Determination of the Amino Acid Sequence of the Plant Cytolysin Enterolobin

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 347, No. 2, November 15, pp. 201–207, 1997 Article No. BB970358 Determination of the Amino Acid Sequence of the Plant Cytolysin Enterolobin Wagner Fontes, Marcelo V. Sousa,1 Jeferson B. Aragão, and Lauro Morhy Centro Brasileiro de Sequenciamento de ProteıB nas, Laboratório de BioquıB mica e QuıB mica de ProteıB nas, Departamento de Biologia Celular, Universidade de BrasıB lia, 70910-900, BrasıB lia-DF, Brazil Received April 9, 1997, and in revised form August 25, 1997 The cytolytic seed protein enterolobin from seeds of Enterolobium contortisiliquum was purified by using FPLC on a Mono Q column giving a single peak in capillary electrophoresis. The complete amino acid sequence of the plant cytolysin was determined by an automated method, yielding a molecular mass of 54,806 Da. Databank searches and sequence alignment demonstrated a high degree of sequence identity and similarity between enterolobin and bacterial aerolysins from Aeromonas hydrophila and A. sobria. Several key residues involved in oligomerization of A. hydrophila aerolysin are conserved in enterolobin. Circular dichroism measurements and structural predictions revealed that enterolobin is very rich in b sheet, like aerolysin. Light-scattering studies revealed that enterolobin oligomerizes as a hexamer at pH levels below 7.0. NaCl concentrations above 50 mM caused dimerization of enterolobin. Dithiothreitol did not cause oligomerization. q 1997 Academic Press Key Words: amino acid sequence; plant protein; cytolysin; enterolobin; Enterolobium contortisiliquum; aerolysin. A previously reported partial amino acid sequence of enterolobin showed probable homology to sequences of aerolysins, cytolytic proteins from the gram-negative bacteria Aeromonas hydrophila and A. sobria (7). The mechanism of action by aerolysin from A. hydrophila has been the object of several studies (8–10). It is synthesized as preproaerolysin with a signal peptide that is cleaved during secretion. Lysis involves several steps, including binding of proaerolysin to a membrane receptor, activation to aerolysin by a cleavage near the C-terminus, insertion into the membrane, and oligomerization to form a pore. The mechanism of action for enterolobin is under investigation to see if it is similar to that of aerolysin. In this paper, further purification of enterolobin was achieved and its complete amino acid sequence was determined. New database searches and sequence analyses including alignment and structural predictions were carried out. Circular dichroism was performed to corroborate secondary structure predictions. Light scattering was used to examine the effect of pH, salt, and DTT2 on the oligomerization of enterolobin. MATERIALS AND METHODS In nature, cytolytic proteins are found in beings from microorganisms to man and are very diverse in size, structure, and mechanism of action (1–3). However, very few examples have been studied in plants. Enterolobin is a cytolytic protein purified from seeds of a brazilian rainforest tree, Enterolobium contortisiliquum (Leguminosae-Mimosoideae) (4). Enterolobin may have a defensive role in the seed, since it has shown insecticidal activity against larvae of the bruchid Calosobruchus macullatus (5) and inflammatory activity upon injection in rats (6). 1 To whom correspondence should be addressed. Fax: 55-612724548. E-mail: [email protected]. 0003-9861/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved. Materials. Seeds of E. contortisiliquum (Vell.) Morong were collected from trees at the Universidade de BrasıB lia campus. Suspensions of human red blood cells type O/ in 50 mM EDTA/150 mM NaCl were obtained from Hemocentro de BrasıB lia. Reagents and solvents for protein sequencing were purchased from the Applied Biosystems division of Perkin–Elmer. Enzymes, reagents, and solvents for purification of enterolobin and production and separation of peptides were all of sequanal grade from several sources. Other reagents were of analytical grades. Milli Q water was used for making the solutions. Purification of enterolobin. The purification of enterolobin was based on the previous method of Sousa et al. (7). Some modifications were introduced. Before extraction, the seed powder was defatted by cold acetone (0207C) for 30 s under agitation in an Omni Mixer and 2 Abbreviation used: DTT, dithiothreitol. 