Effects on Bovine Serum Albumin and Porcine Insulin Immobilization

of Internal Medicine, School of Medicine, Federal University of Minas Gerais, Rua Espirito Santo,. 35/2 andar, Centro, Belo Horizonte, MG, Brazil. Rec...
8 downloads 0 Views 194KB Size
Download PDF
Biomacromolecules 2000, 1, 789-797

789

Surface Functionalization of Porous Glass Networks: Effects on Bovine Serum Albumin and Porcine Insulin Immobilization Herman S. Mansur,*,† Ze´ lia P. Lobato,‡ Rodrigo L. Ore´ fice,† Wander L. Vasconcelos,† Cintia Oliveira,† and Lucas J. Machado§ Department of Metallurgy and Materials Engineering, Department of Veterinary Medicine, and Department of Internal Medicine, School of Medicine, Federal University of Minas Gerais, Rua Espirito Santo, 35/2 andar, Centro, Belo Horizonte, MG, Brazil Received September 29, 2000

Biomolecules can be immobilized in many different ways. They can also be entrapped or tightly adsorbed within porous gels, clays, membranes, resins, and several other materials, but it is crucial that they retain their active conformation after the incorporation procedure. Porous gel matrixes with functionalized surfaces offer unlimited possibilities to control the protein-substrate interaction behavior. In the present work, we have studied the adsorption and the relative stability of bovine serum albumin (BSA) and porcine insulin (PI) incorporated in gels of SiO2 glass matrixes. The porous gel matrixes were obtained using tetramethoxysilane (TMOS)/methanol and functionalized with (3-mercaptopropyl)trimethoxysilane and (3aminopropyl)triethoxysilane. The relative adsorption kinetics and stability of BSA and PI incorporated in glass networks were evaluated by immersion in phosphate buffer saline (PBS) and alkaline elution media for different periods of time. The kinetics of protein release from the gel matrix was monitored by UVvisible spectroscopy. A significantly larger PI release was observed compared to BSA in PBS solutions. We believe this is mainly associated with the difference on protein interactions with the modified surface, according to the characterization results of porosity, surface area, and contact angle of different functionalized gel matrixes. We could not observe any evidence of denaturation with either proteins after their desorption from gel matrixes using the ultraviolet spectroscopy technique. These results have also been confirmed with the strong bioactivity response from “in vivo” tests conducted in rats, where porous gels with PI incorporated were implanted, showing that released proteins retained their native conformation. 1. Introduction The adsorption of protein from solution onto solid surfaces is a fascinating and complex process playing a major role in biological systems. Surface-immobilized proteins have drawn the attention of the research community in the last 2 decades. The high efficiency presented by biological macromolecules in selecting chemical species has motivated the development of devices that combine synthetic materials with biological entities. Proteins can de immobilized in many different ways, but it is crucial that they retain their active conformation after the incorporation procedure.1,2 There are three major methods for immobilizing biomolecules and cells. Two of them are physically based, physical adsorption and physical entrapment. The third method is based on covalent (chemical) attachment.2 Thus, it is important to note that the term immobilization can refer either to a temporary or to a permanent localization of the biomolecule on or within a support. Biological macromolecules, such as proteins, have very particular chain configurations and conformations that * To whom correspondence should be addressed. Tel: +55 31 2381843. Fax: +55 31 238-1815. E-mail: [email protected]. † Department of Metallurgy and Materials Engineering, Federal University of Minas Gerais, Brazil. ‡ Department of Veterinary Medicine, Federal University of Minas Gerais, Brazil. § Department of Internal Medicine, School of Medicine, UFMG, Brazil.

promote high levels of specificity during chemical interactions. The immobilization of proteins onto surface-functionalized substrates has been one of the most promising areas in the bioengineering field as discussed by several authors.3-8 The performance of artificial materials in contact with biological systems is to a large extend determined by surface interactions. Therefore considerable efforts have been made in the last years to improve surface compatibility of materials used for biological and biomedical applications.9 It has an enormous potential for application as biomaterial implants, immunological kits, drug delivery systems, and biosensors. In the present work, we aimed to go one step further in the analysis of this phenomenon of protein adsorption/ desorption. We have studied the incorporation of bovine serum albumin (BSA) and porcine insulin (PI) into porous gel networks of SiO2 and into gels of SiO2 with functionalized surface. We chose to study the adsorption of bovine serum albumin because albumin is dominantly present in body fluids among other proteins. Insulin was chosen because of its activity of regulating the glucose concentration level found in the blood of most mammals. The relative chemical stability of BSA and PI to remain incorporated into the SiO2 glass matrixes was evaluated using UV-vis spectroscopy. “In vitro” tests were used to characterize the proteinadsorption process onto the porous glass matrixes followed

