WO2004113924A2 - Method and system for generating and capturing peptides on a target surface - Google Patents

Abstract

A method for generating and capturing peptides on a target surface. The method comprises electrophoresing a protein on a high resolution reversibly polymerized polyacrylamide gel and vacuum depositing the electrophoresed protein on a digestion membrane for digestion into peptides. The digestion membrane is a substantially chemically inert membrane having a proteolytic enzyme immobilized thereon. The method further comprises vacuum depositing the generated peptides on the target surface. A target surface for use in a mass spectrometer comprises a porous plate that does not ionize or volatize in the charge conditions of the mass spectrometer. In its preferred form, the porous plate has a coating of organic matrix and nitrocellulose. Further, the invention relates to a high vacuum stage for procuring biopolymers attached to a membrane and depositing them on a porous plate target surface.

Description

METHOD AND SYSTEM FOR GENERATING AND CAPTURING PEPTIDES ON A

TARGET SURFACE

FIELD OF THE INVENTION

[0001] In at least one of its aspects, the present invention relates to a method and a system for generating and capturing peptides on a target surface to enable their subsequent analysis by mass spectrometry (MS). Other aspects of the present invention relate to various devices incorporated within and/or processes used in the present system.

DESCRIPTION OF THE PRIOR ART

[0002] The ability to identify and analyze large numbers of proteins in complex biological mixtures has developed in recent years by improvements in MS instrumentation and the translation of genomically encoded information into amino acid sequences. For a basic review of this area see Liebler [1]. The ability to accurately mass and sequence peptides derived by proteolytic digestion of proteins has enabled identification thereof and permitted more complex studies of their structure and function.

[0003] One of the current challenges to achieving higher throughput proteomics analysis is the inability to rapidly separate complex biological mixtures and achieve peptide samples that are compatible with MS. Various automation techniques have been employed to speed the process of protein separation, peptide generation and cleanup prior to MS.

[0004] An important protein separation technique that is widely used is polyacrylamide gel electrophoresis (PAGE).

[0005] Isoelectric focusing is one electrophoretic technique used to separate proteins based on their intrinsic amino acid charge properties. By establishing a pH gradient throughout a polyacrylamide matrix, proteins will migrate in an electric field until their net charges are neutralized. Such separations also provide a means of identifying a protein based on its theoretically calculated isoelectric point.

[0006] A more commonly used electrophoretic technique involves using a polyacrylamide gel matrix as a molecular sieve to separate proteins based on their individual molecular weights [2]. This separation technique relies on the use of highly charged detergents like sodium dodecyl sulfate (SDS) that bind to proteins, and negate any charge differences resulting from differences in amino acid compositions or post-translational modifications [3]. A uniform mass to charge state of individual proteins in this system permits separation of proteins based solely on their molecular weights.

[0007] The combined use of isoelectric focusing of proteins in a first dimension and

SDS-PAGE in a second dimension has provided scientists with a separation tool for resolving complex protein mixtures [4].

[0008] The ability to separate complex biological mixtures by electrophoresis has permitted the immunochemical identification and sequence analysis of proteins. The electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose paper [5] routinely permits antigenic determinants to be detected by corresponding antibodies.

[0009] The transfer of proteins from polyacrylamide gels to nitrocellulose paper is commonly achieved by electrophoretic transfer. The electrophoretic transfer of high molecular weight proteins from polyacrylamide gels to nitrocellulose paper is inefficient because their relatively slow electrophoretic migration rates become even slower as ionic detergent/protein complexes dissociate in detergent- free buffers. To enhance the transfer of high molecular weight proteins out of polyacrylamide gels, proteases have been used during electrophoretic transfer [6]. Reversible cross-linkers have also been used to increase the porosity of the polyacrylamide gel matrix during electrophoretic transfer and improve the recovery of proteins on nitrocellulose paper [7]. While electrophoretic transfer is most commonly used to transfer proteins or peptides from polyacrylamide gels to nitrocellulose, other methods, including vacuum transfer of proteins from polyacrylamide gels to nitrocellulose, have been used [8]. However, vacuum transfer of proteins from polyacrylamide gels to nitrocellulose is not very efficient since most proteins remain trapped in low porosity gels.

