Application of monoliths as supports for affinity chromatography and fast enzymatic conversion

J. Biochem. Biophys. Methods 49 Ž2001. 153–174 www.elsevier.comrlocaterjbbm

Review

Application of monoliths as supports for affinity chromatography and fast enzymatic conversion Djuro Josic´ ) , Andrea Buchacher Research and DeÕelopment, Octapharma Pharmazeutika Produktions Ges.m.b.H., Oberlaaerstraße 235, A-1100 Vienna, Austria

Abstract Monoliths are useful chromatographic supports, as their structure allows improved mass transport. This results in fast separation. Once the ligand of interest has been immobilized, chromatographic separation can also be accomplished in affinity mode. Ligands with low molecular mass have been shown to be the easiest to immobilize. Nowadays, ligands with low molecular mass are often designed by combinatorial chemical techniques. In addition, many applications have been described where ligands with high molecular mass, such as Proteins A and G, antibodies, lectins and receptors are used. The immobilization of an enzyme on the monolithic support creates a flow-through reactor. Small proteins, such as carbonic anhydrase, can be directly immobilized on the support. However, in the case of large molecules, the active center of the enzyme is no longer accessible at all or only to a limited degree. An improvement can be achieved by introducing a spacer, which allows maximum enzymatic conversion. Fast conversion of substrates with high molecular mass has been investigated with immobilized trypsin. It was shown that in case of high-molecular-mass substrates, the conversion rate depends very much on the flow-rate. Most applications described have been performed on an analytical or semi-preparative scale. However, the technical problems of up-scaling are close to being definitely solved, enabling enzymatic conversion on a preparative scale in the future. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Monolithic supports; Membranes; Affinity chromatography; Enzymatic conversion

1. Introduction Monolithic supports can be regarded as a further development of the application of membranes in chromatography. The capacity of single membranes is limited by the )

Corresponding author. Fax: q43-1-61032-295. E-mail address: [email protected] ŽD. Josic´..

0165-022Xr01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 2 X Ž 0 1 . 0 0 1 9 5 - 6

154

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

thickness of their layers. If membranes with large surfaces are used, major technical difficulties arise, especially with regard to sample distribution in the separation unit w1x. The problem is solved in different ways, in most cases by stacking w1,2x or rolling w3x of the membranes. Tennikova et al. w4x as well as Hjerten ´ et al. w5x have offered another solution by polymerization of different synthetic supports and by producing monolithic disks and columns. This approach has allowed the construction of chromatographic units with varying degrees of thickness, thin disks w2,6x as well as monolithic columns of various lengths w7x. Subsequently, Strancar et al. w8x have developed a cylinder-shaped monolithic column, in which the mobile phase pervades the stationary phase in a radial flow. This approach offers further scale-up of the system. The developments in the field of natural and synthetic polymers have been complemented by monolithic supports on the basis of silica gel w9,10x. The first successful experiments with membranes in chromatography have been carried out in affinity mode w3,11x. Also, the first sound data on mass transfer in such supports were obtained in systems with an affinity ligand, especially Protein A, and Immunoglobulin G as ligate w12x. Mass transfer was seen to be much faster than in the case of bulk supports. The experiments also showed that, in the case of membranes, there is practically no limitation of transport by diffusion w12,13x. The corresponding data with affinity ligands were shortly obtained afterwards for monolithic disks as well w14x. However, the initial experiments were carried out with a kind of hardware that was not quite suitable for routine use in the laboratory w12,14x. It took more time to design the first units with immobilized affinity ligands for use in the laboratory w2,15x, and the road to the first operational, commercially available monoliths with such ligands was even longer w16x. Subsequent experiments with affinity-chromatographic separations on membranes and monoliths have almost exclusively been carried out with low-molecular-mass ligands w16–20x. Of the ligands with high molecular mass, Proteins A and G were the first to be used for the purification of antibodies. Immobilized gelatin was used for the isolation of plasma fibronectin, and immobilized collagen was used for the isolation of calcium-binding proteins, the annexins w3,11,20–24x. The use of other high-molecularmass ligands presented some difficulties at the beginning, especially when membranes were used for their immobilization w25x. These problems can be solved by improved immobilization chemistry and by the introduction of spacers, similar to those used for bulk supports w26,27x. Now, it is possible to use to its full extent practically every ligand with low or high molecular mass for affinity-chromatographic separations on both membranes and monoliths w28x. Increasingly, the use of short peptides and other low-molecular-mass ligands becomes important w28–32x. They are designed by combinatorial chemistry for immobilization on monoliths and for purification of the corresponding target substances. They are the subjects of a separate section in this review. The so-called molecular imprinting represents, slightly simplified, a reverse of the method described above. The ligand is not immobilized on the surface, but instead, an artificial receptor is imprinted on the surface of the support w33x. Combining a molecular imprinting concept with a combinatorial chemistry strategy, combinatorial libraries of molecularly imprinted polymers could be prepared and screened w34x. Monolithic supports were regarded as well-suited for such applications at an early stage w35x. In a

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

155

way, this chromatographic method can also be considered affinity chromatography. So far, the use of such supports is still limited to the separation and detection of rather small molecules w34–36x. Shortly after the introduction of monolithic supports made of poly Žglycidyl methacrylate., the first successful conversions of synthetic substrates in flow-through mode were carried out with the enzyme carbonic anhydrase, immobilized on such a monolith. In this system initial, kinetic investigations were also carried out w14x. However, systematic experiments concerning a possible conversion of substrates by enzymes immobilized on a monolith have been carried out only recently w27,28,37,38x. It was shown that such reactors can be used on the smallest scale as well as on a large, preparative scale w39x. Recently, Klein w40x has published a review on the use of affinity membranes over the last 10 years. In his opinion also, commercialization remains meagre in the face of a large number of membrane preparation and process applications described in literature. The situation of the monolithic supports had been similar, and despite the recent progress, a final breakthrough has not been reached yet. Therefore, this review will concentrate on the use of monoliths. Membranes and technical solutions in the context of their use will only be included, if a comparison with monolithic supports appears to be useful. Furthermore, this review deals primarily with the use of supports, which are produced by block polymerization in the shape of a monolith. Apparently, neither monoliths based on silica gel nor monoliths based on polyacrylamide, the UNO-columns ŽBioRad, Munich, Germany., have so far been much used in affinity chromatography mode w23x. However, detailed reports exist on the use of membranes in this field w13,20,41,42x. This review therefore deals with the latest developments connected with the use of monoliths made of poly Žglycidyl methacrylate. in affinity chromatography, and with fast conversion of substrates through immobilized enzymes.