10.1021/bm0056198 CCC: $19.00 © 2000 American Chemical Society Published on Web 11/03/2000

790

Biomacromolecules, Vol. 1, No. 4, 2000

Figure 1. Schematic representation of functionalized gel networks with (a) (3-aminopropyl)triethoxysilane (APTES) and (b) (3-mercaptopropyl)trimethoxysilane (MPTMS).

by the kinetics of the protein-desorption process in phosphatebuffered saline (PBS) and alkaline elution media. “In vivo” experiments were conducted to evaluate the bioactivity of the released proteins from the SiO2 glass matrixes. 2. Experimental Procedure Tetramethoxysilane Si(OCH3)4 (TMOS, >98%), (3-aminopropyl)triethoxysilane (APTES), (3-mercaptopropyl)trimethoxysilane (MPTMS), and bovine serum albumin (BSA, fraction V, >99.5%, Mr ) 67 000) were supplied by SigmaAldrich. Porcine insulin (PI, >99.5%, Mr ) 5778) was provided by Biobras SA. PBS solution (phosphate-buffered solution) was prepared using the reagents Na2HPO4 (>99.0%), NaH2PO4 (>99.0%), Na2CO3 (>99.5%), and NaCl (>99.0%). 2.1. Porous Matrix. A porous glass matrix was obtained using TMOS in methanol and PBS solution with pH ) 7.40 ( 0.05. The gel surface was chemically patterned by adding the silane reagents MPTMS and APTES during the process of porous glass network formation. The schematic representation of the two different functionalized structures obtained for gel networks is shown in Figure 1. The sols were cast into 96-well-plate molds, where gelation occurred. After thermal treatment of aging and drying, the gels of silica were immersed in PBS solution containing PI and BSA, to reach a final protein concentration of 1.0 wt % in SiO2. The impregnated gels were subsequently dried for 48 h/40 °C. These glass disks were produced with an average weight of 12 ( 2 mg. All glass disks were accurately weighed before adsorption and desorption experiments, needed for mass

Mansur et al.

balance calculations to quantify the concentration of BSA and PI immobilized into the porous glass matrixes. Dried gels were characterized using mercury picnometry, surface area and average porous size were estimated by N2 adsortion using the BET method. 2.2. FTIR Spectroscopy. Fourier transform infrared spectroscopy (FTIR) was used to characterize the presence of specific chemical groups in the materials (Perkin-Elmer, Paragon 1000). Gels were milled and mixed with dried KBr powder. FTIR spectra were obtained within the range between 4000 and 400 cm-1 during 64 scans. The incorporation of protein within the gels was also monitored by FTIR spectroscopy. We would like to point out that FTIR spectra were used as a qualitative reference of protein incorporation into the gel matrix. FTIR was also used to characterize the presence of specific chemical groups in the gel network, reflecting the effectiveness of the developed procedure for functionalization of porous glass. 2.3. Protein Release Procedure. Protein release study was performed by incubating gel matrixes with BSA and PI incorporated in PBS and alkaline (8 < pH < 10) media. The SiO2 gels with protein were weighed and then immersed in 1.0 mL of PBS medium for 2 and 24 h at 37 °C. After that, PBS solution was collected for absorbance measurement and protein concentration calculation. An aliquot of 1.0 mL of alkaline medium was then added to these same gel matrixes previously treated with PBS. After 30 min in alkaline solution at room temperature, duplicate samples were collected and absorbance readings were used for BSA and PI desorption evaluation. We have also crushed the SiO2 gels, with BSA and PI immobilized, until reaching a fine powder. The powder samples were weighed and transferred to a plastic tube. PBS solution (1.0 mL) was added to each tube and the mixture was allowed to rest for a 24 h period, at 37 °C. Tubes were centrifuged at 1350g using an International refrigerated centrifuge (model PR-2), and the supernatant was carefully collected, filtered using a 0.45-µm membrane, and stored for latter protein measurement. Then, 1.0 mL of alkaline medium was added to each tube, with the remaining gel powder. After 30 min, at room temperature, each tube was centrifuged, and supernatant was collected, filtered in 0.45-µm paper, and stored for protein measurement as described above. 2.4. Protein Release Kinetics Studies, “in Vitro”. The kinetics of protein desorption from the gel matrix was monitored by UV-vis spectroscopy, with the concentration estimated through Lowry method readings at 660 nm.10 Glass matrixes incorporated with protein were weighed and incubated with 1.0 mL of PBS medium at 37 °C. After 10 min of immersion, an aliquot of approximately 1.0 mL of PBS solution was collected from each tube for protein concentration estimation, and 1.0 mL of fresh PBS medium was then added to the same matrix. This procedure was again repeated for 30, 60, 120, 240, 360, 640, and 1440 min of incubation. 2.5. Colorimetric Determination of Protein Concentration (Lowry Method). This study employed a colorimetric method to determine protein concentrations, known as the Lowry assay.10 In this method, a protein solution, which was