[0010] The efficient generation, recovery and analysis of peptides by mass fingerprinting and/or fragmentation ion sequencing is fundamental to the identification of large numbers of proteins by MS. Proteolytic cleavage of proteins at specific amino acids (for example, C- teπninal lysine and arginine residues in the case of trypsin) produces a series of uniquely sized peptides that can be accurately measured using a mass spectrometer. Source ion fragmentation or collisionally induced fragmentation of peptides can provide sufficient information to enable amino acid sequence assignments.

[0011] Peptide mass fingerprint information coupled with ion fragmentation data permit searches of sequence databases for corresponding matches that can unequivocally identify proteins. The ability to procure large numbers of peptide samples from proteins for analysis by MS is important to facilitate an understanding of cellular gene expression, the etiology of diseases and the identification of potential pharmacological targets. To date, gel electrophoresis remains the most common method of protein separation and analysis performed on biological samples. The in gel digestion of proteins to generate peptides for MS analysis following SDS- PAGE is a lengthy procedure. It normally involves gel staining, imaging, band or spot excision, washing of gel slices, subsequent enzymatic digestion with a protease of relatively low molecular weight to enable it to diffuse into the gel bed, extraction of the resulting peptides and sample cleanup requiring removal of salts [9].

[0012] There have been previous attempts to form peptide replicas from proteins separated by PAGE in order to enable their analysis by MS [10] [11]. Typically, nitrocellulose and polyvinylidene difluoride (PVDF) membranes are used to form these replicas. Identification of individual proteins using these prior art methodologies has been hampered by poor peptide ionization signals during Matrix Assisted Laser Desorption Ionization (MALDI) MS. Despite extremely high electrophoretic protein sample loads, it has only been possible to identify the most abundant sample constituents [12]. Poor recoveries and resolution of peptides on the replica surface, as well as signal to noise problems in mass spectral accumulation, have slowed the adoption of this technology as a high throughput proteomics application.

[0013] United States patent 6,221,626 [Bienvenut et al. (Bienvenut)] teaches the use of a hydrophilic membrane containing an immobilized protease such as trypsin between an electrophoretic gel and a hydrophobic membrane to form a peptide capture surface in an electroblotting "sandwich". Bienvenut teaches that PVDF, nylon and nitrocellulose are suitable for the hydrophobic membrane and that an active carbonyl-modified or carboxyl-modified PVDF is suitable for the hydrophilic membrane. Polypeptides separated on the gel by electrophoresis are electroblotted from the gel through the hydrophilic membrane where they are cleaved by the protease into fragments, and the fragments are collected on the hydrophobic membrane where they are identified, such as by MALDI-TOF (Time of Flight) MS analysis. Bienvenut also teaches that the hydrophilic membrane may be provided with functional groups to which the protease is immobilized by covalent bonding. Further, Bievenut teaches that the protease can be immobilized on the membrane by affinity bonding. It specifically teaches that the protease can be covalently attached to avidin or streptavidin and the resultant conjugate attached to a biotinylated membrane by affinity bonding between avidin/streptavidin and biotin. It also teaches that in the alternative, avidin or streptavidin can be attached to the membrane and the protease can be reacted to provide biotinyl terminations for reaction with a membrane to which avidin or streptavidin has been attached.

[0014] Attachment of enzymes to polylysine side chains can occur by a number of chemical conjugation methods. While bi-functional cross-linkers like glutaraldehyde can be readily used to attach enzymes to a polylysine-coated membrane surface [13], the ability to control this conjugation reaction is rather limited. Inappropriate cross-links (lysine to lysine) reduce the number of available side chains that are available to conjugate the enzyme. Glutaraldehyde inactivation of the enzyme can also become a problem particularly if the strategy does not involve pre-activation of the surface with the cross-linker followed by its removal.

[0015] The use of an electrotransfer apparatus for generating and capturing peptides on a membrane surface suffers several drawbacks. Replicas formed by proteolytic digestion and electrotransfer of proteins from gels show reduced recoveries and a loss of resolution of peptides on the target membrane surface. This problem is particularly evident for high molecular weight proteins that electrophorese slowly out of gels and therefore do not proteolyze efficiently. Various attempts have been made to rectify these problems, such as the use of square wave alternating voltages [14] and the inclusion of SDS in alkaline transfer buffers or the incorporation of trypsin into the gel matrix [10].

