Aprotinin recovery: comparison between biospecific and pseudobiospecific affinity adsorptions

APROTININ RECOVERY: COMPARISON BETWEEN BIOSPECIFIC AND PSEUDOBIOSPECIFIC AFFINITY ADSORPTIONS

R. E. TAMAGAWA¹, A. R. AZZONI¹, E. A. MIRANDA¹^* and M. A. VIJAYALAKSHMI²

¹Departamento de Processos Biotecnológicos, FEQ, UNICAMP, C.P. Box 6066, 13083-970 Campinas - SP, Brazil. Phone/fax: [phone], E-mail: [e-mail] ²Laboratoire d'Interaction Moléculaire et des Technologie des Séparations, Université Technologie de Compiègne, Centre de Recherches de Royallieu, BP 529; 60206 Compiègne Cedex - France, E-mail: [e-mail]

(Received: January 19, 1999; Accepted: March 23, 1999)

Abstract - Two adsorption techniques were studied as potential methods for the recovery of aprotinin: a biospecific adsorption (using immobilized trypsin and chymotrypsin as ligants) and a pseudobiospecific adsorption of aprotinin-trypsin complex onto IMAC matrix. These studies indicated that ionic strength has an important effect on aprotinin adsorption on immobilized trypsin, and the pH was the most significant variable in aprotinin desorption from both immobilized enzymes. The aprotinin-trypsin complex adsorption onto the IMAC matrix was not as significantly affected by changes in pH as it was for changes in ionic strength. For the desorption step, both variables significantly affected the complex recovery. However, the effect of ionic strength was markedly stronger for this desorption step.

Keywords: Aprotinin recovery, adsorptions, IMAC matrix.

INTRODUCTION

Aprotinin is a bovine protein with medical applications – mainly as an anticoagulant in cardiac surgery - due to its property of inhibiting serine proteases (Baufreton et al., 1996). In this work, two adsorption methods were compared in order to develop a process to recover aprotinin from its natural (bovine tissues) and recombinant sources (Barthel and Kula, 1993). In the first method (biospecific adsorption onto enzymes inhibited by aprotinin), trypsin and chymotrypsin were chosen as ligants since these enzymes are inhibited by aprotinin. In the second method (pseudobiospecific adsorption onto immobilized metal affinity chromatography, IMAC, matrix), we planned to adsorb aprotinin as a complex with trypsin. The reason for the use of the complex is that protein adsorption onto IMAC matrix requires that at least one histidine residue on the protein surface be available for interaction with the chelated metal. Aprotinin does not have histidine residues; swine trypsin does (four histidine residues).

MATERIALS AND METHODS

Adsorption Matrix and Chemicals

The Sepharose 4B and CNBr-activated Sepharose 4B were Pharmacia (Sweden) products. The IMAC matrix used, Prosep Chelating-I (PCI), a silica matrix containing iminodiacetic acid (IDA) as the chelating agent, was donated by Bioprocessing (England). The other products were trypsin and chymotrypsin (swine and bovine) donated by Biobrás S/A (Brazil); aprotinin, benzoil-arginina-etil-ester (BAEE), 1,4 butanodiol diglicidil ether, and benzamidine from Sigma (EUA); and glycine and aprotinin from Merck (Germany).

Determination of Aprotinin Concentration

Aprotinin concentration was determined by inhibition of trypsin esterase activity using the method presented by Burck (1970). One inhibition unit is the amount of inhibition that causes a reduction in BAEE hydrolysis by one unit of absorbance at 253 nm per minute.

Immobilization of Trypsin and Chymotrypsin on Agarose

Sepharose oxirane activation was carried out according to the procedure presented by Hermanson et al. (1992). The enzyme immobilization on activated agarose was carried out in a 0.5 M bicarbonate buffer, pH 9.5 at 25 °C for 24 hours. The resin was sequentially washed with 1.0 M NaCl in 0.05 M tris-HCl, pH 8.0 and 25 mM HCl, 0.01 M CaCl₂ solutions. Nonreacted sites were quenched with 1.0 M glycine solution, pH 9.0. The two proteases were also immobilized using CNBr-activated Sepharose 4B according to manufacturer's instructions. For trypsin immobilization, benzamidine was added to the buffer solution at a 1:1 benzamidine/trypsin molar ratio.