201 202 FONTES ET AL. FIG. 1. FPLC anionic-exchange chromatography. A sample of 200 mg from the fraction from DEAE–cellulose batch separation was dissolved in 500 ml of 0.2 M ammonium bicarbonate buffer, pH 8.7, and injected onto a Mono Q column. The flow rate was 1 ml/min. A gradient from 0.2 to 1 M ammonium bicarbonate, pH 8.7, was used for elution. Peak 2 contained all the hemolytic activity and was assigned to enterolobin. then dried at room temperature. The extraction and precipitation with ammonium sulfate were done as in Sousa et al. (7), excepting that the saturation range was now between 25 and 40%. Similarly, this material was submitted to DEAE–cellulose batch separation in 50 mM ammonium bicarbonate pH 8.2 buffer. Nonadsorbed protein was filtered, dialyzed against water, and freeze-dried. The fraction from the DEAE–cellulose batch separation was dissolved in 20 mm ammonium bicarbonate pH 8.7 buffer at a final protein concentration of 0.4 mg/ml and applied onto a semi-preparative Mono Q HR 10/10 column fitted in a FPLC system (Pharmacia). A concentration gradient was used for elution: from 0 to 10% 1 M ammonium bicarbonate, pH 8.7, in 25 min and from 10 to 100% in 10 min. Flow rate was 1 ml/min and detection was at 280 nm. The material from the hemolytic peak (enterolobin) was dialyzed and dried down in a Speed Vac (Savant). Hemolysis assays were carried out as previously described (4, 7). Capillary electrophoresis. Capillary electrophoresis (CE) was used for checking the purity of the samples during purification and for the determination of molecular weight and pI of enterolobin. A 270HT Applied Biosystems CE apparatus was employed following the instructions and utilizing the kits provided by the manufacturer. For purity evaluation a free solution system (40 mM sodium phosphate, pH 2.5 ) was used. Samples volumes of 10 nl were applied by vacuum to a 50-mm capillary filled with buffer. A tension of 20 kV was kept during the run at 257C. Detection was at 215 nm. Molecular weight determination was carried out by using the Pro Sort kit. The Pro Focus kit was used for pI determination. Amino acid analysis and protein sequencing. Enterolobin was reduced and S-pyridylethylated according to (11). Nine samples of 100 mg of reduced and S-pyridylethylated enterolobin were submitted to automatic hydrolysis (vapor-phase HCl at 1607C for 75 min), precolumn PITC derivatization, and amino acid analysis on an Applied Biosystems Model 420H/130A amino acid analyzer. Parallel amino acid analysis in a post-column ninhydrin derivatization system was also carried out. Three samples of 300 mg of the protein were hydrolyzed in a PicoTag workstation using vapor-phase HCl at 1097C for 24 h. Analyses were done in an Hitachi L-8500 amino acid analyzer. Tryptophan was determined spectrophotometrically (12, 13). Enzymatic cleavages (with trypsin, chymotrypsin, V8 protese, and elastase) and chemical cleavages (with cyanogen bromide, BNPSskatole, and iodosobenzoic acid) were conducted as in (14). Separations of peptides were carried out on Vydac C4 (digests from chemical cleavages) or Applied Biosystems C18 (enzymatic digests) columns fitted to a Model 2150 LKB HPLC using acetonitrile gradient schemes as previously described (7). Sequencing was performed in an Applied Biosystems Model 477A/120A protein sequencer. Database searches and sequence analyses. The amino acid sequence of enterolobin was searched against most sequence databases, using three different algorithms: Blast at http://expasy.hcuge. ch/cgi-bin/BLAST.pl (15), Blitz at http://www.ebi.ac.uk/searches/ blitz_input.html (16), and Bic at http://sgbcd.weizmann.ac.il/Bic/Exe cAppl.html. Clustal V (17) was used for sequence alignment. Amino acid composition and other properties derived from the sequence were used as search parameters by the software PropSearch