CA2102237C - Method and apparatus for immobilized enzyme reactions - Google Patents

Abstract

Perfusive matrices (10) useful in immobilized enzyme reactors (IMERs) are disclosed. The matrices define first and second interconnected sets of pores and an intraparticle high surface area (18) capable of immobilizing enzymes, in fluid communication with the second set of pores. The first and second sets of pores are embodied, for example, as the interstices (12) among particles and throughpores (14) within the particles. The particles also contain diffusive transport pores (16). The dimensions of the first and second sets of pores are such that, at fluid flow rates and pressure drops practical for immobilized enzyme reactions, convec-tive flow occurs in both pore sets, and the convective flow rate exceeds the rate of diffusion of the substrate and product to and from immobilized enzymes within the particles. This approach allows rapid enzyme loading and substrate conversion reactions at optimal catalytic flow rates.

Description

~ WO92/1~36 PCT/US92/0~7 ~ 2102~3~

METHOD AND APPARATUS FOR IMMOBILIZED ENZYME REACTIONS

Field of the Invention This invention relates generally to methods and apparatus for conducting enzymatic reac_ions. In particular, the invention relates to methods ~nd apparatus for conducting highly productive enzymatic reactions utilizing enzymes immobilized on a perfusive matrix.

Backqround of the Invention The art has placed increasing emphasis on the use of immobilized enzymes and on developments in immobilized enzyme technology. Enzyme immobilization provides a reaction system with high specificity and sensitivity, and generally increases enzyme stability, thereby converting enzymes into versatile tools useful in a wide variety of applications. Immobilized enzyme reactors ("IMER"s) are used in many industrial processes, including synthesis and analysis in the chemical, pharmaceutical, food, and bioremediation industries, as well as for clinical analysis and therapeutic purposes in medicine.

Currently there are a number of limitations to the enzyme reactors in the art. Chief among these are the current methods of enzyme immobilization, which affect IMER capacity, cost and speed.

WO92/1~ ~ - 2 PCT/US92/02~87 21~2237 Of the various ways known in the art for immobilizing enzymes, one of the most common and potentially versatile methodologies useful in IMER
t~chnology is the use of matrix-bound enzymes. The nature of the support matrix for an immobilized enzyme system is important to its function. In continuous flow reactors, ~or example, materials with poor dimensional stability should be avoided. In addition, a large surface area-to-volu~e ratio generally is desired. The ideal support matrix is one which promotes substrate binding, decreases product inhibition, shifts the apparent pH optimum to the desired value, increases enzyme stability, and discourages microbial growth. Cost of production and ease of preparation also should be considered.

Typically, matrix supports useful in liquid chromatography are used. Particularly useful matrices comprise porous particles which provide the desired high surface area-to-volume ratio. Traditional materials for liquid chromatography also are characterized by operational constraints based on their geometric, chemical, and mechanical properties. Soft, porous particles, for example, cannot be subjected to pressure drops exceeding about 50 psi because they are easily crushed. This implies that there are similar constraints in effecting enzymatic reactions by passing substrate solutions through enzymes immobilized on such matrix materials.

In high performance liquid chromatography (HPLC), instead of employing as matrices soft, particulate, gel-like materials having mean diameters on the order of 100 ~m, one employs smaller, rigid, porous and substantially homogeneous beads of about 10 to 20 ~m in ~ W092/1~36 _ 3 _ 2 ~ Q ~ 2 3 7 PCT/USg2/02~7 .

diameter and made of an inorganic material such as silica, or a rigid polymer such as a styrene di~inylbenzene. R~c~use the dense ~ACki ng of these smaller beads creates a high resistance to liquid flow, the equipment is designed to operate at hiqh pressures, which allows rapid fluid transfer.

Unfortunately, flowthrough speeds of solutes other than small molec~ r weight solutes become limiting in conv~ntion~ HPLC matrices, prim~rily because mass transfer within the pores of the ~PLC
particles is diffusive, as compared to the mass transfer between particles, which is convective.
Accordingly, the time it takes for an enzyme's substrate to diffuse to an enzyme locus within a pore from a region where convective transport dominates, and for the product to diffuse back, becomes a reaction rate-limiting factor. Beyond a certain opt~mal flow velocity, further increases serve only to increase breakthrough of unreacted substrate. In addition, these diffusion limitations within pores can result in significant substrate and end product inhi~ition. To overcome this inhibition, enzyme reactors of the art often employ a "stop-flow" protocol and/or multiple sequential subtrate injections, to limit the substrate concentration provided to the enzyme sites at any given time. While these protocols can limit the effects of inhibition, they are time-consuming and can introduce unwanted errors into the analysis. The limiting flowthrough speeds also substantially limit the speed with which enzymes can be loaded onto the particle surfaces. Currently, this step alone can takes hours or days to complete.