Surface Functionalization of Porous Glass

treated with alkaline copper sulfate, was subsequently treated with another reagent called Folin-Ciocalteau phenol reagent. The former reagent reacts with the nitrogen of the peptide bond, resulting in a faint pink-purple color. The latter reagent intensifies the color by reacting with tyrosine and tryptophan residues in the polypeptide. The addition of these reagents to solutions that have undergone serial dilutions results in a blue-green complex. The intensity of the color complex is directly proportional to the protein concentration of the sample due to the higher number of tryptophan and tyrosine residues. The absorbances of both BSA and PI solutions were measured using a Shimadzu UV-160A spectrophotometer set at 660 nm. 2.6. Spectroscopic Study of Protein Native Conformation. We have used the solvent perturbation spectroscopy technique11 (SPS) to verify denaturation effects of the developed process for incorporation of BSA and PI into the gels of silica and silica with a functionalized surface. The SPS analysis was carried out through the UV spectra shift (blue shift) obtained in the range of 250-300 nm when 4 M guanidine hydrochloride (Gd-HCl) was added to BSA and PI solutions. We have used the absorption differences between folded (native) and unfolded (denaturated) states of proteins. The effectiveness of the developed method for incorporation of proteins in porous gel glasses was evaluated by comparing pure gels of silica with gels with a chemically functionalized surface. 2.7. Subcutaneous Bioactivity of Insulin-Immobilized Gels, “in Vivo”. Since the immobilization process may denature or otherwise inactivate the incorporated insulin, the bioactivity was investigated. Male rats were fasted overnight, anesthetized using ethyl ether reflux, and weighed, and 0.4mL blood samples were collected by intravenous catheter inserted in the jugular vein. Subcutaneous implants with insulin incorporated of either pure silica gel disks or surfacemodified gels were placed under the back skin of each rat. The rat was denied food, and blood samples were collected at intervals of 20, 40, 60, 90, 120, and 150 min postimplanting. The heparinized blood samples were centrifuged at 7000g for 10 min, and plasma was removed and saved for glucose monitoring. 2.8. Glucose Analysis. Plasma glucose levels were obtained by using the glucose GOD-ANA colorimetric assay (Labtest Diagnostica S.A., Minas Gerais, Brazil), following the manufacturer’s protocols. 2.9. Hydrophobic/Hydrophilic Behavior via Contact Angle Measurement. The influence of surface functionalization on the hydrophilic/hydrophobic behavior of glass substrate was estimated via contact angle measurements. We have evaluated the average contact angle of Milli-Q water (18.0 MΩ) spread over hydroxy-terminated, amine-terminated, and thiol-terminated gels. 2.10. SEM (Scanning Electron Microscopy)/EDX (Energy-Dispersive X-ray Spectroscopy) Microanalysis. The influence of functionalization of substrate on the adsorption behavior of BSA and PI was observed by SEM/EDX microanalysis. SEM micrographs were obtained using a JEOL JSM 5410. Gel samples were coated with Au/Pd films

Biomacromolecules, Vol. 1, No. 4, 2000 791

Figure 2. Photograph of SiO2 porous glass disks produced via the sol-gel route.