[0016] A more fundamental problem encountered with electrotransfer replicas is the differential charge properties of the resulting peptides owing to their particular amino acid compositions. Basic peptides tend to electrophorese cathodally while acidic peptides migrate toward the anode. Neutral hydrophilic peptides fail to migrate in an electric field, as do hydrophobic peptides that additionally suffer reduced solubilities in aqueous buffers. The differential chemical properties of peptides produced by proteolytic digestion of proteins cannot always be readily accomodated by buffers used to electrophoretically transfer them to a replica surface.

[0017] Reversible cross-linkers have been used to increase the porosity of polyacrylamide gel matrices during electrotransfer and improve the recovery of proteins on nitrocellulose paper [7]. Several types of reversibly cross-linked polyacrylamide gel systems are known [7] [15] [16] [17] [18] [19]. Some of these systems, however, are not compatible with a replica system employing a MALDI target surface pre-coated with an organic matrix. For example, the conditions necessary to effect the hydrolysis of cross-links, particularly the alkaline buffers used to dissolve (l,2-dihydroxyethylene)bisacrylamide (DHEBA) cross-linked gels [15] [19], present a problem for a replica system employing a pre-coated MALDI target surface. Most organic acid matrices would be extensively solubilized from pre-coated MALDI target surfaces under the alkaline conditions used to effect depolymerization of DHEBA cross-linked gels during replica formation.

[0018] Subsequent improvements in the electrophoretic resolution of proteins on more advanced depolymerizing gel systems or the use of tissue sections for molecular localization studies will demand a higher degree of resolution and recovery of peptides on the target surface. In the electrotransfer replica, small proteins and peptides are often lost through high porosity capture membranes (>0.05 μM) especially when high electric field strengths are used over extended periods of time. This problem is exacerbated when membrane-binding sites become saturated with peptides arising from proteolytic digestion of high abundance proteins. Reducing membrane porosity increases resistance across the electrotransfer replica leading to increased heating. Likewise, increases in field voltages often result in increased current leakage through channels that bypass the gel.

[0019] There is therefore a need for a system that is an alternate to the electrophoretic transfer method of peptide replica membrane formation. SUMMARY OF THE INVENTION

[0020] It is an object of the present invention to provide a method and system for the efficient generation of peptide replicas from protein electrophoretic gels that mitigates one or more of the bias for certain classes of peptides, poor recoveries and resolution of peptides on the target replica surface and the accompanying signal to noise problems in mass spectral accumulation inherent in prior art methods and systems.

[0021] In one of its aspects, the present invention provides a method and system for the high resolution separation of proteins on a polyacrylamide gel, their transfer to a proteolytic membrane and the transfer of the resulting peptides to a novel target surface for subsequent analysis by a MALDI mass spectrometer.

[0022] The components of the current invention include (i) a reversibly cross-linked electrophoretic gel system and an associated buffer system and running conditions; (ii) a proteolytic replica membrane; (iii) a porous target replica surface compatible with MS; and (iv) a high vacuum device for procuring peptide replicas.

[0023] The reversibly cross-linked high resolution N,N-bisacrylylcystamine (BAG) horizontal polyacrylamide gel of the present invention is a reduced dimension gel, which reduces scan time of the peptide replica in the mass spectrometer.

[0024] The associated buffer system uses anodal and cathodal buffer wick compositions having an elevated salt , concentration and glycerol, which, in combination with a reduced running current overcome the problems associated with the high voltages typically required for such a gel.

[0025] The proteolytic membrane of the present invention is formed from a substantially chemically inert polymer, preferably having a minimal membrane caliper and uniform distribution and size of pores. Preferably, the polymer is polycarbonate or polyester.

[0026] The target surface of the present invention is a novel porous plate coated with matrix. [0027] The formation of the proteolytic replica membrane and target replica surface is achieved through vacuum deposition using a high vacuum stage disclosed as a component of the present invention.

[0028] While the target surface and vacuum deposition technology of the present invention is particularly applicable to protein separations involving PAGE, there may be additional applications, including the molecular imaging of cells, tissue sections or other complex macromolecular structures where a conversion of analytes to a more volatile and readily ionizable state is subsequently required for MS analysis.

BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Embodiments of the present invention will be described with reference to the accompanying drawings, wherein like reference numerals denote like parts, and in which:

Figure 1 is a photograph of a high vacuum stage for generating peptide replicas on porous plate MALDI target surfaces;

Figure 2 an autoradiograph of a 0.05 μm nitrocellulose membrane subjected to the deposition of S-Methionine/cysteine radiolabelled COS cell protein extract from a Tricine- BAC mini-gel with an interposed nitrocellulose membrane with or without trypsin immobilized thereon;

Figure 3 shows autoradiographs of identical samples of ³⁵S-Methionine/cysteine radiolabelled COS cell protein extract electrophoresed on a Tricine-BAC mini-gel and tryptic peptides derived thereon and recovered on a replica 0.05 μm nitrocellulose membrane;

Figure 4 is an autoradiograph of a 0.05 μm nitrocellulose membrane of tryptic peptides derived from a ³⁵S-Methionine/cysteine radiolabelled COS cell protein extract electrophoresed on a Tricine-BAC Phast-gel and vacuum transfered through a proteolytic nitrocellulose membrane;

Figure 5 is a Coomassie Brilliant Blue stained electrophoretic gel of the results of bovine serum albumin (BSA) digestion by biotinylated trypsin derivatized by different ratios of biotin to trypsin; Figure 6 is a Coomassie Brilliant Blue stained electrophoretic gel of the results of BSA digestion by trypsin in the presence of different concentrations of soluble polylysine;

Figure 7 is a chemiluminescent film exposure of biotinylated and non-biotinylated polylysine coated membranes incubated with avidin/biotinylated peroxidase;

Figure 8 illustrates the loss of Coomassie Brilliant Blue stained gelatin homogeneously incorporated in a polyacrylamide slab gel by proteolysis with biotinylated trypsin complexed to polycarbonate membranes by avidin and biotinylated polylysine;

Figure 9 illustrates the digestion of DL-BAPA over a 25 hour period by biotinylated trypsin complexed to biotinylated polylysine-coated polycarbonate membrane disks with avidin;

Figure 10 illustrates the digestion of DL-BAPA over a 25 hour period by trypsin coupled to polylysine-coated polycarbonate membrane disks with sulfo-SANPAH;

Figure 11 illustrates the digestion of DL-BAPA for an initial incubation and when transferred to fresh substrate by trypsin coupled to ^"polylysine-coated polycarbonate membrane disks with sulfo-SANPAH and complexed to polylysine-coated polycarbonate membrane disks using avidin/biotin;

Figure 12 illustrates the proteolytic activity of glutaraldehyde-polymerized trypsin conjugated to polylysine-coated polycarbonate membranes with sulfo-SANPAH;

Figure 13 shows MALDI TOF MS analysis of lμg of BSA deposited on conventional and porous plate target surfaces;

Figure 14 shows MALDI TOF MS analysis of BSA tryptic digests spotted on a solid stainless steel MALDI plate;

Figure 15 shows MALDI TOF MS analysis of BSA tryptic digests vacuum deposited on porous stainless steel MALDI plates pre-coated with nitrocellulose and either - CHCA orHABA; Figure 16 shows MALDI TOF MS analysis of BSA tryptic digests vacuum deposited on porous stainless steel MALDI plates pre-coated with various concentrations of nitrocellulose and α-CHCA; and

Figure 17 shows MALDI TOF MS analysis of various amounts of BSA tryptic digests vacuum deposited on porous stainless steel MALDI plates pre-coated with nitrocellulose and α-CHCA.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0030] For convenience, the description of the preferred embodiments of the present invention will be set out initially with reference to the major components thereof.

REVERSIBLY CROSS-LINKED GEL SYSTEM AND ASSOCIATED BUFFER SYSTEM

[0031] One component of the present invention is a high resolution, reversibly cross- linked gel system and an associated buffer system, which enables proteins to be rapidly transferred out of the depolymerizing polyacrylamide matrix under high vacuum.

[0032] The reversibly cross-linked polyacrylamide gel is formed of N,N - bisacrylylcystamine (BAC) by a modification of the method of Hansen et al. [7]. 1% Tetramethylethylenediamine (TEMED) produces a gel that yields high resolution of proteins, while still being amenable to depolymerization by dithiothreitol (DTT).

[0033] To gain higher resolution of proteins, the discontinuous Tricine gel buffer system developed by Schagger and von Jagow [20] for separating peptides was modified for use both with reversibly cross-linked BAC-polyacrylamide vertical slab mini-gels and Phast gels.