W092/1~36 PCT/US92/0~7 - 21û22~
Efforts to reduce these diffusional limitation effects have led to the development of methods for loading the enzyme only onto the outer edge of a porous particle. While this can reduce the enzyme loading timeframe and limit diffusional pa~hlengths~ it can limit the capacity of the system substantially.
Alternatively, the particle size may be enlarged to increase capacity, but this generally only results in longer diffusional path lengths within the particle.
Reducing or eliminating the ~er~nde~ce of Lmmobilized enzyme reactors on the diffusive flow rate of solutes would enh~nce their productivity.

It is therefore an object of this invention to provide an enzyme reactor that can be rapidly loA~d, and which effectively is not diffusion limited. It is another object of this invention to provide a method and apparatus for immoblized enzyme reactions wherein substrate and product inhibition is limited. Still another object of the invention is to provide a method and apparatus for immobilized enzyme reactions that is adaptable and cost-efficient to perform.

~ WO92/18636 PCT/US92/0~7 ~ 21~%37 Summary of the Invention This invention pertains to perfusive matrices capable of immobilizing enzymes and to the use of such matrices to perform enzymatic reactions. Perfusive matrices comprise rigid, porous, high surface area materials such as particles which may be of the same mean diameter as are employed in conventional chromatography matrices. The geometry of perfusive matrices are configured to allow convective fluid transfer both within and between the par~ s.
Typically, 10-20 ~m diameter particles of per~usive matrices have throughpores of relatively large mean diameter (e.g., 6,000 to 8,000 A) and a high surface area network of internal, blind subpores of smaller mean diameter (500 to 1500 A) within the throughpores.
The enzymes can be immobilized on all avA;~hle surface areas, including within the throughpores and the subpores.

Perfusive matrices are characterized by a relatively small ratio of the mean diameters of the interparticle flow paths to the intraparticle throughpores, thereby permitting intraparticle convective flow at accessible fluid flow velocities.
The resulting network limits the diffusional path lengths within the particles so that mass transfer within the particle pores is governed by convection rather than diffusion over a large range of high flow rates. Where the perfusion matrix comprises packed particles, the diameter of the particles determines the mean diameter of the interparticle spaces in a packed bed. In preferred embodiments, the ratio of the mean particle diameter to the mean diameter of the intraparticle throughpores is less than 70, most ~ 21 ~2~37 preferably less than 50. Preferred subpore diameters are within the range of about 300-700 A . In addition, the low ratio (and correspondingly larger intraparticle pore size) substantially reduces particle pore effects.
Preferred ratios of convective flow velocities through the interparticle and intraparticle pores are between about 10:1 to 100:1.

In a perfusive matrix containing immobilized enzymes, the velocity of a mobile phase can be increased, for example, to greater than 10 to 100 times that of conventional HPLC systems without substantial loss of enzyme binding capacity or substrate conversion efficiency. Typically, mobile phase velocities of greater than 1000 cm/hr can be achieved. In addition, the significantly reduced diffusional pathlengths that characterize perfusive matrices substantially eliminate substrate or product inhibition even when the system is operated in a non-perfusive mode. A description of perfusive matrix material in chromatogra~hic contexts is provided in co-pending Canadian patent application serial number 2,018,393 filed June 16, 1990 and in Afeyan et al., ~1990) Bio Technoloqy 8:203-206. Perfusive matrix materials are available commercially from PerSeptive Biosystems, Inc. of Cambridge, Massachussetts, U.S.A.

In one aspect, this invention is a method for conducting an enzymatic reaction using a perfusive matrix formed by packing a multiplicity of particles defining therewithin through pores and substrate interactive surface regions within the throughpores comprising immobilized enzymes. A solution of enzyme substrate is passed through the matrix at a velocity WO92/1~36 PCT/US92/02~7 ~ 7 sufficient to cause convective fluid flow in the throughpores at a rate greater than the rate of substrate diffusion through the throughpores.

In another embodiment, a matrix is provided defining inter-~o~n~cted first and second throughpore sets wherein the members of the first throughpore set have a greater mean diameter than the members of the second throughpore set. Enzymes are immobilized, using chemistries known per se, or novel methods, on all surfaces ~f the matrix, including on surface regions in fl~id communication with the members of the second throughpore set. A solution of enzyme substrate is passed through the matrix at a rate sufficient to induce convective flow through both throughpore sets, and to induce a rate of convective flow through the second throughpore set that is greater than the rate of diffusion of the substrate within that set.

In another aspect, the invention provides a method for conducting an enzymatic reaction utilizing a matrix defining interconnected first and second throughpore sets dimensioned to allow convective fluid flow through both throughpore sets. Both throughpore sets comprise a multiplicity of throughpores for channeling through the matrix an enzyme solution and a substrate solution reactive with that enzyme. The matrix also includes interactive surface regions capable of immobilizing enzymes and which are in fluid communication with the members of the second pore set.