2. Affinity chromatography 2.1. Affinity chromatography with low-molecular-mass ligands The most frequently used ligands are dyes, inhibitors, co-enzymes and other substances with low molecular mass, which can subsequently enter into specific interactions with the components of the sample w13,14,17,18,20,40–44x. Fig. 1 shows an early application of affinity chromatography with low-molecular-mass ligands for the isolation of the enzyme carbonic anhydrase from human erythrocytes w14x. A monolithic disk with the immobilized inhibitor of carbonic anhydrase, p-Žamino methyl. benzol sulfonamide, was used. After binding at higher pH and washing-out of non-specifically bound components, the active enzyme is eluted at a pH of 5.5. This is a very elegant method, which exploits the fact that the interaction between the enzyme and the inhibitors is pH-dependent. At higher pH, the binding is stable. At pH below 6.0, it is increasingly unstable, and can result to dissociation. The protein is not denatured

156

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

Fig. 1. Purification of carbonic anhydrase from human erythrocytes by the use of a monolithic disc with immobilized p-Žaminomethyl. benozoyl sulfonamide. Reprinted from Ref. w14x with permission.

under these conditions, neither through its binding nor through its elution, and therefore, retains its enzymatic activity. In contrast to the widespread use of thin membranes in the field of immobilized affinity ligands with low molecular masses for the isolation of a number of proteins w13,17,45x, monoliths were rarely applied in this context in subsequent years. However, the latest results by Amatschek et al. w30x showed that ligands with low molecular

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

157

masses can also be used on monolithic supports. Hahn et al. w46x have shown that the choice of spacer and the structure of the matrix, along with immobilization chemistry, are decisive for the performance of monoliths with such ligands. The most interesting applications in the field of low-molecular-mass ligands at present are the following: immobilized-metal affinity chromatography ŽIMAC. on membranes for the isolation of recombinant proteins with a polyhistidine sequence at the end of the polypeptide chain w17x and the use of immobilized histidine for the removal of endotoxins from protein solutions w47x. A detailed review of these applications has been written by Tennikova and Freitag w42x. Young et al. w48x have carried out IMAC with monolithic supports of cellulose with grafted acrylic polymers. They purified catalase from bovine liver with a disc cartridge, which is similar to that described by Josic´ et al. w2x. They also showed that such monolithic supports allow enrichment comparable to that achieved with bulk supports but about three times faster. Over the last 3–4 years, the use of monoliths for affinity chromatography with low-molecular-mass ligands has concentrated on the application of ligands designed by combinatorial chemistry. They are the subjects of the next section. 2.2. Combinatorial ligands The widespread production and use of therapeutical recombinant proteins of high value requires adequate techniques for their isolation. Until well into the 1990s, affinity chromatography, mostly in the immunoaffinity mode, was the method of choice for the isolation of these proteins w49,50x. However, monoclonal as well as polyclonal antibodies of animal origin are rather sensitive molecules. Through hydrolysis of the matrix, they may contaminate the product. In contrast, synthetic affinity ligands, according to Huang et al. w51x and Fassina et al. w52x, are ideally suited for the purification of high-value therapeutic proteins, since they are inexpensive, chemically defined, non-toxic, contain no hydrolyzable bonds, resistant to chemical and biological degradation, and sterilizable and cleanable in situ. Plasma proteins, in particular, offer plenty of options for the use of such combinatorial synthetic ligands. Their design has been greatly assisted by increased access to structural data, computer-aided molecular design, and novel combinatorial chemical techniques w29,30,53x. This technology for the purification of proteins comprises four steps. The first step is the mapping of the binding site of a natural binding partner for linear epitopes w29x or the development of a peptide library, either biological w51x or synthetic w30x. Mixture-based combinatorial peptide libraries in various formats based on solid phase peptide synthesis w54x have been developed to an essential high-throughput discovery tool w55,56x. Deconvolution of the library can be performed using an iterative approach w57x or the less costly and time-consuming positional scanning approach w58x. In any case, a spacer and a support for ideal performance in affinity chromatography has to be found w29,30x. The last step would be the optimization of chromatographic conditions. Once the ligand is available, it is immobilized in a fourth step, and the appropriate separation conditions for the target molecule are optimized w29,30,53,57x. In this way, synthetic ligands for the human insulin precursor and for Immunoglobulin G were found. The proteins were subsequently purified by affinity chromatography w52,53x. However, if the three-dimensional structure of a protein is not yet known,