[0034] One way of achieving a greater concentration of proteins in gels and consequently peptides recovered on the replica surface is to reduce the dimensions of the BAC gels used. A smaller gel still capable of producing a high degree of electrophoretic resolution of proteins reduces scan times of target plates in the MALDI TOF mass spectrometer. A Phast gel format competent for formation of peptide replicas was therefore developed. While the application of the Tricine buffer system and BAC-cross-linked gel formulations, designed for the vertical mini- gel to a Phast system format, would have represented a simple solution to this problem considerable modifications were necessary to ensure good electrophoretic separations of proteins. Electrophoresis of a Phast gel, which is composed of an open-faced, horizontal polyacrylamide bed on a Gelbond backing, poses significant technical problems when high resolution BAC gels are formulated. The high voltages used to separate proteins is accompanied by significant electro-osmosis that inevitably leads to current fluctuations and distortion or even burning of gels.

[0035] To overcome the problems associated with a reversibly cross-linked BAC Phast. gel system, anodal and cathodal buffer strip compositions that employ higher than normal salt concentrations and glycerol to act as a stabilizer may be used. Preferably, buffer strips formed from the following compositions are used with the BAC Phast gel system of the present invention: Anode [0.4 M Tris (pH 8.9), 20% Glycerol, 3% (w/v) LEF agarose] and Cathode [0.2M Tris, 0.2M Tricine (pH 8.4), 20% (v/v) Glycerol, 3% (w/v) D3F agarose, 0.1% SDS)].

[0036] Additionally, a reduced running current improves resolution of proteins on the reversibly cross-linked BAC Phast gel system.

[0037] The following gel running conditions yielded an acceptable resolution of proteins on the reversibly cross-linked BAC Phast gel of the present invention: a pre-electrophoresis step was conducted for 5 amperes-volts-hours (aVh) with the apparatus set at 70 volts (V), 25 milliamperes (mA) and 3 watts (W) per gel. The samples were then programmed to be applied from 5 aVh to 15 aVh at the same voltage, current and power settings. The gels were run for 65 aVh and when the power fell below 0.8 W, the voltage was increased to 100 V (reading 18 mA and 1.2-1.5W/gel) The gels were run until the dye front reached the anode buffer strip (about 190-200 aVh total). Running time was about 1.5 hours.

PROTEOLYTIC MEMBRANE

[0038] A second component of the present invention is an improved proteolytic membrane interposed between the reversibly cross-linked BAC gel and the MS target surface.

[0039] This improved proteolytic membrane combines a variety of desirable properties that are difficult to achieve with nitrocellulose or PVDF membranes. [0040] This interposed proteolytic membrane may be formed from a substantially chemically inert membrane with a small caliper (thickness), low peptide binding, uniformity of pore size and distribution, and durability. Suitable membranes include those made from polycarbonate and polyester. The calipers of commercially-available polycarbonate and polyester membranes (10 μm) are less than a tenth that of nitrocellulose membranes (125-150 μm ) and the pores are much more uniform owing to the track-etch method used to form them [21]. Polycarbonate and polyester are much less reactive to a variety of chemical and biological agents than nitrocellulose or PVDF and while this is a desirable attribute for reducing non- specific surface adsorption of proteins and peptides, it can pose a logistic problem for the attachment of biological enzymes like trypsin.

[0041] Preferably, the membrane used to form the proteolytic membrane is a polycarbonate membrane and most preferably, a polycarbonate membrane having a O.Olμm - 10 μm average pore diameter. The polycarbonate membrane may be polyvinypyrrolidone (PVP) treated or polyvinypyrrolidone-free (PVPF). Preferably, the proteolytic enzyme is trypsin.

[0042] To the knowledge of the present inventor, chemically derivatized polycarbonate membranes are not commercially available, but there have been previous attempts to attach biomolecules to use them as biosensors [13]. Any suitable method may be used to attach the enzyme to the membrane. Preferably, a covalently attached polymer acts as a backbone for enzyme attachment. Preferably, the polymer backbone is polylysine. Polylysine can be readily derivatized by hetero- or homo-bifunctional crosslinkers. Depending on its dispersion index, a single polylysine molecule can contain tens of thousands of lysine side chains that can be readily derivatized.