Effect of pH and Ionic Strength on Aprotinin Adsorption and Desorption onto Biospecific Matrix

Aprotinin adsorption and desorption experiments were carried out according to the factorial experimental design method presented by Barros Neto et al. (1995) (Table ¹). For the adsorption assays 40 mg of swine trypsin immobilized resin (oxirane activated) or 20 mg of bovine chymotrypsin immobilized resin (CNBr activated) were suspended in 1.0 ml of a 33 µg/ml or a 56 µg/ml aprotinin solution (for trypsin resin or chymotrypsin resin, respectively) in 1.5 ml Eppendorf tubes. These aprotinin solutions were prepared in 0.05 M tris-HCl buffer containing 0.01 M CaCl₂, at different pH values (from 7.3 to 8.7) and NaCl molarities (from 0.018 M to 0.582 M) as shown in Table ¹. These ranges were set based on the literature (Barthel and Kula, 1993; Fiorucci et al., 1995; Kassel, 1970) and preliminary experiments. The flasks were agitated at 25°C for 24 h and the aprotinin concentrations in the supernatants were determined via trypsin inhibition.

*µ , ionic strength. x1 and x2, codified variables (pH and ionic strength, respectively) according to the following equations: x1 = (pH - 8.0)/ 0.5 and x2 = (µ - 0.300)/ 0.200, where µ is given in terms of NaCl molar concentration.

For the desorption studies, the resins were first saturated with aprotinin in large single batches using 50 ml Erlenmeyers where the resins were suspended in 33 µg/ml or 56 µg/ml of aprotinin solutions (for trypsin resin or chymotrypsin resin, respectively) prepared in 0.05 M tris-HCl buffer containing 0.01 M CaCl₂, pH 8.0. The suspensions were agitated at 25°C for 24 h and then washed extensively with 5 mM tris-HCl buffer with 0.01 M CaCl₂, pH 8.0. The desorption assays were carried out dividing the aprotinin-saturated resin (40 and 20 mg of swine trypsin and bovine chymotrypsin immobilized resins, respectively) in 1.0 ml of 0.02 M glycine buffer at different pH values (from 2.1 to 3.5) and NaCl molarities (from 0.018 M to 0.582 M). The flasks were agitated at 25°C for 24 h and centrifuged. Aliquots of the supernatants were collected, and the pH values of these solutions were adjusted to 8.0 by adding small volumes (25% of the supernatant volumes) of 0.5 M tris-HCl buffer, pH 8.2. The aprotinin concentrations in these solutions were determined via trypsin inhibition. Statistical analysis of the experimental data was carried out using the Modreg software (software that comes with the Barros Neto et al. 1995 publication) and Statistica (Statsoft, USA). Linear or quadratic polinomial models were utilized to fit the experimental data.

Loading the IMAC Matrix

The PCI matrix was loaded with Cu²⁺ by mixing it with 20 mM copper sulfate (5 volumes) after a washing with water (10 volumes). Metal loading was followed by washing with 100 mM sodium acetate, pH 4.0, NaCl 1.0 M (10 volumes) and 20 mM sodium phosphate, pH 7.0, 1.0 M NaCl (10 volumes).

Adsorption Isotherm onto IMAC Matrix

20 mg of PCI-Cu²⁺ (dry matrix basis) were added to 1.0 ml of aprotinin-trypsin complex solutions (1.0 to 20.0 mg/ml in 20 mM sodium phosphate) inside 3 ml syringes with a filter at their tips. After 1 hour at 25 ºC under agitation, the liquid phase was collected by ejection. This adsorption step was followed by five washing steps with 1.0 ml of adsorption buffer. Next, the adsorbed protein was eluted by mixing the matrix with two volumes (1.0 ml each time) of 100 mM sodium acetate, pH 4.0, 1.0 M NaCl. Complex concentration was determined by absorbance at 280 nm.

Studies of Complex Desorption from IMAC Matrix

For the desorption studies, the PCI-Cu²⁺ matrix was first loaded with aprotinin-trypsin complex in large batches where the matrix was suspended in 10 mg/ml of complex solution prepared in 20 mM phosphate buffer, pH 7.0, 1.0 M NaCl. The suspension was agitated at 25 °C for 1 hour and then washed extensively with the adsorption buffer. The desorption assays were carried out by dividing the complex loaded matrix between syringes (20 mg of wet matrix) and adding to each one 1 ml of a specific desorption buffer (buffers with different pH values with or without 1.0 M NaCl). After incubation at 25 °C for 1 hour, the solution was ejected, fresh desorption solution was added and the process was repeated. The buffers used for desorption were 20 mM phosphate buffer for pH 6.0 and 5.0, 100 mM sodium acetate for pH 4.0 and 3.5; and 20 mM glycine for pH 2.1. A final desorption using EDTA (two volumes of 1 ml of 50 mM solution at pH 8.0) was carried out with a similar procedure.