to enhance resolution and stabilize the electron beam mainly on protein-rich areas. 3. Results We have obtained the SiO2 porous glass disks via a solgel route. SiO2 glasses and SiO2-functionalized glasses were cast in 96-well-plate molds, where gelation occurred. Figure 2 shows a photo of a few of the glass disks produced. All samples of porous glass disks presented no major difference in apperance. They were found to be optically transparent to visible light and mechanically stable. We did not observe any significant effect on the porosity of the chemically functionalized gel matrix when compared to pure silica gels. The average porous diameter was found to be in the range of 1-2 nm. Surface area results obtained through N2 adsorption using the BET model showed an average value of 500 ( 100 g/m2 for the analyzed samples. Contact angle measurements indicated a strong difference on the hydrophilic behavior of the pure silica gel compared to surface-modified gels. We obtained average contact angle values of 29°, 59°, and 71° for pure silica gel, amineterminated gel, and thiol-terminated gel, respectively (Figure 3). We have used the FTIR technique to characterize the protein incorporation in porous sol-gel matrixes. The spectrum of pure silica gels (Figure 4a(1)) clearly reveals the peaks due to Si-O-Si bonds (1080 and 450 cm-1) and Si-OH bonds (3500 and 950 cm-1). The peaks associated with amide-I (1620-1680 cm-1) and amide-II (ν ) 14801580 cm-1) were observed on the spectrum of pure bovine serum albumin used as reference (Figure 4a(3)). The FTIR spectrum in Figure 4a(2) shows the results of the porous silica gel network after protein (BSA) incorporation, where all major important amide stretching vibration bands are present. Therefore, we could confirm the immobilization of protein in gel network by using FTIR spectroscopy. In Figure 4b, FTIR spectra of silica gels modified with mercaptopropyltriethoxysilane (MPTMS) are shown. Again,

792

Biomacromolecules, Vol. 1, No. 4, 2000

Mansur et al.

Figure 3. Contact angle mesurements of Milli-Q water (18.0 MΩ) over pure SiO2 gel (a), -SH modified gel (b), and -NH2 modified gel (c).

Figure 4. (a) FTIR spectra of porous glass and porous glass with BSA incorporated: (1) pure silica gel glass without protein; (2) BSAimmobilized porous glass network; (3) reference spectrum of BSA. (b) FTIR spectra of porous glass and porous glass with PI incorporated: (1) reference spectrum of PI; (2) pure silica gels; (3) the silica gel modified with 1% MPTMS and 1% porcine insulin incorporated.

the spectrum of pure silica gels (Figure 4b(2)) reveals the peaks due to silicon bonds in the glass network. The spectrum in Figure 4b(3) is associated with the silica gel modified with 1% molar MPTMS and 1% porcine insulin (PI) incorporated. This spectrum (Figure 4b(3)) shows peaks at 2550 cm-1 that are characteristic of the MPTMS and peaks at 1650-1680 cm-1 associated to porcine insulin amide-I vibration band. It can also be verified the presence of the -CH stretching vibration bands (2850-3000 cm-1) mainly associated with the propyl group introduced by the func-

tionalization of the gel surface. The spectrum of pure PI used as reference can be observed in Figure 4b(1). SiO2 gels and chemically functionalized SiO2 gels were analyzed by SEM/EDX. SEM micrographs of the surface of a pure silica gel (Figure 5a), MPTMS functionalized gel (Figure 5b), and APTES functionalized gel (Figure 5c) that were previously impregnated with porcine insulin (PI) are exhibited in Figure 5. These micrographs revealed that the morphology and distribution of protein agglomerates on the surface of the gels are drastically affected by the type of chemical functionality introduced into the samples. Also noted was a “dewetting”, characteristic of the surface of the functionalized gel, usually forming droplets of PI.13,14 3.1. Kinetics of the Protein Desorption Study. Surfacefunctionalized glass with APTES and MPTMS presented significantly less BSA released in PBS when compared to the pure gel of silica (Figure 7a). On the contrary, gelfunctionalized substrates with APTES and MPTMS have clearly showed an increase of PI release in PBS media when compared to pure gel of silica (Figure 6a). As we increased the molar fraction of amino-terminated groups (R-NH2) from 1% to 3%, we could observe that BSA release in PBS was reduced over 50% (Figure 7a). An opposite trend was verified for thiol-terminated groups (RSH), where we have noticed a significant increase on BSA release when the MPTMS molar fraction was raised from 1% to 3% and then to 10% (Figure 6a). It was also observed that BSA was incorporated and kept immobilized onto the gel powder after crushing with a pestle when compared to the original gel (Figure 7a solid and dotted bars). The gel of silica immersed in PBS for 24 h was later incubated in alkaline medium. Protein release (PI and BSA) was observed from all three silica substrates, pure glass, thiolfunctionalized glass, and amino-functionalized glass (Figure 6 and Figure 7). We have also observed that BSA concentration released in alkaline solution was larger than the amount of PI released (Figure 6b and Figure 7b). The kinetics of protein released from the gel matrix was monitored by UV-vis spectroscopy, with the concentration estimated through the Lowry method at 660 nm. The results revealed a different behavior of adsorption and stability of