158

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

another strategy must be chosen. The easiest option is to find appropriate peptide sequences that can interact with the target protein. This approach has been used to design and synthesize a ligand that mimics a key dipeptide motif on Fragment B of Protein A, which is known to play a key role in the interaction with the Fc fragment of IgG w58x. Necina et al. w29x used the binding region of von Willebrand factor ŽvWF., a carrier protein for clotting Factor VIII ŽFVIII., to generate a panel of octapeptides. They identified four peptides with affinity to FVIII and used them for affinity chromatography of this protein. All the applications of immobilized combinatorial ligands listed here have been carried out with AconventialB bulk supports only. However, Amatschek et al. w30x have shown that combinatorial peptides are also suited for immobilization on monolithic supports. They have immobilized a number of newly designed combinatorial octapeptides with affinity to FVIII on different bulk supports and on monoliths. Compared with the bulk supports under investigation, the monolithic material showed much better performance, both in terms of capacity and selectivity. A characteristic separation is shown in Fig. 2. Platonova et al. w31x have successfully used synthetic pentadecapeptides and hexadecapeptides as well as the nonapeptide hormone bradykinin, all of them immobilized to monolithic, CIM-convective interaction media ŽBIA Separations., for the purification of monospecific polyclonal antibodies against corresponding antigens. The model investigations were to show that this technology can be used for fast fractionation of monospecific polyclonal antibodies in serum. This would recommend it for medical diagnostics and for the identification of pathogenic immunoglobulins. Experience with immobilized combinatorial ligands suggests so far that their performance is more sensitive with regard to immobilization chemistry, matrix composition and type of spacer than with regard to ligands with high molecular mass. This applies to bulk supports as well as monolithic supports. Hahn et al. w46x have experimented with the above-mentioned parameters such as immobilization chemistry, matrix composition and type of spacer, using a model peptide with affinity to chicken egg lysozyme. With this model ligand, they investigated representative matrices as supports in order to find an optimal ligand spacer support combination. Different spacer lengths and types were included in this investigation. The performance of each affinity sorbent was characterized by breakthrough curves. The model peptide immobilized on monolithic CIM disks had the highest capacity per micromole ligand. The highest capacity per volume was observed in the case of a tentacle gel with a 16-atom spacer. If the model peptide was immobilized on a tentacle gel without a spacer, the capacity of the support with affinity to lysozyme was considerably reduced. All the other bulk supports used in the experiments had much lower capacity and selectivity than the two supports mentioned above. The investigations indicate that it is exactly the immobilization of small molecules, where selectivity and capacity are linked to the optimal presentation of the ligand, which in turn depends on the characteristics of the chosen support. 2.3. Affinity chromatography with high-molecular-mass ligands As mentioned above, the separation by affinity chromatography on membranes with immobilized Protein A and gelatin was among the first successful applications of this

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

159

Fig. 2. ŽA. Affinity chromatography of plasma-derived clotting Factor VIII Žpd FVIII. by the use of selected candidates from a peptide library, immobilized on to CIM monoliths ŽBIA Separations, Ljubljana, Slovenia.. ŽB. SDS-PAGE. ŽC. Immunoblot with an anti-FVIII antibody mixture. ŽD. Immunoblot using anti-von Willebrand factor ŽvWF. antibody. Lane 1: sample; lane 2: eluate from a blank column; lane 3: eluate from a column with Peptide No. 12; lane 4: eluate from a column with Peptide No. 18; lane 5: eluate from a column with Peptide No. 24; lane 6: eluate from a column with Peptide No. 30; lane 7: eluate from a column with Peptide No. 35; lane 8: eluate from a column with Peptide No. 36; lane 9: recombinant vWF. Reprinted from Ref. w30x with permission.

160

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

technology for protein purification w28x. Subsequently, Mandaro et al. w3x have used rolled membranes with immobilized Protein A or G for the isolation of immunoglobulins. Brandt et al. w11x have carried out similar experiments with hollow-fiber membranes. Apart from Protein A, this group has also immobilized gelatin on the same type of membrane. The resulting units were subsequently used for the isolation of fibronectin from blood plasma. With this rather complex adhesion protein, the early experiments showed instantly a surprisingly high speed of separation and a high capacity of the membrane supports. Detailed investigations, which were carried out shortly afterwards, have shown that low capacity can be a problem, especially in the case of cellulose-based membranes. The reason is probably the low density of active groups on the surface of the support w25x. In the case of affinity chromatography with monolithic supports, which carry immobilized ligands with high molecular mass, the dynamic capacity of the separation unit has been observed to be influenced by the flow-rate. This may be explained by the rather slow interaction between large molecules w27x. In the course of the 12-year history of this type of support, membranes with immobilized ligands such as Proteins A and G w3,11,12,21,22,27x, antibodies w2,22,59x, receptors w59,60x, and other proteins with miscellaneous biological functions w3,22x have been used. This has been reported in detail, also in several reviews w13,20,28,38,40–42x. Josic´ et al. w27x have shown that monolithic disks, made of epoxy-activated poly Žglycidyl methacrylate., can be used for the immobilization of ligands with high molecular mass. The units with immobilized heparin and collagen were used for the separation of membrane proteins and annexins from the liver carcinoma, Morris hepatoma 7777. In the early years, monolithic disks were not used very often for the immobilization of high-molecular-mass ligands. The reason may have been that the disks were expected to separate biopolymers within seconds w15,16,61x. Affinity chromatography is rarely used for such fast, analytical separations. Consequently, among the first applications in this context, one is found with an analytical purpose. A disk with immobilized heparin was used for fast quality control of preparations of the plasma proteins Antithrombin III and clotting Factor IX w62x. The option of also using monolithic supports by means of scale-up for semi-preparative and preparative purposes w39x has opened the way for the application of affinity chromatography in this field. Fig. 3 shows the isolation of Protein G from a cell lysate of Escherichia coli by affinity chromatography on a CIM disk ŽBIA Separations. with immobilized Immunoglobulin G. With this application, Kasper et al. w63x have shown that direct immobilization of such large proteins and immunoglobulins on epoxy-activated disks is possible without the need for inserting a spacer between the support and the protein. The separations could also be carried out on a semi-preparative scale. Here again, the great advantage of affinity chromatography was shown in the isolation of target substances from highly diluted solutions. The special advantage inherent in monolithic supports is also obvious: the significant reduction in the time required for separation w64x. Josic´ et al. w27x have shown that monolithic supports, CIM disks ŽBIA Separations., can also be used for the immobilization of other ligands, such as lectins and various enzymes. A disk with the immobilized lectin concanavalin A ŽCon A. was used successfully for the purification of highly hydrophobic proteins, among them are proteins from plasma membranes of the liver.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

161

Fig. 3. Fast isolation of Protein G from diluted Ž1:5 and 1:10. cell lysate by the use of a monolithic disk with immobilized IgG. Reprinted from Ref. w31x with permission.

