M.P. Pener, Comp. Biochem. Physiol. 18 (1966) 783795. [11] C. Kellenberger, C. Boudier, I. Bermudez, J.G. Bieth, B. Luu, H. Hietter, J. Biol. Chem. 270 (1995) 2551425519. [12] N. Nakakura, H. Hietter, A. Van Dorsselaer, B. Luu, Eur. J. Biochem. 204 (1992) 147153. [13] E. Kromer, N. Nakakura, M. Lagueux, Insect Biochem. Mol. Biol. 24 (1994) 329331. [14] A. Hamdaoui, S. Wataleb, B. Devreese, S.J. Chiou, J. Vanden Broeck, J. Van Beeumen, A. DeLoof, L. Schoofs, FEBS Lett. 422 (1998) 7478. [15] Z. Malik, S. Amir, J. P al, Z. Buz as, E. V arallyay, J. Antal, Z. th, A. Patthy, L. Gr Szil agyi, K. Vek ey, B. Asbo af, Biochim. Biophys. Acta 1434 (1999) 143150. [16] S. Kurata, H. Saito, S. Natori, Dev. Biol. 153 (1992) 115 121. derh [17] C. Leonard, K. So all, N.A. Ratclie, Insect Biochem. 15 (1985) 803810. [18] Y. Tryselius, D. Hultmark, Insect Mol. Biol . 6 (1997) 173181. [19] U.K. Laemmli, Nature 227 (1970) 680685. [20] M.M. Bradford, Anal. Biochem. 72 (1976) 248254. ~or, Arch. Biochem. Biophys. 124 [21] J.R. Whitaker, J. Perez-Villasen (1968) 7078. [22] H. Jiang, M.R. Kanost, Insect Biochem. Mol. Biol. 30 (2000) 95 105. [23] J.H. Law, P.E. Dunn, K.J. Kramer, Insect proteases and peptidases, in: A. Meister (Ed.), Adv. Enzymol., vol. 35, Wiley, New York, 1977, pp. 389425. [24] A.M. Peterson, C.V. Barillas-Mury, M.A. Wells, Insect Biochem. Mol. Biol. 24 (1994) 463471. [25] C. Barillas-Mury, R. Graf, H.H. Hagedorn, M.A. Wells, Insect Biochem. 21 (1991) 825831. [26] J. Hanzon, Ph.D thesis, The Hebrew University of Jerusalem, 1997. [27] D. Koga, T. Funakoshi, K. Mizuki, A. Ide, K.J. Kramer, K.C. Zen, H. Choi, S. Muthukrishnan, Insect Biochem. Mol. Biol. 22 (1992) 305311. [28] R.I. Samuels, S.E. Reynolds, Arch. Insect Biochem. Physiol. 24 (1993) 3344.

deduced on the basis of isolation of cDNA from Manduxa sexta [24] and Aedes aegypti [25] midgut. Preliminary evidence [26] indicates that the activity prole of LmHP-1 and LmHP-2 in the hemolymph during the fth larval instar is coincident with apolysis and the molting process. This supports the notion that these hemolymph proteinases may be transferred at the appropriate time into the integumental molting uid and might participate in cuticle degradation. Chitinases of the silkworm (B. mori) and the tobacco hornworm (M. sexta) are presumably synthesized during the larval to pupal transition as inactive zymogens, which are apparently activated by limited proteolysis [27]. During the progress of cuticle degradation preceding eclosion of the tobacco hornworm to the adult form, the developmental proles of two molting uid proteases are in correlation to progression of cuticle degradation whereas chitin-degradative enzymes did not parallel weight loss from the old cuticle. Protease activity is a proposed to be prerequisite to chitin degradation by chitinase [28].

[1] L. Polg ar, in: A. Neuberger, K. Brocklehurst (Eds.), Hydrolytic Enzymes, Elsevier, Amsterdam, 1987, pp. 159170. [2] A.R. Khan, M.N.G. James, Protein Sci. 7 (1998) 815836. [3] K. Brocklehurst, F. Willenbrock, E. Salih, in: A. Neuberger, K. Brocklehurst (Eds.), Hydrolytic Enzymes, Elsevier, Amsterdam, 1987, pp. 39158. [4] C.L. Smith, R. DeLotto, Nature 368 (1994) 548551. [5] F.C. Kafatos, C.M. Williams, Science 146 (1964) 538540. [6] Y. Katsumi, H. Kihara, M. Ochiai, M. Ashida, Eur. J. Biochem. 228 (1995) 870877. [7] E. Sakal, S.W. Applebaum, Y. Birk, Int. J. Pept. Protein Res. 32 (1988) 590598.