Enzymes preferably are immobilized on the surfaces of the matrix by passing a solution containing the enzyme through the matrix. In preferred embodiments of this invention, the fluid mixture is WO92/1~36 PCT/US92/02~7 2 10~;~3-7 8 -passed through the matrix at a rate sufficient to produce convective fluid flow through both pore sets, the velocity through the first set being greater than the velocity through the second set, and the convective fluid flow velocity through the second pore set being greater than the diffusive flow rate of the enzyme within the second pore set. The dimensions of the members of the second pore set and the interactive surface regions permit flow through the members of the second pore set at a rate such that the time for a solu~e to diffuse to and from the interactive surface regions is comparable to or shorter than the time for the solute to flow convectively past the region.

To bring about reaction between ~he enzyme substrate and the immobilized enzyme, the substrste solution then is passed through the matrix at a fluid flow rate sufficient to allow catalysis of the substrate by the immoblized enzyme. In preferred embodiments, this fluid flow r~te is sufficient to produce convective fluid flow through both pore sets.
This allows the enzymatic reaction to take place under kinetically very favorable conditions. Specific dimensions and flow velocities are described in greater detail below.

In yet another aspect, the invention provides a method for performing a series of enzyme reactions in successive zones within the matrix. In this method each enzyme is loaded at a perfusive fluid flow velocity and at a concentration insufficient to saturate all available binding sites on the matrix. In addition, the enzymes are loaded in the order in which they are to be used. Because the perfusive flow rate allows the enzyme solution to encounter all available WO92/1~36 PCT/US92/02~7 21~237 binding sites it flows past, the first enzyme added will saturate the avaliable binding sites it first encounters, forming a discrete zone of immobilized enzyme. The second enzyme, added subsequently, then will flow past this zone of saturated binding sites cont~i n i ng the immobilized first enzyme and will begin binding at the first available binding sites encountered below, forming a second discrete zone composed of immobilized second enzyme. Similarly, the next enzyme added will flow past both these first ~nd second zones and wili occupy available bin~ing sites further down the column. In this way multiple, different enzymes may be bound in successive, discrete zones. The method is particularly useful where multiple, different enzymatic reactions with ~ given substrate are desired. For example, the IMER may be constructed to mimic the actions of an enzyme complex found in nature. Alternatively, multiple different reactions may be required to obtain a product of interest. In these cases the product formed by reaction of a substrate with the first enzyme becomes the substrate for the reaction with the immobilized enzyme in the zone below. In preferred embodiments, optimal kinetics for all reactions in the system can be achieved with one given flow rate.

Alternatively, the multiple enzyme system may be used to perform catalytic reactions on different substrates. A particularly useful application of this method is in clinical analyses, to assay the presence and/or concentration of different solutes in a body fluid sample. In this case the different enzymes need not be se~regated into discrete zones.

WO92/1~ ~ PCT/US92/02887 -- 10 -- ~

In still another aspect, the invention provides a novel enzyme reactor including a rigid matrix defining interconnected first and second pore sets dimensioned to allow convective fluid flow through both throughpore sets, each pore set comprising a multiplicity of pores for chAnneling through the matrix a solution of enzyme or enzyme substrateO The matrix further defines surface regions comprising immobilized enzyme and in fluid communication with the members of the second throughpore set. The relative dimensions of the members of the first and second throughpore sets and the surface regions are fixed so that when fluid is passed through the matrix at a preselected velocity, there is convective fluid flow through both pore sets.
As a result, even though diffusion is required, the system is not "diffusion bound" except at very high flow rates. At practical flow rates, the rate of diffusion is not rate limiting in the productivity of the system.

In one embodiment of the apparatus of the invention, plural different enzymes may be immobilized on the matrix surfaces. The different enzymes further may be immobilized on the matrix in successive zones.

In another embodiment of the apparatus of the invention, the rigid matrix comprises a multiplicity of interfacing particles defining an interstitial volume which constitutes the first throughpore set. Each individual particle defines a plurality of throughpores, constituting the second pore set, and a plurality of blind pores in fluid communication with the throughpores, within which enzymatically active surface regions are located. Preferably, throughpores, subpores, and any interconnecting pores are ~ WO~2/1~ ~ 11 - PCT/US92/0~7 ~iO22~7 anisotropic. Currently preferred particles have a mean diameter greater than about 50 ~m, most preferably greater than about 100 ~m, and have a ratio of me~n particle diameter to mean intraparticle throughpore diameter of less than 70. A preferred geometry of the particles comprises adhered clusters of smaller similar clusters made up of small interadhered particles called porons. The preferred material for manufacture of the particles is polysLylel,e divinyl benzene copolymer.
Such particles are available commercially from PerSeptive Biosystems, Inc. of Cambridge, MA.