Surface Functionalization of Porous Glass

Biomacromolecules, Vol. 1, No. 4, 2000 793

Figure 5. SEM micrographs of the surface of (a) silica gels, (b) (mercaptopropyl)triethoxysilane (MPTMS)-modified silica gels impregnated with porcine insulin, and (c) (3-aminopropyl)triethoxysilane (APTES)-modified silica gels impregnated with porcine insulin.

Figure 6. Histogram of porcine insulin release: (a) PBS solution (solid bar) 2 h immersion, (dashed bar) 24 h immersion, (dotted bar) crushed powder, 24 h immersion; (b) immersion in alkaline solution (solid bar) after pretreated for 2 h in PBS solution, (dashed bar) after pretreated for 24 h in PBS solution, (dotted bar) crushed powder after pretreated for 24 h in PBS solution.

BSA and PI proteins incorporated into the porous glasses. A larger PI release was observed compared to BSA release in PBS solutions. The median time to release 50% of the incorporated protein (T50) is shown in Table 1. Glass matrixes incorporated with BSA have presented a minimum of 30 min incubation period before any significant detection of protein release. On the other hand, the PI delivery study has indicated almost immediate protein release (less than 10 min). Glasses with thiol-functionalized surfaces (R-SH) have presented stronger PI adsorption compared to aminoterminated groups (R-NH2). An opposite trend was verified with a thiol-functionalized surface that presented less adsorption with BSA than the amino-patterned gel of SiO2.

Figure 7. Histogram of bovine serum albumin release: (a) PBS solution (solid bar) 2 h immersion, (dashed bar) 24 h immersion, (dotted bar) crushed powder, 24 h immersion; (b) immersion in alkaline solution (solid bar) after pretreated for 2 h in PBS solution, (dashed bar) after pretreated for 24 h in PBS solution, (dotted bar) crushed powder after pretreated for 24 h in PBS solution.

The kinetics study of PI release in PBS solution with different functionalized glass substrates compared to pure silica gel is shown in Figure 8. Figure 8a shows the curve associated with PI release from a pure SiO2 matrix; Figure 8b represents the amine-terminated glass release curve, and Figure 4c indicates the PI release profile from thiolterminated glass. Figure 9 shows the accumulated percentage of PI release in PBS solution with different functionalized glass substrates. Again, PI presented stronger immobilization behavior with thiol chemically patterned glass than with SiO2 aminoterminated glass.

794

Biomacromolecules, Vol. 1, No. 4, 2000

Mansur et al.

Figure 8. Kinetics of PI release in PBS solution with different functionalized glass substrates: (O) SiO2; (9) R-NH2 terminated/APTES; (2) R-SH terminated/MPTMS. Inset detail: initial stage of protein release.

Figure 9. Accumulated percentage of PI release in PBS solution with different functionalized glass substrates: (b) SiO2; (0) R-SH terminated/ MPTMS; (2) R-NH2 terminated/APTES. Inset detail: initial stage of protein release with T50 evaluation. Table 1. Average Release Time of Bovine Serum Albumin (BSA) and Porcine Insulin (PI) in Three Different Porous Gel Substrates: SiO2; SiO2 Chemically Functionalized with APTES (Amino Terminated); SiO2 Chemically Functionalized with MPTMS (Thiol Terminated) median release time/min (T50) substrate

PI

SiO2 SiO2-APTES (R-NH2) SiO2-MPTMS (R-SH)

30