at http:// www.embl-heidelberg.de/prs.html (18) using the SwissProt database. Theoretical pI and molecular mass were calculated from enterolobin sequence by ProtParam at http://expasy.hcuge.ch/sprot/ protparam.html. Secondary structure prediction was performed by the method GOR (19) at http://molbiol.soton.ac.uk/compute/GOR.html. Circular dichroism. A circular dichroism (CD) spectrum of enterolobin (108 mg/ml) in Milli Q water was obtained in a Cary 61 differential dichrograph. Quartz cells had a 0.05-cm pathlength. The FIG. 2. Free solution capillary electrophoresis. Samples from three different purification steps of enterolobin were checked for purity, namely ENT-N (fraction obtained from ammonium sulfate precipitation), ENT-D (fraction from DEAE–cellulose batch separation), and ENT-F (fraction from FPLC/Mono Q chromatography). Aqueous solutions of 1 mg/ml of the samples were injected onto a 50-mm capillary filled with sodium phosphate buffer, pH 2.5, and 20 kV was applied to the capillary ends. Detection was at 215 nm. 203 AMINO ACID SEQUENCE OF ENTEROLOBIN TABLE I Amino Acid Compositions of Enterolobin and Preproaerolysin Analysis method Amino acid Precolumn Postcolumn Enterolobin sequence Preproaerolysin sequence Ala Arg Asx Cys Gly Glx His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val 35 26 57 6 25 52 14 29 32 35 6 19 16 33 41 18a 20 41 38 20 57 n.d. 44 58 11 23 32 27 5 21 21 47 38 09a 09 35 34 21 59 9 25 46 9 27 32 30 6 21 21 33 41 15 16 40 35 22 63 4 48 42 6 25 35 24 7 11 26 39 31 18 21 36 Note. n.d., not determined. a Determined by the spectrophotometric methods of Goodwin and Morton (12) and Bencze and Schmid (13). recording range was between 185 and 260 nm. The algorithm of (20) was employed for the determination of percentage of secondary structures. Light scattering. Light scattering of enterolobin was obtained at 257C in a Perkin–Elmer MPF 2-A spectrofluorometer with both monochromators set to the same wavelength (450 nm) and same slit width (6 mm). Three factors were tested independently. pH was adjusted from 10.6 to 5.5 by additions of 2.0 M HCl to a solution of enterolobin (84.9 mg/ml in unbuffered Milli Q water). Ionic strength was adjusted by increasing the NaCl concentration up to 200 mM (62.8 mg/ml in unbuffered Milli Q water). DTT concentration was increased up to 10 mM (55.0 mg/ml in unbuffered Milli Q water). The sample concentration was kept constant since the volumes of HCl, NaCl, and DTT solutions added at each step were very low (approximately 1 ml) compared to the total volume of 3 ml. To convert light scattering to molecular mass, the expression used was M Å Ms (R907/C) (R907/C)s in which M means molecular mass to be determined, Ms means molecular mass of the standard, R907 means Rayleigh scattering, and C means concentration (21). BSA was used as standard under the same conditions of the sample for all studies. RESULTS The introduction of some modifications to the original method of purification of enterolobin (7) converted it to a faster and more efficient process. In particular, the FPLC on a Mono Q column was of great utility, since in a single step it provided enterolobin (Fig. 1) in a pure state as shown by a single peak obtained by capillary electrophoresis (Fig. 2). Capillary electrophoresis (using the Pro Focus kit) determined the pI of enterolobin to be 7.2, in agreement with a former pI determination of 7.0 by polyacrylamide gel isoelectric focusing (4). The pI predicted from the amino acid sequence of enterolobin was 7.5, which is close to the experimental pI. A molecular mass of 55,200 Da was determined by using the Pro Sort kit in capillary electrophoresis. This mass is also very close to the previously determined mass of 55,000 Da by means of gradient SDS–PAGE (4) and to the molecular mass calculated from the amino acid sequence of enterolobin (54,806 Da). The amino acid composition of enterolobin, determined by two different methods and also derived from the sequence (Table I), was compared with the compositions of preproaerolysins from A. hydrophila and A. sobria. Although there is some similarity in composition among enterolobin and