2.4. Isolation of antibodies and immunoaffinity chromatography For years, antibodies have been among the most important tools for the isolation and characterization of proteins. Purification of IgG with immobilized Proteins A and G has a long tradition. It is therefore hardly surprising that the first application of membrane chromatography was also carried out with immobilized Protein A. In this context, other uses of membranes and monoliths, made of various materials, such as cellulose, silica

162

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

gel and several synthetic polymers, were described rather early. All these supports had been supplied with active groups on their surfaces at different densities, on which Protein A or G could be immobilized with varying degrees of success. A review on their use was written by Langlotz and Kroner w21x. Later investigations revealed that some of the membranes and monoliths, especially those based on cellulose and silica gel, could not be recommended for such separations because of low capacity andror selectivity w25x. Some supports have disappeared from the market for this very reason. However, some membranes and monoliths with immobilized Proteins A and G have held their ground and are the materials of choice for very fast and effective separation as well as purification of monoclonal and polyclonal antibodies w21,25,31x. Within the framework of a very interesting application, Zhou et al. w65x prepared a membrane made of a composite of cellulose, grafted with acrylic polymers and formed by polymerizing a glycidyl methacrylate in the presence of dispersed cellulose fiber. A column with immobilized Protein A was packed with this membrane, which had previously been cut to pieces. Fig. 4 shows that the resulting column allowed the separation of IgG from human serum in less than 30 s. In an interesting study carried out by the same group, a similar monolithic support with immobilized Protein A was used as a model for immune adsorption therapy for hemoperfusion w66x. In order to facilitate the flow through the support, a tangential-flow unit was used. In this way, IgG and immune complexes were removed from blood and plasma. Experiments in vitro and in vivo with animal proteins confirmed that the Protein A monolith adsorbed mainly IgG and only small amounts of other plasma proteins. Such an extracorporeal circulation system proved to be safe and reliable. The isolation of antibodies by affinity chromatography with Protein A or G always carries a certain risk of denaturation, as elution occurs under rather harsh conditions. Fast separation on monoliths, achieved within seconds Žcf. Fig. 4., minimizes the risk. Another approach which has been described above, is the use of ligands from the combinatorial library. None of the bacterial proteins used so far is able to bind IgM specifically. For this antibody, as well as for IgA, the ligands produced by combinatorial chemistry appear to be an ideal solution w52,53,57x. The so-called multidimensional chromatography has been used specifically for the isolation of antibodies. As the single monolith exhibits low back pressure, several units with different ligands can be stacked in one cartridge. Using the method, Josic´ et al. w22x have isolated monospecific, polyclonal antibodies against the calcium-binding protein annexin VI. The antibodies that cross-react with other, similar proteins from the annexin group, were removed from the antiserum by passing it through the disk with the corresponding immobilized antigens Žannexins with low molecular mass.. Using immobilized synthetic peptides, Platonova et al. w31x have isolated the corresponding monospecific polyclonal immunoglobulins from human serum. For this purpose, several disks with different peptide ligands were used, which had been put into one housing. This allowed the adsorption of various, monospecific antibodies in a single step. Fig. 5 shows another example of the use of multidimensional chromatography on monoliths for the isolation of immunoglobulins. Monoclonal antibodies from ascites fluid were purified in a tandem configuration, consisting of an anion-exchange disk and a disk with immobilized Protein A.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

163

Fig. 4. Fast analysis of human IgG on a Protein A monolithic column. ŽA. Twenty microliters of human IgG solution Ž2.5 mgrml. and ŽB. 10-fold diluted human serum were injected. Reprinted from Ref. w31x with permission.

The isolated antibodies can be immobilized and used for the isolation of the corresponding antigen. It has been found that in the case of monoliths, just as in the case of bulk supports, the method of immobilization of antibodies, i.e. their presentation on the surface of the support, can sometimes be critical. For the immobilization of protein ligands, epoxy-activated monoliths CIM-disks ŽBIA Separations. were used w22,31x. These monoliths can easily be modified by spacers of different composition and length

164

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

Fig. 5. Separation of IgG from other proteins in mouse ascites fluid by two-dimensional chromatography, i.e. anion-exchange chromatography and Protein A affinity chromatography. A 1-ml volume of ascites fluid with monoclonal Antibody No. 69.31 was applied to a QA disk ŽCIM disk, BIA Separations. and a Protein A disk in tandem. A large part of the proteins binds to the QA disk, except the IgG, which is subsequently captured by the Protein A disk. The accompanying proteins are eluted from the QA disk with a salt gradient. Then the IgG is eluted from the Protein A disk. Reprinted from Ref. w27x with permission.

in order to achieve the best presentation of ligand molecules and binding of the antigen w46x. Another option is to bind the antibody across the Fc-part to the Protein A, immobilized on a monolith, and to bind it covalently by means of a cross-linker w67x. This warrants the presentation of the antigen-binding part of the immunoglobulin. For most applications, simple immobilization of antibodies on an epoxy-activated monolith was sufficient to allow this technique to be used successfully w22,31x. In all cases, it was possible to isolate the respective antigens within a very short time, even from very complex biological matrices. Schuster et al. w68x have made a direct comparison of monoliths and glass beads containing gigapores, both with immobilized antibodies. The method was immunoaffinity chromatography, exploiting the Ca2q-dependent interaction of the anti-Flag antibody and Flag-tagged proteins. They found that both supports are suited for high-throughput immunoaffinity chromatography. The isolation of the antigen was carried out in less than 2 min. For both supports, efficiency of separation was independent of flow-rate. However, in the case of the bulk support Žporous glass beads., the dynamic binding capacity depended on the flow-rate, while the dynamic binding capacity of the anti-Flag ŽCIM. monolithic disk was not influenced by flow. The reason for this advantage of the CIM disk may be that its convective mass transport is combined with minimal diffusion w28x.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

165

3. Immobilization of enzymes and fast conversion of substrates Monolithic supports and membranes are characterized, as often pointed out, by an almost complete lack of diffusion resistance during mass transfer. Therefore, they represent ideal supports for the immobilization of enzymes and fast conversion of substrates. Initial applications of these supports took place rather early w14x, but subsequent development has been sluggish, at best. Developments in the last 2 years have shown that the enzymes immobilized on the monolith can be used for analytical as well as preparative purposes. The progress of monoliths towards miniaturization makes them suitable for in-process analysis. The above-mentioned possibilities for scaling up in the case of monoliths w39x and of separation units with thin membranes w66,69x have shown that they are also suitable for preparative use. 3.1. Fast conÕersion of low-molecular-mass substrates Abou-Rebyeh et al. w14x carried out the first conversion of substrate with immobilized carbonic anhydrase in flow-through mode as early as 1991. The enzyme was immobilized on a disk with epoxy-activated poly Žglycidyl methacrylate.. By using such a disk with immobilized carbonic anhydrase, they created a kind of flow-through reactor with an enzyme immobilized on a monolith. Although at that time, no optimized hardware was available for this application, kinetic experiments concerning substrates with low molecular masses under dynamic conditions were carried out. In these experiments, rather low flow-rates up to 1.2 mlrmin were used, but it was still shown that higher flow-rates led to an increase in enzymatic activity Žcf. Fig. 6.. In experiments similar to those shown in Fig. 6, Petro et al. w69x found within the framework of a comparative