RESULTS AND DISCUSSION

Effect of pH and Ionic Strength on Aprotinin Adsorption and Desorption onto Biospecific Matrix

The dependence of aprotinin adsorption or desorption on pH and ionic strength were displayed in three-dimensional graphs (Figures ¹ and ²). Adsorption of aprotinin on immobilized trypsin was favored by decreasing the ionic strength (Figure ¹). An increase of 40% in binding capacity was verified by decreasing the ionic strength from 0.582 M to 0.018 M. This was not the case for aprotinin adsorption on immobilized chymotrypsin, where variation in ionic strength had little influence. These results could be explained by the mechanism suggested by Vincent and Lazdunski (1972) for aprotinin-trypsin association, where the primary event is recognition, by ion-pair formation (salt bridge), of the a-amino group of the Lys₁₅ of the aprotinin by the a-carboxilate of Asp₁₇₇ of the trypsin. The fact that adsorption of aprotinin on immobilized trypsin was disfavored when ionic strength was increased could be due to ionic interference in ion-pair formation. This does not occur in aprotinin-chymotrypsin interaction, where the Asp₁₇₇ is substituted by a Ser₁₇₇ that does not have a charged group.

Figure 1: The effect of pH and ionic strength (µ) on aprotinin adsorption. Adsorbent: (a) immobilized swine trypsin (oxirane-activated agarose) and (b) immobilized bovine chymotrypsin (CNBr-activated agarose). The R² obtained were 0.963 and 0.845, respectively. Q is the amount of aprotinin adsorbed per gram of resin.

Figure 2: The effect of pH and ionic strength (µ) on aprotinin desorption. Adsorbent: (a) immobilized swine trypsin (oxirane-activated agarose) and (b) immobilized bovine chymotrypsin (CNBr-activated agarose). The R² obtained were 0.995 and 0.953, respectively. D is the amount of aprotinin desorbed per gram of resin.

The adsorption data also showed that pH has a small effect on the trypsin-aprotinin association, specially for high values of ionic strength; the maximum difference in adsorption capacity observed was an increase of 13% when the pH was increased from 7.5 to 8.5 at an ionic strength of 0.1 M. This effect was not verified for the chymotrypsin-aprotinin association where the binding capacity was not significantly affected by pH variations in the range studied. Immobilized trypsin had the highest aprotinin adsorption capacity: 12.6 mg/g of wet matrix.

Desorption of aprotinin from immobilized trypsin or chymotrypsin resins was strongly influenced by pH, with higher desorption at lower pH values (Figure ²). The difference between aprotinin-protease affinities could be evidenced by the difference between the pH values where almost 100% of the aprotinin was desorbed: 2.1 for the stronger aprotinin-trypsin association and 2.8 for the aprotinin-chymotrypsin association (data not shown). The influence of ionic strength on aprotinin desorption was not significant for both immobilized enzymes.

The best conditions for aprotinin adsorption and desorption on immobilized trypsin obtained from this study were adsorption at 0.018 M NaCl and pH 8.7 and desorption at 0.018 M NaCl and pH 2.1. For immobilized chymotrypsin, the best conditions found were adsorption at 0.582 M NaCl and pH 8.0 and desorption at 0.582 M NaCl and pH 2.1.

Effect of pH and Ionic Strength on Complex Adsorption onto IMAC Matrix

The isotherm for complex adsorption was determined at 25 °C in the 20 mM phosphate buffer, pH 7.0 and 1.0 M NaCl. In order to investigate the effect of pH, we maintained the salt concentration constant and changed the pH to 8.0 and 8.5, whereas to study the effect of ionic strength, we maintained the pH at 7.0 and changed the salt concentration to 0.50 and 0.04 M (Figure ³). The changes in pH did not have a significant effect on complex adsorption capacity. However, the data indicated a strong dependence of complex adsorption capacity on ionic strength, with higher capacities obtained at a lower salt concentration. The adsorption capacity of aprotinin-trypsin complex was as high as 18.0 mg/g of dry matrix (20 mg/g of wet matrix).

Figure 3: Effect of pH and ionic strength on the aprotinin-trypsin complex adsorption on PC-I-Cu²⁺ at 25 °C in 20 mM sodium phosphate. Adsorption capacities at 1.0 M NaCl, pH 7.0 (n), 8.0 (¡) and 8.5 (D) and at pH 7.0 with 0.5 M (o ) and 0.04 M (Ñ) of NaCl. Capacities determined by the eluted protein.