These and other aspects of the invention will be understood more fully by reference to the following detailed description in conjunction with the att~ch~d drawing in which like-referenced characters refer to like parts throughout the several views.
-WO92/18636 - 12 - PCT/US92/0~7 21~22~7 ` ~
Brief Description of the Drawing:

FIGURE 1 is a schematic representation of a particle suitable for use in forming a perfusive matrix contAini~g immobilized enzymes in accordance with the ingS of the present invention;

FIGURE 2 is a schematic representation of an apparatus embodying the present invention; and FIGURE 3A-C are graphic representations for perfusive immobilized enzyme reactiong performed under various reaction conditions.

WO92/1~36 - 13 - PCT/US92/02~7 21~2~7 Detailed Description The matrix of the enzyme reactors of this invention is characterized by a geometry which is bi-modal or multi-modal with respect to its porosity snd which has interactive surface regions capable o~
immobilizing enzymes. The matrix defines a set of pores of larger di~meter, such as are defined by the interstices among a bed of particles, which determine pressure gradient and fluid flow velocity through the bed, and a set of pores of smaller diameter (e.g., anisotropic throughpores). The smaller pores penmeate the individual particles. At fluid flow velocities above a threshold level, these pores serve to deliver by perfusion a solution of sn enzyme or enzyme subs~rate to substantially all surface regions within the particles.

FIGURE 1 shows schematically a matrix p~rticle 10 suitable for use with the present invention. As illustrated, in addition to extra-particle pores or channels 12, which have a relatively large mean diameter (defined by the interstices amon~ particles), a matrix utilizing the particle 10 also comprises a second set of throughpores 14, here embodied as pores defined by the body of the particle 10. Also de~ined by the particle 10 is a set of diffusive transport pores 16. The mesn diameter of the pores 12 is larger than the throughpores 14. In accordance with the immobilized enzyme reactor of the invention, the ratio of the mean diameters of pores 12 and 14 is such that, when the particles are close-packed in a bed, there exists a threshold of fluid velocity which can practically be achieved in the bed which induces a convective flow within pores 14 that is faster than the WO92/1~36 4 PCT/US92/0~87 -- 1 -- ~
2~ Q2237 rate of diffusion within pores 14. Above the threshold bed velocity the bed is said to be operating in the perfusive domain where contact between enzyme and substrate is no longer bound by diffusive transport.
Precisely where this threshold of perfusion occurs depends on many factors, but primarily it ~ep~n~s upon the ratio of the mean di~meters of the first and second pore sets, here, pores 12 and 14 respectively. The smaller th~t ratio, the lower the threshold velocity.
In preferred embodiments, the ratio of the mean particle diameter to mean intraparticle throughpore diameter is less than 70, most preferably less than 50.

In accordance with the invention, the particle 10 inclu~es a large surface region 18 within the particles onto which enzymes can be immobilized.
Enzymes may be covalently att~ched to the surfaces 18 at high ron~tration using any one of a nu~ber of techniques well known to those skilled in the art.
DepenAin~ on the composition of the matrix material and the enzyme, covalent attachment can be through functional groups such as carboxyl, amino, hydroxyl, sulfhydryl, hydroxyphenyl groups and/or other groups.
In addition, the matrix first may be derivatized to create the desired functional group for a given attachment protocol. Additionally, a pendant chain can be attached to the surface having a terminal functional group distal to the surface to which the enzyme can be coupled resulting in an immobilized enzyme which is attached by a "leash~' at a distance away from the surface. Enzymes also may be covalently bound to the matrix surface by means of difunctional crosslinkers.
Further details on methods for covalent immobilization of matrices can be found in a number of references in the art. Included among these are Falb, R . D .

WO92/1~36 15 PCT/US92/02~7 21022'~7 pp. 67-76, in Enzyme Enqineerinq Vol 2, Pye, E. at al.
eds., (1973) Plenum Press, NY, and White et al., (1980) Enzyme Microb. Technol. 2:82-90.

Alternatively, enzymes may be noncovalently attached to the matrix support. Noncovalent attachment provides a number of advantages over covalent attachment, chief among them being the cost efficiency and adaptability this approach provides. A given enzyme can be bound to the matrix using any of a number of well-known noncovalent interactions. Thereafter the enzymatic conversion is run, and the enzyme subsequently eluted. The matrix then can be reloaded with the same or a different enzyme, thereby substantially increasing the life of a given enzyme and/or reactor system. Alternatively, multiple substrate conversions may be performed with a given enzyme bound to the matrix, which then is eluted and replaced. As removal of the enzyme does not require breaking strong chemical bonds, harsh extraction or "stripping" conditions are not needed, allowing the enzyme to be reused as needed. In addition, noncovalent attachment is thought to minimize conformational distortion of the enzyme ne~e~ to achieve binding. It will be understood by those skilled in the art that substrate conversions performed using noncovalently immobilized enzymes will need to be run under conditions that do not induce elution of the enzyme. This may be achieved by manipulation of various reaction parameters, such as buffer pH and/or salt conditions. Particularly useful matrices for non-covalent enzyme immobilization include ion exchangers, and matrices that bind by hydrophobic interactions (e.g., "reverse phase" matrices.) .