Fig. 6. Relationship between flow and activity of carbonic anhydrase, immobilized on a carrier membrane, 2-chloro-4-nitrophenyl acetate being used as substrate. Substrate solution was pumped through the membrane at different flow-rates. The enzymic activity increased at greater flow-rates. Reprinted from Ref. w14x with permission.

166

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

study that trypsin, immobilized on macroporous beads as well as on a monolithic support, not only has higher catalytic activity when bound to a monolith, but also in a much higher throughput. In these first experiments, the enzymes were immobilized directly on epoxy-activated supports. Both carbonic anhydrase and trypsin are rather small proteins. Apparently, they are easy to immobilize without damaging their activity or the enzyme–substrate interaction. As seen in Fig. 7, certain limitations apply in the case of complex enzymes, such as invertase. The experiment shown there indicates that the conversion of the substrate saccharose, which has a low molecular mass, is incomplete in the case of higher concentrations. The solution with a higher concentration had to be pumped several times through the unit with the immobilized enzyme for complete conversion w27x. Yeast invertase is a glycoprotein with a high portion of oligosaccharides in its molecule. It is therefore possible that this kind of immobilization of the molecule causes its partial inactivation or a partial loss of its enzymatic activity. The influence on enzymatic activity of the spacer used for immobilization on the monoliths has been investigated with the model system of glucose oxidase ŽGOX. w27x. The results are shown in Fig. 8. Glucose oxidase from yeast is also a glycoprotein with a rather high portion of carbohydrates. Fig. 8 also shows that enzymatic conversion is rather low if the enzyme is immobilized directly on the epoxy groups of the support. The corresponding detector response was about 55–60% of that of the commercially available detector with immobilized GOX. The response was improved, if a spacer like ethylendiamine and glutaraldehyde ŽEDA–GA. was added between GOX and the surface of the support. But still, it did not approach the response of the commercially available detector. If the enzyme was immobilized at its disaccharide part on previously immobilized lectin Con A, the necessary space between the enzyme and the support was established, and the active center was fully accessible to the substrate. Applying this optimized immobilization, a detector response comparable to that of the commercially available detector was achieved Žcf. Fig. 8.. Vodopivec et al. w37x have immobilized GOX to monoliths ŽCIM disks, BIA Separations. for the determination of glucose by on-line monitoring in two model systems, first, cultivation of the yeast Saccharomyces cereÕisiae and second, the production of citric acid by Aspergillus niger. A CIM disk with immobilized GOX, having a diameter of 12 mm and layer thickness of 3 mm, was built as an enzyme reactor into a flow-injection analysis ŽFIA. system and used for on-line glucose monitoring in both of the above-mentioned model systems. This so-called CIM GOD disk FIA system Žcf. Fig. 9. exhibited a good signal reproducibility and linear response in the range of 10–200 mgrl. Although less stable than the commercial-packed bed

Fig. 7. Enzymatic conversion of saccharose on a disk with immobilized invertase. ŽA. Sample before enzymatic conversion with 10% Žwrv. saccharose; ŽB. enzymatic conversion of a 5% Žwrv. saccharose solution. Peak 1s fructose, retention times 3.15 min; peak 2 s glucose, retention times 3.52 min; saccharose could not be detected; ŽC. enzymatic conversion of a 10% Žwrv. saccharose solution, conditions as in ŽB.. In this case, nonconverted saccharose was detected Žpeak 3, retention times 5.58 min.. Reprinted from Ref. w27x with permission.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

167

168

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

Fig. 8. Calibration curves for glucose standard solutions performed on different disks with the commercially available enzyme reactor. Curve 4 sGOX, immobilized on an epoxy-activated disk; curve 3sGOX, immobilized on a disk with ethylendiamine glutaraldehyde as spacer; curve 2 sGOX, immobilized on a disk with Con A; curve 1s commercial enzyme reactor. Reprinted from Ref. w27x with permission.

Fig. 9. On-line FIA system for glucose monitoring, consisting of: 6-W, six-way valve; I1, I2 injection valves; FIX1, FIX2, VARIO, peristaltic pumps; M, mixing chamber; B, CIM GOD disk; MC, mixing coil; VIS, detector; MP, data recording system; D, conical glass separator; C1, C2, carrier; R, reagent; W, waste. Reprinted from Ref. w37x with permission.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

169

GOX reactor, it shows greater sensitivity and it can be applied at higher buffer flow-rates Ž) 3 mlrmin.. 3.2. Fast conÕersion of high-molecular-mass substrates Only a few data are available on the use of monoliths with immobilized enzymes for conversion of high-molecular-mass substrates w27,70,71x. So far, all investigations have been carried out with immobilized trypsin. The results show that in the case of high-molecular-mass substrates, the enzymatic conversion depends to a large extent on the flow-rate Žcf. Fig. 10.. Therefore, the situation is fundamentally different from the

Fig. 10. Enzymatic conversion of high-molecular-mass substrates on a disk with immobilized trypsin. ŽA. Conversion of transferrin. Lane 1: calibration proteins; lane 2: sample before enzymatic conversion; lane 3: sample Ž5 ml, concentration: 1 mgrml., boiled in the sample buffer for SDS-PAGE under denaturing conditions and pumped through the disk Žtrypsin. at a flow-rate of 0.5 mlrmin Žpressure, 0.1 MPa.; lane 4: the same sample as in lane 3, but pumped through the disk Žtrypsin. at a flow-rate of 0.1 mlrmin Žno back pressure.. Reprinted from Ref. w27x with permission.