preproaerolysins, computational searches of amino acid composition did not reveal any significant similarity. However, enterolobin composition was shown to be similar to that of the Vibrius vulnificus cytolysin precursor. The complete amino acid sequence of enterolobin was assembled from anteriorly sequenced regions (7) and some newly sequenced peptides obtained from enzymatic and chemical cleavages (Fig. 3). The enterolobin sequence was aligned against sequences of preproaero- 204 FONTES ET AL. FIG. 3. Amino acid sequence of enterolobin. The arrows indicate the sequence extension on the fragments produced by cleavages with chymotrypsin (C), trypsin (T), elastase (E), S. aureus V8 protease (V), cyanogen bromide (CB), BNPS-Skatol (S), and iodosobenzoic acid (IB). Fragments labeled with OV are overlaps obtained previously as well as the cyanogen bromide fragment CN H-1 (7). lysins from A. hydrophila (44.6% identity, 59.4% similarity) and A. sobria (42.7% identity, 57.6% similarity), the only sequences obtained with high similarity scores regardless of the search algorithm used. The sequences of enterolobin and the better studied preproaerolysin from A. hydrophila are aligned in Fig. 4. Secondary structures of enterolobin were determined by circular dichroism and predicted from the amino acid sequence by the GOR method (Table II). The programs suggest that enterolobin and A. hydrophila proaerolysin are rich in b structures. Light-scattering experiments (Fig. 5) showed that oligomerization of enterolobin molecules was promoted by low pH and high ionic strength. DTT did not cause the association of the protein. As shown in Fig. 5a, the average molecular mass of enterolobin increased as pH decreased, mainly in the range between 7 and 5. The average molecular mass also increased as a function of increased NaCl concentration (Fig. 5b). Although the scattering intensity had a little variation in the presence of DTT, no significant mass variation of enterolobin was observed (Fig. 5c). DISCUSSION In a previous report, amino acid analysis and partial amino acid sequencing of enterolobin revealed that enterolobin and aerolysins from A. hydrophila and A. sobria were probably similar proteins (7). In the present report, amino acid analyses by both precolumn and postcolumn derivatization methods (Table I) confirm the previous results. PITC pre-column derivatization AMINO ACID SEQUENCE OF ENTEROLOBIN 205 FIG. 4. Sequence alignment between enterolobin (ent) and preproaerolysin (aer) from Aeromonas hydrophila. Numbering of amino acid residues was based on the preproaerolysin sequence (9) as deposited in the SwissProt database (Accession No. P09167). The signs Å and / indicate identical and similar residues, respectively. Under specific amino acid residues (bold), symbols @ , # , l, and r adjacent to numbers indicate sites important for oligomerization, binding, secretion, and proteolytic activation of aerolysin, respectively. Symbol * is related to residues with no defined function. The signal peptide of preproaerolysin and the similar region in enterolobin are in italics. The site of proteolytic processing of the signal peptide in preproaerolysin (23) is represented by D. 206 FONTES ET AL. TABLE II Percentage of Secondary Structures of Enterolobin Compared to Proaerolysin Proaerolysin Enterolobin Structure type GOR CD GOR (CD) (24) (X-ray) (25) a b sheet b turn b total Random 24.30 30.50 27.10 57.60 21.50 8.40 32.90 21.40 54.30 37.30 16.80 31.70 27.30 59.00 27.70 11.00 50.00 20.00 70.00 19.00 17.00 40.00 — — — followed by reverse-phase HPLC separation proved to be a more reliable method, since it provided an amino acid composition of enterolobin closer to the actual one (obtained from the sequence) than the ninhydrin postcolumn reaction with ion-exchange separation. In addition, the alignment between the complete amino acid sequences of enterolobin and preproaerolysin from A. hydrophila (Fig. 4) corroborates the high similarity and identity scores for the two cytolysins, despite the distant phylogenetic relationship of these species. Moreover, several key residues involved in oligomerization of A. hydrophila aerolysin