WOg2/18636 - 16 - PCT/US92/02$87 '~ 102237 In any enzyme immobilization procedure, a pri~ary design criterion is to maximize enzyme activity of the resulting enzyme reactor. Even though the reactor may be highly stable and permit repeated use of the enzyme, a significant loss of enzymatic activity in the attachment step often can make the reactor economically unfeasible. Accordingly, the chemistry for immobilization should be compatible with the enzyme.

The amount of enzymatic activity per unit weight of reactor matrix is another important factor and, for most purposes to avoid excessive h~ i ness, a reactor should have an enzyme activity of at least 100 U/g of matrix material, where a unit of enzyme activity (U) is defined as one micromole of substrate converted per minute.

An important consideration for enzyme reactors int~ndeA for continuous operation is the suitability of the matrix for column flow. A high surface area usually is required to obtain the desired levels of enzymatic activity. High surface area can be attained by employing microporous materials, but this can result in inefficient substrate transport by diffusion.
Increases in surface area also can be achieved by reducing particle size, but the column flow properties of the reactor are usually unsatisfactory.

To circumvent these problems, highly porous, rigid materials are used. It becomes apparent that a matrix comprising a multiplicity of particle 10 is particularly well suited for enzyme immobilization.
The exceptionally high surface area-to-volume ratio achievable with perfusive matrix particles (typically on the order of about 30-50 m2/ml for 10-20 ~m -W092/1~36 - 17 - PCT/US92/02~7 210~37 particles) allow immobilized enzyme reactions to be performed in small volumes without compromising capacity. In addition, the rapid throughput of perfusive systems allow enzymes to be loaded rapidly onto a perfusive reactor system (e.g., in seco~ or minutes) in a fraction o~ the time currently required without loss of binding capacity. Moreover, as will be appreciated by those skilled in the art, the rapid throughput of the system and its small size allow one to ~ni~ nte reaction parameters rapidly and easily when developing a reaction protocol.

Virtually any enzyme can be utilized in the enzyme reactor of this invention for an equally large variety of applications. In addition, multiple different enzymes may be immobilized on the matrix to catalize reactions with multiple, different subtrates. Useful enzymes include oxido-reductases, trsnsferases, hydrolases, lyases and isomerases. Useful applications include a wide range of industrial applications, including those in the chemical, food and fragrance industries; in bioremediation, including the treatment of waste water by immobilization of pesticide-detoxificstion enzymes; snd in pharmsceutics, particulsrly for the sepsration of chiral isomers from a racemic mixture. Among the many useful applications in medicine are the use of IMERs as part of extracorporeal devices, and in a variety of clinical analyses, including assaying the presence and/or concentration of one or more solutes of interest in a body fluid sample. In addition, the IMERs of this invention are useful for the biochemical analysis of particular enzymatic and metabolic reactions.

WO92/1~ ~ - 18 - PCT/US92/02~7 2~2237 In addition to isolated and purified enzymes, it should be noted that relatively impure and/or heterogeneous enzyme preparations, such as those derived from cell extracts, cell lysates, partiaily purified enzyme isolates and whole cells can also be used, albeit at some reduction in the enzymatic activity. Accordingly, the tenm "enzyme" as used herein is meant to broadly include catslytic enzymes in all of these forms.

It also is possible to immobilize plural enzymes of different types in successive zones on the matrix surfaces to affect a series of enzymatic changes in a single pass. For example, enzymes normally associated as a complex to perfonm sequential enzymatic reactions in nature may be immobilized in series in a perfusive matrix in the following manner. A solution contAin;ng a r~n~Pntration of the first enzyme in the series insufficient to saturate all available bindinq sites on the matrix is provided to the perfusive matrix, e.g., to packed particles in a column. The solution further is provided at a fluid flow velocity sufficient to produce perfusion. At or above this flow rate the enzyme will encounter virtually all available binding sites it passes, thereby saturating the first available sites, forming a discrete zone of immobilized enzyme. A second solution containing a second enzyme, also at a concentration insufficient to saturate all available binding sites, then is added. The enzymes in the second solution, also provided to the matrix at a perfusive flow rate, will flow past the zone of saturated binding sites containing immobilized first enzyme, and will begin binding the first available sites encountered further down the column, forming a second zone of immobilized enzyme. Other, different WO92/1~36 19 PCT/US92/02~7 ~ 23~2'~7 enzymes then may be added in the same manner. The number of enzymes that may be added will be limited primarily only by the capacity of the column and the concentration of enzymes needed to effect efficient catalysis. It will be appreciated by those skilled in the art that optimal kinetics for enzyme reactions performed in series will require use of enzymes with similar reaction rates, and/or limiting the flow r~te to a rate which will provide optimal kinetics for the slowest reaction.