170

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

case where substrates with low molecular mass are converted in such an enzyme reactor. There, no such flow-dependency was found w37x. The fact that it is necessary to denature the substrate before enzymatic hydrolysis by adding SDS and mercaptoethanol shows that in this case, steric effects, i.e. accessibility to the cleavage site in the protein, play an important part in the interaction with the enzyme Žsee Fig. 10.. The experiments were reproducible and could be carried out in a single unit with immobilized enzyme ŽCIM Disk Trypsin.. For this reason, the use of such disks with immobilized proteins was recommended for the conversion of the investigated proteins and for subsequent peptide-mapping w70x. Xie et al. w71x have used a monolith made of polyŽ2-vinyl-4.4-dimethylazlactone-coacrylamide-co-ethylene dimethacrylate., with immobilized trypsin as enzymatic reactor, for the conversion of substrates with high and low molecular masses. The catalytic activity of a monolithic reactor is maintained even at a flow velocity of 180 cmrmin, if a substrate with a low molecular mass such as L-benzoyl arginine is used. In comparatative experiments with casein, used as a model for substrates with high molecular masses, such results could not be expected to be obtained. The authors explain this phenomenon by pointing to the high viscosity of casein solutions and their collodial character, characteristics which contribute significantly to the increase of resistance to flow within the monolithic reactor and which do not permit the use of the high flow velocities observed in substrates with low molecular masses. In a long-term experiment using a monolithic reactor with immobilized trypsin and casein as substrates, the group found that the initial enzymatic activity did not change during the first 24-h period of continuous operation. However, the caseinolytic activity decreased after 48 and 72 h. This loss of activity is probably due to an accumulation of collodial particles within the reactor and concurrent plugging of the pores. In a preliminary paper, Lim et al. w72x reported successful application of immobilized elastase on CIM disks for the preparative hydrolysis of the inter-a-trypsin inhibitor from human plasma. Recently, Zhou et al. w73x have immobilized the enzyme pectinase on the above-mentioned GMA-modified cellulose. However, the column was not used for the conversion of the respective substrate, but for the isolation of inhibitors with high molecular mass, the so-called polygalacturonase-inhibiting proteins.

4. Conclusion Owing to their special characteristic, i.e. an almost total lack of mass-transport resistance because of diffusion w28,74x, monoliths with immobilized ligands with either high or low molecular mass can be used for the fast separation of proteins and in-process control in affinity mode. If molecules with high molecular mass are used as ligands, the technology of immobilization must be optimized by inserting adequate spacers. Enzymatic conversion of substrates with high or low molecular mass can be achieved with units that contain enzymes immobilized on the surface of a monolith. New technological approaches will allow the scaling up of monolithic columns w39x and consequently, their use in preparative separations.

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

171

References w1x Reif O-W, Freitag R. Characterization and application of strong ion-exchange membrane adsorbers as stationary phases in high-performance liquid chromatography of proteins. J Chromatogr, A 1993;654: 29–41. w2x Josic´ D, Reusch J, Loster K, Baum O, Reutter W. High-performance membrane chromatography of ¨ serum and plasma membrane proteins. J Chromatogr 1992;590:59–76. w3x Mandaro R, Roy S, Hou K. Filtration supports for affinity separation. BiorTechnology 1987;5:928–32. w4x Tennikova TB, Belenkii BG, Svec F. High-performance membrane chromatography: a novel method of protein separation. J Liq Chromatogr 1990;13:63–70. w5x Hjerten ´ S, Liao J-L, Zhang R. High-performance liquid chromatography on continuous polymer beds. J Chromatogr 1989;473:273–5. w6x Svec F, Tennikova TB. Polymeric separation media for chromatography of biopolymers in a novel shape: macroporous membranes. J Bioact Compat Polym 1991;6:393–405. w7x Petro M, Svec F, Frechet JMJ. Molded continuous polyŽstyrene-co-divinylbenzene. rod as a separation ´ medium for the very fast separation of polymers: comparison of the chromatographic properties of the monolithic rod with columns packed with porous and non-porous beads in high-performance liquid chromatography of polystyrenes. J Chromatogr, A 1996;752:59–66. w8x Strancar A, Barut M, Podgornik A, Koselj P, Schwinn H, Raspor P, et al. Application of compact porous tubes for preparative isolation of clotting factor VIII from human plasma. J Chromatogr, A 1997;760: 117–23. w9x Minakuchi H, Nakanishi K, Soga N, Ishizuka N, Tanaka N. Octadecylsilylated porous silica rods as separation media for reversed-phase liquid chromatography. Anal Chem 1996;68:3498–501. w10x Schulte M, Lubda D, Delp A, Dingenen J. Preparative monolithic silica sorbents ŽPrepRODx. and their use in preparative liquid chromatography. J High Resolut Chromatogr 2000;23:100–5. w11x Brandt S, Goffe RA, Kessler SB, O’Connor JL, Zale SE. Membrane-based affinity technology for commercial scale purifications. BiorTechnology 1988;6:779–82. w12x Unarska M, Davies PA, Esnouf MP, Bellhouse BJ. Comparative study of reaction kinetics in membrane and agarose bead affinity systems. J Chromatogr 1990;519:53–67. w13x Thommes J, Kula M-R. Membrane chromatography—an integrative concept in the downstream process¨ ing of proteins. Biotechnol Prog 1995;11:357–67. w14x Abou-Rebyeh H, Korber F, Schubert-Rehberg K, Reusch J, Josic´ D. Carrier membrane as a stationary ¨ phase for affinity chromatography and kinetic studies of membrane-bound enzymes. J Chromatogr 1991;566:341–50. w15x Strancar A, Koselj P, Schwinn H, Josic´ D. Application of compact porous disks for fast separations of biopolymers and in-process control in biotechnology. Anal Chem 1996;68:3483–8. w16x Strancar A, Barut M, Podgornik A, Koselj P, Josic´ D, Buchacher A. Convective interaction media: polymer-based supports for fast separation of biomolecules. LC GC Int 1998;10:660–70. w17x Reif O-W, Nier V, Bahr U, Freitag R. Immobilized metal affinity membrane adsorbers as stationary phases for metal interaction protein separation. J Chromatogr, A 1994;664:13–25. w18x Champluvier B, Kula M-R. Sequential membrane-based purification of proteins, applying the concept of multidimensional liquid chromatography ŽMDLC.. Bioseparation 1992;2:343–51. w19x Kim M, Saito K, Furusaki S, Sato T, Sugo T, Ishigaki I. Adsorption and elution of bovine y-globulin using an affinity membrane containing hydrophobic amino acids as ligands. J Chromatogr 1991;585:45– 51. w20x Tennikova T, Freitag R. High-performance membrane chromatography of proteins. In: Aboul-Enein HY, editor. Analytical and preparative separation methods of biomacromolecules. New York: Marcel Dekker; 1999. p. 255–300. w21x Langlotz P, Kroner KH. Surface-modified membranes as a matrix for protein purification. J Chromatogr 1992;591:107–13. w22x Josic´ D, Lim Y-P, Strancar A, Reutter W. Application of high-performance membrane chromatography for separation of annexins from the plasma membranes of liver and isolation of monospecific polyclonal antibodies. J Chromatogr, B: Biomed Sci Appl 1994;662:217–26.