are conserved in enterolobin. Residues His107, His132, and Cys159 (22, 23) are involved in the oligomerization of the bacterial cytolysin, and the aligned residues His127, His154, and Cys183 are found in the plant cytolysin. In proaerolysin, Trp371 and Trp373 are important for the maintenance of the dimer in solution (24), which must dissociate prior to oligomerization (24). Enterolobin contains Trp386 and Trp387 at related positions. All these residues are located in domain 2 of the tertiary structure of proaerolysin (25), which is suggested to be an important domain for the oligomerization of the cytolysin. Enterolobin also shows oligomerization under certain conditions in vitro as shown by the light-scattering results. Histidine residues related to the binding of proaerolysin to the membrane receptor (22) are not conserved in enterolobin, since His186 is deleted in enterolobin and His332 is substituted by Glu351. Mutated proaerolysins with Asn in such positions reveal a considerable lower affinity for the erythrocyte receptor (22, 25). These substitutions in enterolobin could mean a different binding mechanism for plant cytolysin or a different receptor. The LysValArg429 region in proaerolysin is the site for tryptic and chymotryptic activation with removal of a C-terminal fragment (9), which has the corresponding nonconserved AspSerCys439 tripeptide in enterolobin. However, there are other possible cleavage sites for trypsin and chymotrypsin around this region in enterolobin, although it is not known if proteolytic cleavage in entero- lobin causes any further increase in hemolytic activity in vitro and in vivo. In proaerolysin from A. salmonicida, Trp227 plays a role in the secretion of the protein into the media (26). In enterolobin Trp247 is aligned with proaerolysin Trp227. PropSearch included amino acid composition and other sequence-derived properties as search parameters. Its FIG. 5. Light scattering of enterolobin. Scattering readings were made at 450 nm with a slitwidth of 6 nm for both monochromators. (a) The pH of a solution of enterolobin in water (84.9 mg/ml) was adjusted to 10.6 with 3 M NaOH and then decreased by steps to 5.5 with 2.0 M HCl. (b) The NaCl concentration of a solution of enterolobin in water (62.8 mg/ml) at pH 10.6 adjusted with 3 M NaOH was gradually increased up to 200 mM. (c) The DTT concentration of a solution of enterolobin in water (55 mg/ml) at pH 10.5 adjusted with 3 M NaOH was increased up to 10 mM by steps. AMINO ACID SEQUENCE OF ENTEROLOBIN output suggests that there could be similarity of enterolobin and the Vibrius vulnificus cytolysin precursor. However, alignment of their sequences (not shown) demonstrates that only the tryptophan cluster between Trp394 and Trp396 shows some similarity. This region is responsible for oligomerization of aerolysin (23, 26). The percentages of secondary structures (Table II) determined by circular dichroism show a high amount of b structures for both enterolobin and A. hydrophila proaerolysin. The crystal structure of proaerolysin shows less b structure than the circular dichroism method, although b sheet structures still form the majority of the molecule. The GOR method gave acceptable predictions for both proteins. This secondary structure similarity between the two cytolysins may reflect tertiary structure conservation, which is normally higher than the sequence similarity. Oligomerization of enterolobin, as analyzed by light scattering (Fig. 5), is influenced by pH. At pH 10.6, enterolobin shows an average molecular mass from 55.2 kDa, representing a predominance of the monomeric form. Its molecular mass increases to 330 kDa at pH 5.5, approximately corresponding to a hexamer. Although there is no evidence in the literature for the influence of the pH on the oligomerization of aerolysin, there is one study showing the oligomerization of the Pseudomonas aeruginosa toxin on erithrocyte membranes in vivo, in a pH range from 6.0 to 6.8 (27). The oligomerization of enterolobin induced by increasing the ionic strength produced only the average molecular mass corresponding to the dimeric form of the protein. Increasing