A schematic representation of a system for practicing the invention is illustrated in FIGURE 2. A
solution of enzyme substrate 20, is pumped by a pump 22 to an enzyme reactor 24. The enzyme reactor 24 includes a matrix comprising a packed bed of particles 10. While a matrix comprising a packed bed of particles 10 is pre~erred, other matrix forms may be used. It is necessary only that the matrix be rigid, so that it can withstand substantial pressure drops, and that it define throughpore sets capable of convectively ch~nnel 1 ing a solution of enzyme substrate throughout the column 24 at reasonable, and preferably low, pressure drops. Immobilized on the interactive surface regions 18 of the particle 10 are enzymes 25.
In the practice o the invention, the solution of enzyme substrate 20 is passed through the reactor 24, thereby inducing an enzymatic reaction resulting in a product stream 26 exiting from the enzyme reactor 24.

The apparatus also may be part of an automated system. In this case, the apparatus preferably further comprises a multi-port sampling valve 28 which may provide the enzyme(s) to be loaded, as well as all necessary solvents or buffers, including washing WO92/18636 20 PCT/US92/0~7 2~Q2~37 solvents, eluting solvents, and recycling or "stripping~ solvents. A valve 30 at the exit of the reactor 24 may direct the product stream 26 to a detector 32, to waste collectors 34, or to product collectors 36. In addition, valve positions may be under computer control As discussed above, the rate at which a solution (e.g., 20J is passed through the enzyme reactor 24 and the dimensions of the throughpores 14 and the enzyme interactive surfaces 18 are important to the practice of the invention. That is, flow rates sufficient to induce convective flow through both the interstitisl pores 12 and the throughpores 14 should be achievable without reguiring excessive pressure drops.
Moreover, for true perfusive operation, the convective flow rate within the throughpores 14 should be greater than the ra~e of diffusion of the solution 20 within the throughpores 14.

In the currently preferred embodiment of this invention, perfusive particles and their throughpores are of a mean diameter sufficient to permit substrate conversion to be performed in perfusion mode. One measure of the mass transfer of a solute through a pore is given by a characteristic Peclet number (Pe), a ~ en~ionless quantity equal to VL/D, where V is the convective velocity through the pore, L is its length, and D is the diffusivity of the solute through the pore. In conventional porous matrix systems, under all regimes, the Peclet number which describes the ratio of convective to diffusive transport within the pores of a material is always much less than one. In perfusive systems, the Peclet number in the second set of pores (e.g., the throughpores) is always greater than one.

WO92/1~ ~ - 21 - PCT/US92/02~7 ~10~7 As should be evident from the foregoing, the size of the particle (and the related intraparticle throughpore) will affect the flow rate and Peclet number. For example, for 10 ~m particles having 4,000A
intraparticle throughpores, and for a solute having a pore diffusi~ity of 10-7 cm2/sec~ the threshold flow rate is about 300 cm/hr. Above this threshold, it will be found that increased pressure drop and velocity permit increased throu~hput per unit volume of matrix above levels heretofore achievable. Extrsordinary productivities are achieved at flow rates of within the range of 1000-4000 cm/hr.

Particle sizes commonly used in IMERs are substantially larger than conven~io~l HPLC matrix particles, often on the order of 100-lOOO~m in diameter. Perfusive matrices comprising particles in this size range can have intraparticle throughpores in the range of aobut 2-20 ~m. Thus, for 200 ~m particles and a solute having a pore diffusivity on the order of lo- 6 cm2/sec, to achieve a Peclet number of grester than one, the threshold flow rate need only be about 0.3 cm/hr to achieve convection within the pores (or, sssuming only 1% bed to pore velocity, 30 cm/hr).
Accordingly, perfusion can be achieved at very low pressure drops and at sufficiently slow flow rates that enzyme conversion reactions are not compromised, even for reactions requiring flow rates at fractions of a ml/min.

For smaller particles (e.g., 10 ~m particles), optimal catalytic kinetics for immbilized emzyme reactions may require reductions in flow rates to ~elow the perfusive threshold. However, it will be appreciated by those skilled in the art that perfusive WO92/1~36 - 22 - PCT/US92/02~7 2~237 matrices comprising smaller particles and requiring concomitantly higher threshold flow rates still will be useful, even when enzyme reactions are run below the perfusion mode. There are several reasons for this.
First, the enzymes themselves may be loaded in the perfusion mode, allowing the column to be loaded in seconds, a fraction of the time required using conventional systems. S~co~A, the increased intraparticle throughpore size substantially reduces pore effects identified with ~ol~ventio~Pl porous matrices. Third, the r~ c~ ratio of interparticle and intraparticle pores, and the substantially reduced intraparticle diffusional pathlengths that characterize these matrices ~irtually eliminate substrate or enzyme inhibition due to solute build up in a st~g~Ant mobile phase.