172

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

w23x Liao J-I. Continuous bed for conventional column and capillary column chromatography. In: Brown PR, Grushka E, editors. Advances in chromatography. New York: Marcel Dekker; 2000. p. 468–502. w24x Bamford C, Al-Lamee K, Purbrick M, Wear T. Studies of a novel membrane for affinity separations: functionalisation and protein coupling. J Chromatogr 1992;606:19–31. w25x Strancar A. Separation of biopolymers with different techniques of liquid chromatography. PhD Thesis. University of Ljubljana, Slovenia, 1997. w26x Turkova´ J. Oriented immobilization of biologically active proteins as a tool for revealing protein interactions and function. J Chromatogr, B: Biomed Sci Appl 1999;722:11–31. w27x Josic´ D, Schwinn H, Strancar A, Podgornik A, Barut M, Lim Y-P, et al. Use of compact, porous units with immobilized ligands with high-molecular masses in affinity chromatography and enzymatic conversion of substrates with high and low-molecular masses. J Chromatogr, A 1998;803:61–71. w28x Jungbauer A, Buchacher A, Josic´ D. Monoliths as stationary phases for separation of proteins and polynucleotides and enzymatic conversion. J Chromatogr, B: Biomed Sci Appl 2001;752:191–205. w29x Necina R, Amatschek K, Schallaun E, Schwinn H, Josic´ D, Jungbauer A. Peptide affinity chromatography of human clotting factor VIII: screening of the vWF-binding domain. J Chromatogr, B: Biomed Sci Appl 1998;715:191–201. w30x Amatschek K, Necina R, Hahn R, Schallaun E, Schwinn H, Josic´ D, et al. Affinity chromatography of human blood coagulation factor VIII on monoliths with peptides from a combinatorial library. J High Resolut Chromatogr 2000;23:47–58. w31x Platonova GA, Pankova GA, Il’ina IY, Vlasov GP, Tennikova TB. Quantitative fast fractionation of a pool polyclonal antibodies by immunoaffinity membrane chromatography. J Chromatogr, A 1999;852: 129–40. w32x Korol’kov V, Platonova G, Azanova V, Tennikova T, Vlasov G. In situ preparation of peptidyl polymers as ready-to-use adsorbents for rapid immunoaffinity chromatography. Lett Polym Sci 2000;7:53–61. w33x Wulff G. Molecular imprinting in cross-linked materials with the aid of molecular templates—a way towards artificial antibodies. Angew Chem, Int Ed Engl 1995;34:1812–32. w34x Takeuchi T, Fukuma D, Matsui J. Combinatorial molecular imprinting: an approach to synthetic polymer receptors. Anal Chem 1999;71:285–90. w35x Takeuchi T, Matsui J. Miniaturized molecularly imprinted continuous polymer rods. J High Resolut Chromatogr 2000;23:44–6. w36x Takeuchi T, Haginaka J. Separation and sensing based on molecular recognition using molecularly imprinted polymers. J Chromatogr, B: Biomed Sci Appl 1999;728:1–20. w37x Vodopivec M, Berovic M, Jancar J, Podgornik A, Strancar A. Application of convective interaction media disks with immobilised glucose oxidase for on-line glucose measurements. Anal Chim Acta 2000;407:105–10. w38x Svec F, Frechet JMJ. New designs of macroporous polymers and supports: from separation to bio´ catalysis. Science 1996;273:205. w39x Podgornik A, Barut M, Strancar A, Josic´ D, Koloini T. Construction of large volume monolithic columns. Anal Chem 2000;72:5693–9. w40x Klein E. Affinity membranes: a 10-year review. J Membr Sci 2000;179:1–27. w41x Josic´ D, Strancar A. Application of membranes and compact, porous units for the separation of biopolymers. Ind Eng Chem Res 1999;38:333–42. w42x Tennikova T, Freitag R. An introduction to monolithic disks as stationary phases for high performance biochromatography. J High Resolut Chromatogr 2000;23:27–38. w43x Freitag R, Splitt H, Reif O-W. Controlled mixed-mode interaction chromatography on membrane adsorbers. J Chromatogr, A 1996;728:129–37. w44x Champluvier B, Kula M-R. Dye-ligand membranes as selective adsorbents for rapid purification of enzymes: a case study. Biotechnol Bioeng 1991;40:33–40. w45x Mandjiny S, Vijayalakshmi MA. Membrane based pseudobioaffinity chromatography of placental IgG using immobilized L-histidine. Biotechnol Blood Proteins 1993;227:189–95. w46x Hahn R, Amatschek K, Schallaun E, Necina R, Josic´ D, Jungbauer A. Performance of affinity chromatography with peptide ligands: influence of spacer, matrix composition and immobilization chemistry. Int J Bio-Chromatogr 2001, in press. w47x Legallais C, Anspach FB, Bueno SMA, Haupt K, Vijayalakshmi MA. Strategies for the depyrogenation