the ionic strength decreases the oligomerization rate for aerolysin (9). Despite all the sequence and structure similarities, the oligomerization mechanism may be different between the cytolysins. The small increase of light scattering of enterolobin upon addition of DTT, with no change on the average molecular mass, suggests that some disulfide bonds were disrupted, but no oligomerization occurred. At the moment, we are trying to obtain crystals of enterolobin. When its crystal structure is solved, a comparison with the proaerolysin structure will be possible and new conclusions on the structure–activity relationship of enterolobin and aerolysin will be made. Site-directed mutagenesis based on structural guidelines will also be of great help to confirm and determine other important sites for the binding, oligomerization, and pore formation by enterolobin. The current data suggest that enterolobin and aerolysin are very similar cytolytic proteins, with some coincident steps in their mechanisms of action. ACKNOWLEDGMENTS The authors thank R. B. Cunha and N. M. Domingues for operating the protein sequencer and the amino acid analyzers, respectively. Dr. Carlos André O. Ricart is acknowledged for critical reading of the manu- 207 script. W. Fontes was a recipient of a CNPq studentship. Financial support came from PADCT (Brazil), FAPDF (BrasıB lia), and IFS (Sweden). REFERENCES 1. Berheimer, A. W., and Rudy, B. (1986) Biochim. Biophys. Acta 864, 123–141. 2. Bernheimer, A. W. (1970) in Bacterial Protein Toxins (Ajl, K. S., and Montile, T. C., Eds.), pp. 183–212, Academic Press, New York. 3. Bhakdi, S., and Tranum-Jessen, J. (1987) Rev. Physiol. Biochem. Pharmacol. 107, 147–223. 4. Sousa, M. V., and Morhy, L. (1989) Ann. Acad. Bras. Ci. 61, 405– 412. 5. Sousa, M. V., Morhy, L., Richardson, M., Hilder, V. A., and Gatehouse, A. M. R. (1993) Entomol. Exp. Appl. 69, 231–238. 6. Castro-Faria-Neto, H. C., Cordeiro, R. S. B., Martins, M. A., Correia-da-Silva, A. C. V., Bossa, P. T., Sousa, M. V., and Morhy, L. (1991) Mem. Inst. Oswaldo Cruz 86, 129–131. 7. Sousa, M. V., Richardson, M., Fontes, W., and Morhy, L. (1994) J. Protein Chem. 13, 659–667. 8. Buckley, J. T. (1991) Experientia 47, 418–426. 9. Van der Goot, F. G., Lakey, J., Pattus, F., Kay, C. M., Sorokine, O., Van Dorsselaer, A., and Buckley, J. T. (1992) Biochemistry 31, 8566–8570. 10. Buckley, J. T. (1992) FEBS Lett. 307, 30–33. 11. Wilson and Yuan (1989) in Protein Sequencing—A Practical Approach (Findlay, J. B. C., and Geisow, M. J., Eds.), pp. 21–22, IRL Press, Oxford. 12. Goodwin, T. W., and Morton, R. A. (1946) Biochem. J. 40, 628. 13. Bencze, W. C., and Schmid, K. (1957) Anal. Chem. 29, 1193– 1196. 14. Aitken, A., Geisow, M. J., Findlay, J. B. C., Holmes, C., and Yarwood, A. (1989) in Protein Sequencing—A Practical Approach (Findlay, J. B. C., and Geisow, M. J., Eds.), pp. 61–71, IRL Press, Oxford. 15. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. Mol. Biol. 215, 403–410. 16. Sturrock, S. S., and Collins, J. F. (1993) Blitz software version 1.5, Biocomputing Research Unit, University of Edinburgh, UK. 17. Higgins, D. G., and Sharp, P. M. (1989) Comput. Appl. Biosci. 5, 151–153. 18. Hobohm, U., and Sander, C. (1995) J. Mol. Biol. 251, 390–399. 19. Garnier, J., Osguthorpe, J. D., and Robson, B. (1978) J. Mol. Biol. 120, 97–120. 20. Compton, L. A., and Johnson, W. C. (1986) Anal. Biochem. 155, 155–167. 21. Tounelat, J., and Guinaud, S. (1973) in Experimental Methods in Biophysical Chemistry (Nicolau, C., Ed.), p. 149, Wiley, NY. 22. Green, M. J., and Buckley, J. T. (1990) Biochemistry 29, 2177– 2180. 23. Wilmsen, H. U., Buckley, J. T., and Pattus, F. (1991) Mol. Microbiol. 5, 2745–2751. 24. van der Goot, F. G., Ausio, J., Wong, K. R., Pattus, F., and Buckley, J. T. (1993) J. Biol. Chem. 268, 18272–18279. 25. Parker, M. W., Buckley, J. T., Postma, J. P. M., Tucker, A. D., Leonard, K., Pattus, F., and Tsernoglou, D. (1994) Nature 367, 292–295. 26. Wilmsen, H. U., Leonard, K. R., Tichelaar, W., Buckley, J. T., and Pattus, F. (1992) EMBO J. 11, 2457–2463. 27. Ohnishi, M., Hayashi, T., Tomita, T., and Terawaki, Y. (1994) FEBS Lett. 356, 357–360.