The invention may be further understood from the following, nonlimiting example.

Calf intestinal alkaline phosphatase (CI-AP) is noncovalently attached to a perfusive reverse phase column (POROS~M RM, PerSeptive Biosystems, Inc., Cambridge, MA, 0.1 x 2 cm, 20 ~m diameter particle, packed at 40 bar). CI-AP, reconstituted to 1 mg/ml with 10 mM tris(hydroxymethyl)aminoethane, pH 8.0, is diluted (1:100) in 10 mM phosphate buffer (PB), pH 7.1, and ng quantities loaded onto the column at a given flow rate, generally between 0.1 ml/min to 0.5 ml~min.
Typically, the column is loaded in about 10 seconds.

The enzyme substrate, p-nitrophenylphosphate, (PNPP, Pierce Co., Rockford, IL) ) is dissolved in 10 mM
diethylanolamine, 0.5 mM MgC12, pH 9.~, and loaded at either 0.1 or 1.0 mg/ml. The substrate reaction is ~ WO92/1~36 - 23 - PCT/US92/02~7 .

2~d2237 run as a continuous flow system at a given flow rate snd ~oncentration of enzyme. Product is identified by absorbance at 405 nm.

The column may be stripped by washing with PB, followe~ by a stripping solvent comprising 80%
acetonitrile/20% acetic acid, and re-equilibrat with PB. PNPP can be passed over the column before a second enzyme is loaded, and the effluent tested for residual enzyme activity.

Fig. 3A and 38 represent plots of product formation versus enzyme concentration for 10 different flow rates and two different substrate concentrations (O.1 mg/ml and 1.0 mg/ml in 3A and 3B, respectively). A lOofold increase in substrate concentration appears to have no effect on the system. As expected, slower flow rates increase sensitivity of the assay. However, while preferred substrate reaction flow rates (e.g., less than about 0.2 ml/min) are below the threshold perfusion flow rate for the size particle used in this experiment (20 ~m), the linearity of the curves at both high and low substrate concentrations (R=l.000-0.990) provides dramatic evidence that no significant substrate or product inhibition occurs in this system, even when high substrate concentrations are provided as a con~inuous flow. As stated above, in the preferred embodiment of this invention, perfusive particles for use in IMERs have a diameter on the order of about 100-1000 ~m, substantially reducing the threshold flow rate for intraparticle convective flow.

In Fig. 3C, 1650 column volumes have been passed through the system (containing 25 ng CI-AP) at a perfusive flow rate (3600 cm/hr) between a first and WO92/1~ ~ PCT/US92/0~7 2~ 02237 second substrate conversion series (curves 1 and 2, respectively, using 1 mg/ml PNPP, and a range of flow ra~es). As is evident from the data plotted in Fig.
3C, no significant leaching of the noncovalently im~obilized enzyme has occurred in the interim. In yet another experiment, a second column capable of ~mmobiliz~ng enzyme is placed $n t~ndem with the first, and the effluent from the fir~t column passed o~er it to test for leaching enzyme. No significant substr~te conversion can be identified in thls econd colu~n when substrate is subseguently passed through this second column, indicat~ng that substantially no enzyme has been leAched from the first column..

The invention may be e~bodied ~n other specifie forms without departing from the spirit and essential characteristics thereof. Accordingly, the invention is to be defined not by the preceding description, which is intended as illustrative, but by the claims that follow.

Claims (25)

1. A method for performing an enzymatic reaction comprising the steps of:

(A) providing a matrix comprising a multiplicity of packed particles defining therewithin throughpores and substrate interactive surface regions comprising immobilized enzyme having an enzyme activity of at least 100 units per gram of matrix, wherein one unit comprises one micromole of substrate converted per minute; and (B) passing a solution of substrate, reactive with the immobilized enzyme, through said matrix at a velocity sufficient to induce convective fluid flow through said throughpores within said particles greater than the rate of substrate diffusion through said throughpores, and to produce a Peclet number in said throughpores greater than 1, said velocity being sufficient to allow catalysis of said substrate; and wherein the Peclet number is defined as VL/D, wherein v is the convective velocity through the throughpore, L is its length, and D is the diffusivity of a solute through the throughpore.