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

w48x w49x

w50x

w51x w52x w53x

w54x w55x w56x w57x

w58x w59x w60x w61x w62x w63x w64x w65x w66x w67x w68x w69x

w70x

173

of contaminated immunoglobulin G solutions by histidine-immobilized hollow fiber membrane. J Chromatogr, B: Biomed Sci Appl 1997;691:33–41. Yang L, Jia L, Zou H, Zhang Y. Immobilized iminodiacetic acid ŽIDA.-type Cu2q-chelating membrane affinity chromatography for purification of bovine liver catalase. Biomed Chromatogr 1999;13:229–34. Te Booy MPWM, Riethorst W, Faber A, Over J, Konig ¨ BW. Affinity purification of plasma proteins: characterization of six affinity matrices and their application for the isolation of human factor VIII. Thromb Haemostasis 1989;61:234–7. Morfini M, Longo G, Messori A, Lee M, White G, Mannucci P. Pharmacokinetic properties of recombinant factor VIII compared with a monoclonally purified concentrate ŽHemofil w M.. Thromb Haemostasis 1992;68:433–5. Huang PY, Baumbach GA, Dadd CA, Buettner JA, Masecar BL, Hentsch M, et al. Affinity purification of von Willebrand factor using ligands derived from peptide libraries. Bioorg Med Chem 1996;4:699–708. Fassina G, Verdoliva A, Palombo G, Ruvo M, Cassani G. Immunoglobulin specificity of TG 19318: a novel synthetic ligand for antibody affinity purification. J Mol Recognit 1998;11:128–33. Sproule K, Morrill P, Pearson JC, Burton SJ, Hejnaes KR, Valore H, et al. New strategy for the design of ligands for the purification of pharmaceutical proteins by affinity chromatography. J Chromatogr, B: Biomed Sci Appl 2000;740:17–33. Merrifield RB. Solid-phase peptide synthesis: I. The synthesis of a tetrapeptide. J Am Chem Soc 1963;85:2149–54. Houghten RA, Pinilla C, Appel JR, Blondelle SE, Dooley CT, Eichler J, et al. Mixture-based synthetic combinatorial libraries. J Med Chem 1999;42:3743–78. Houghten RA, Wilson DB, Pinilla C. Drug discovery and vaccine development using mixture-based synthetic combinatorial libraries. Drug Discovery Today 2000;5:276–85. Teng SF, Sproule K, Husain A, Lowe CR. Affinity chromatography on immobilized AbiomimeticB ligands: synthesis, immobilization and chromatographic assessment of an immunoglobulin G-binding ligand. J Chromatogr, B: Biomed Sci Appl 2000;740:1–15. Li R, Dowd V, Stewart DJ, Burton SJ, Lowe CR. Design, synthesis, and application of a protein A mimetic. Nat Biotechnol 1998;16:190–5. Nachman M. Kinetic aspects of membrane-based immunoaffinity chromatography. J Chromatogr 1992; 597:167–72. Nachman M, Azad ARM, Bailon P. Membrane-based receptor affinity chromatography. J Chromatogr 1992;597:155–66. Wang QC, Svec F, Frechet JM. Macroporous polymeric stationary-phase rod as continuous separation medium for reversed-phase chromatography. Anal Chem 1993;65:2243–8. Josic´ D, Bal F, Schwinn H. Isolation of plasma proteins from the clotting cascade by heparin affinity chromatography. J Chromatogr 1993;632:1–10. Kasper C, Meringova L, Freitag R, Tennikova T. Fast isolation of protein receptors from streptococci G by means of macroporous affinity discs. J Chromatogr, A 1998;798:65–72. Hagedorn J, Kasper C, Freitag R, Tennikova T. High performance flow injection analysis of recombinant protein G. J Biotechnol 1999;69:1–7. Zhou D, Zou H, Ni J, Jia L, Zhang G, Zhang Y. Membrane supports as the stationary phase in high-performance immunoaffinity chromatography. Anal Chem 1999;71:115–8. Jia L, Yang L, Zou H, Zhang Y, Zhao J, Fan C, et al. Protein A tangential flow affinity membrane cartridge for extracorporeal immunoadsorption therapy. Biomed Chromatogr 1999;13:472–7. Schneider C, Newman R, Sutherland D, Asser U, Greaves M. A one-step purification of membrane proteins using a high efficiency immunomatrix. J Biol Chem 1982;257:10766–9. Schuster M, Wasserbauer E, Neubauer A, Jungbauer A. High speed immuno-affinity chromatography on supports with gigapores. Bioseparation 2000;9:259–68. Petro M, Svec F, Frechet JMJ. Immobilization of trypsin onto AmoldedB macroporous polyŽglycidyl ´ methacrylate-co-ethylene dimethacrylate. rods and use of the conjugates as bioreactors and for affinity chromatography. Biotechnol Bioeng 1996;49:355–63. Josic´ D., Baum O., Loster K., Reutter W., Reusch J. Application of compact porous disks for separation ¨ of proteins and immobilization of enzymes. PREP ’92, 6–8 April 1992, Nancy, France, p. 113–7.

174

D. Josic, ´ A. Buchacherr J. Biochem. Biophys. Methods 49 (2001) 153–174

w71x Xie S, Svec F, Frechet JMJ. Design of reactive porous polymer supports for high throughput bioreactors: ´ polyŽ2-vinyl-4,4-dimethylazlactone-co-acrylamide-co-ethylene dimethacrylate. monoliths. Biotechnol Bioeng 1999;62:30–5. w72x Lim Y-P, Callanan H, Hixson D, Barut M, Podgornik A, Josic´ D, et al. Immobilized elastase–CIM disks for a rapid preparative cleavage of inter-alpha-inhibitor complex proteins. ISPPP2000. Ljubljana, Slovenia. w73x Zhou D, Zou H, Ni J, Wang H, Yang L, Zhang Y. Membrane affinity chromatography for analysis and purification of biopolymers. Proc. ISPPP 2000. Chromatographia 1999;50:27–34. w74x Meyers JJ, Liapis AI. Network modeling of the convective flow and diffusion of molecules adsorbing in monoliths and in porous particles packed in a chromatographic column. J Chromatogr, A 1999;852:3–23.