Effect of pH and Ionic Strength on Complex Desorption from IMAC Matrix

Desorption experiments indicated a strong effect of pH and ionic strength on aprotinin-trypsin complex recovery (Figure ⁴). In both desorption conditions regarding ionic strength (presence and absence of 1.0 M NaCl), the complex recovery increased with protonation. However, this effect was stronger for the case where NaCl was added to the desorption buffer: virtually all the complex was desorbed at pH 2.1 with buffer containing NaCl, whereas only 40% was desorbed with buffer when NaCl was not added. After the desorption step, the matrix was regenerated by washing it with 50 mM EDTA, pH 8.0. This procedure promotes metal stripping from the matrix, and then all the remaining protein adsorbed by interaction with the metal is also removed (Figure ⁴). Although washing with EDTA allowed complete desorption, it is not usually applied for protein elution, particularly in large-scale processing, due to the inclusion of metal species in the system and the fact that it removes the ligant from the matrix.

Figure 4: Complex desorption in the absence (a) and in the presence (b) of 1.0 M NaCl. Hatched area: desorption by protonation (pH change); buffer solutions: 20 mM sodium phosphate (pH 6.0 and 5.0), 100 mM sodium acetate (pH 4.0 and 3.5), and 20 mM glycine (pH 2.1). Blank area: desorption with 50 mM EDTA, pH 8.0 after the desorption at low pH.

CONCLUSION

Two adsorption methods were studied as potential methods for the recovery of aprotinin. Both methods, biospecific and pseudobiospecific adsorption, were affected by changes in pH and ionic strength. The capacities of both systems were of the same magnitude. Therefore, in order to continue process development, both systems should be evaluated for selectivity when fed with aprotinin containing fermentation broth or tissue extract.

ACKNOWLEDGEMENTS

The authors are grateful to FAPESP, Brazil for its financial support and to Biobrás S.A., Brazil for its donation of enzymes. We thank Dr. Cesar Costapinto Santana (UNICAMP, Brazil) in whose laboratories this work was developed; Dr. Paulo de Tarso Vieira e Rosa (UNICAMP, Brazil) for advice and stimulating discussions; and Dr. Zivko Nikolov (Iowa State University, USA) for the donation of aprotinin. We also thank Dr. Sônia Maria Alves Bueno (UNICAMP, Brazil) for reviewing the paper.

• Barthel, T. and Kula, M., Rapid Purification of DesPro(2)-Val15-Leu17-Aprotinin from Culture Broth of a Recombinant Saccharomyces cerevisia Biotech. Bioeng. 42, 1331-1336 (1993).
• Baufreton C., Velthuis, H. T., Jansen, P. G. M., Wildevuur, C. H. R. and Loisance, D. Y., Reduction of Blood Activation in Patients Receiving Aprotinin During Cardiopulmonary Bypass for Coronary Artery Surgery. ASAIO Journal 42m, M417-M423 (1996).
• Barros Neto, B.; Bruns, R. E. and Scarminio, I. S., Planejamento e Otimizaçăo de Experimentos. Campinas: Editora da Unicamp (1995).
• Burck, P. J., Pancreatic Secretory Trypsin Inhibitors. Meth. Enzym. 19, 906-914 (1970).
• Fiorucci, L.; Erba, F.; Falasca, L.; Dini, L. and Ascoli, F., Localization and Interaction of Bovine Pancreatic Trypsin Inhibitor and Tryptase in Granule of Bovine Mast Cells. Biochem. Biophys. Acta. 1243, 407-413 (1995).
• Hermanson, G. T., Mallia, A. K. and Smith, P. K., Immobilized Affinity Ligand Techniques, pp. 118-119. San Diego: Academic Press (1992).
• Kassel, B., Bovine Trypsin-Kallikrein Inhibitor (Kunitz Inhibitor, Basic Pancreatic Trypsin Inhibitor, Polyvalent Inhibitor from Bovine Organs). Meth. Enz. 19, 844-852 (1970).
• Vincent, J. P. and Lazdunski, M., Trypsin-Pancreatic Trypsin Inhibitor Association. Dynamics of the Interaction and Role of Disulfide Bridges. Biochem. 11, 2967-2977 (1972).

Table 1: Experimental design for the study of the effects of pH and ionic strength on aprotinin adsorption and desorption (a 2² plus star, central composite design).

* To whom correspondence should be adderessed

To whom correspondence should be adderessed

Publication Dates

• Publication in this collection
  15 Sept 1999
• Date of issue
  June 1999

History

• Received
  19 Jan 1999
• Accepted
  23 Mar 1999