2. A method for performing an enzymatic reaction comprising the steps of:

(A) providing a matrix defining:

interconnected first and second throughpore sets dimensioned to allow convective fluid flow through both said first and second throughpore sets, each of the interconnected throughpore sets comprising a multiplicity of throughpores for channeling through the matrix an enzyme solution and a substrate solution reactive with the enzyme, and surface regions in fluid communication with the members of the second throughpore set and capable of immobilizing an enzyme, (B) passing a solution of an enzyme through said matrix to load sufficient enzyme onto at least a portion of said surface regions to provide an enzyme activity of at least 100 units per gram of matrix, wherein one unit comprises one micromole of substrate converted per minute; and (C) passing a solution of a substrate reactive with said immobilized enzyme through said matrix at a velocity sufficient to produce a Peclet number in the throughpores of the second throughpore set greater than 1, to allow catalysis of the substrate by the immobilized enzyme; and wherein the Peclet number is defined as VL/D, wherein V is the convective velocity through the throughpore, L is its length, and D is the diffusivity of a solute through the throughpore.

3. The method of claim 2 wherein said enzyme solution or said substrate solution is passed through said matrix at a fluid flow rate to produce:

convective fluid flow through both throughpore sets, a convective fluid flow velocity through said first throughpore set greater than the fluid flow velocity through the second throughpore set, and a convective fluid flow velocity through said second throughpore set greater than the diffusive flow rate of said enzyme or said substrate within the members of said second throughpore set.

4. The method of claim 1 or 2 wherein the matrix defines a multiplicity of subpores comprising said surface regions.

5. The method of claim 3 wherein step B or C is conducted by passing said solution through said matrix at a bed velocity greater than about 300 cm/hr.

6. The method of claim 3 wherein step B or C is conducted by passing said solution through said matrix at a bed velocity greater than about 1000 cm/hr.

7. The method of claim 3 wherein step C is conducted by passing said solution through said matrix at a bed velocity greater than about 30 cm/hr.

8. The method of claim 2 wherein the first throughpore set is defined by packed particles having a mean diameter of at least about 8 µm, and said second throughpore set comprises throughpores within the particles having a mean diameter greater than about 2000.ANG.

9. The method of claim 8 wherein the ratio of the mean diameter of the particles to the mean diameter of the second throughpores is less than about 70.

10. The method of claim 9 wherein the ratio of the mean particle diameter to the mean diameter of the second throughpores is less than about 50.

11. The method of claim 8 wherein said particles have a mean diameter of at least about 100 µm, and said second pore set comprises throughpores within the particles having a mean diameter greater than about 2 µm.

12. The method of claim 4 wherein said subpores have a mean diameter less than about 700.ANG..

13. The method of claim 1 or 2 wherein said surface regions comprise plural different enzymes immobilized thereon.

14. The method of claim 13 wherein said different enzymes are immobilized in successive zones on said matrix.

15. The method of claim 2 wherein the Peclet number in the throughpores of the second throughpore set is greater than 5.

16. An enzyme reactor comprising, a rigid matrix defining interconnected first and second throughpore sets, each of which comprise a multiplicity of throughpores for channeling through said matrix a substrate solution reactive with an enzyme, and surface regions comprising immobilized enzyme sufficient to provide at least 100 units per gram of matrix, wherein one unit comprises one micromole of substrate converted per minute and in fluid communication with the members of the second throughpore set, the relative dimensions of the members of said first and second throughpores sets and said surface regions being fixed to permit, when said substrate solution is passed through said matrix at a preselected velocity, convective fluid flow through both throughpore sets, a convective fluid flow velocity through said first throughpore set greater than the fluid flow velocity through the second throughpore set, and a convective fluid flow velocity through said second throughpore set greater than the diffusive flow rate of said substrate within the members of said second throughpore set, thereby to produce a Peclet number in the throughpores of said second throughpore set greater than 1; and wherein the Peclet number is defined as VL/D, wherein V is the convective velocity through the throughpore, L is its length, and D is the diffusivity of a solute through the throughpore.

17. The enzyme reactor of claim 16 wherein said surface regions comprise plural different enzymes immobilized thereon.

18. The enzyme reactor of claim 17 wherein said plural different enzymes are immobilized in successive zones on said matrix.

19. the enzyme reactor of claim 16 wherein the matrix comprises a multiplicity of interfacing particles defining an interstitial volume constituting said first throughpore set, each of said particles defining:

a plurality of throughpores comprising said second throughpore set, and a plurality of pores closed at one end and comprising said surface regions.

20. The enzyme reactor of claim 19 wherein said particles have a mean diameter of at least about 50 µm and a ratio of particle diameter to intraparticle throughpore diameter of less than 70.

21. The enzyme reactor of claim 20 wherein said particles have a mean diameter of at least about 100 µm.

22. The enzyme reactor of claim 20 or 21 wherein said ratio is less than 50.

23. The enzyme reactor of claim 19 wherein said particles define a plurality of anisotropic throughpores.

24. The enzyme reactor of claim 19 wherein said particles comprise small interadhered particles.

25. The enzyme reactor of claim 19 wherein the ration of the convective flow velocity through said first throughpore set to the convective flow velocity through said second throughpore set is within the range of 10:1 to 100:1.