[0043] One suitable coupling method is to use affinity bonding between avidin and biotin. Trypsin can be biotinylated and it will remain proteolytically active [22]. A suitable trypsin is bovine pancreatic trypsin, 3X recrystallized from Bioshop, catalogue # TRP003. A multivalent, high affinity, biotin-binding protein, avidin (a suitable avidin is catalogue # 9275, Sigma Chemical Co., St. Louis, MO), can be used to generate a large multimeric protein complex. A potent proteolytic membrane can subsequently be assembled on a biotinylated, polylysine-coated polycarbonate surface. [0044] The ability to biotinylate trypsin without a subsequent loss in enzymatic activity does not require prior protection of the active site with soybean trypsin inhibitor as was previously thought [22]. Using the HABA/Avidin detection reagent (catalogue # H2153, Sigma Chemical Co., St. Louis, MO) it was determined that 4-5 out of the 6 lysines in bovine pancreatic trypsin were derivatized _' at a high biotin to protein weight ratio (i.e., 1:1). This high level of derivatization did not inactivate the enzyme, but reproducibly increased enzymatic activity presumeably by blocking autolysis at lysines that had subsequently become biotinylated.

[0045] In addition, recognition by trypsin of the C-terminal lysine peptide bond structure in polylysine does not appear to competitively inhibit the proteolysis of protein substrates

[0046] To optimize (increase) the amount of biotinylated trypsin incorporated into the polylysine backbone, tests were performed to optimize conditions for biotinylation of the polylysine side chains and to establish the correct molar ratios of avidin and biotinylated trypsin necessary to build a potent proteolytic membrane. Biotinylated tr psm-immobilized polycarbonate PVP membranes possess slightly higher proteolytic activity than similarly treated polycarbonate PVPF membranes. Various coupling conditions were assessed in order to optimize the proteolytic activities of the two types of membranes (PVP and PVPF). These included varying the temperature for trypsin complexation with avidin and biotinylated polylysine, altering the coupling pH (1 to 5), and adjusting the avidin:biotinylated-trypsin molar ratios (10 to 0.01). The preferred coupling conditions are pre-formation of the avidin/biotinylated trypsin complex at 4°C for 24 hours using a mole ratio of 0.8:1 (avidi biόtinylated trypsin) followed by incubation of the complex with biotinylated polylysine- coated membranes. These conditions are likely ideal since trypsin incorporation into the complex is maximized while a sufficient number of vacant bϊotin-binding sites on avidin are maintained to allow the subsequent interaction of the complex with the biotinylated polylysine- coated polycarbonate membrane surface.

[0047] Another mild coupling technology suitable for immobilizing trypsin on polylysine-coated polycarbonate membranes is to use a heterobifunctional cross-linker. Ideally, the reactive groups on the cross-linker should be capable of independent activation to allow the sequential derivatization of polylysine side chains followed by covalent attachment of trypsin. A suitable cross-linker would possess both a chemical and a photo-active group. Heterobifunctional, photosensitive cross-linking agents like sulfosuccinimidyl 6-[4 -azido-2 - nitrophenylamino]hexanoate (sulfo-SANPAH) have previously been used to cross-link polypeptide hormones to their respective membrane receptors [23]. A suitable sulfo-SANPAH is Pierce Chemical Co., Rockford, LL, catalogue # 22589. The succinimide group reacts readily with free amines such as those present in the side chains in polylysine. The azido group can be activated thereafter with high intensity light to facilitate the covalent attachment of trypsin.

[0048] In the case of polycarbonate proteolytic membranes formed using sulfo-

SANPAH, PVPF membranes are more active than PVP membranes following trypsin conjugation. The basis of this difference appears to be the greater amount of polylysine initially conjugated to PVPF membranes.

[0049] Proteolytic PVPF membranes having trypsin-conjugated by the sulfo-SANPAH method possess significantly higher proteolytic activity than those produced by complexation of trypsin using the avidin/biotin technique. In addition, the proteolytic activity provided by the sulfo-SANPAH conjugation method is more stably associated with the proteolytic membrane (Example 8). A more trypsin-resistant avidin molecule generated either by site-directed mutagenesis or chemical derivatization of sensitive lysine and/or arginine residues might improve the stability of avidin thereby increasing the fidelity of this type of membrane complexation method.

[0050] The proteolytic activity of the proteolytic membrane may also be improved by controlled polymerization of trypsin with glutaraldehyde. Glutaraldehyde modification of lysine residues in trypsin stabilizes the protein from unfolding, reduces autolysis and improves its hydration state [24]. These favourable properties increase its proteolytic activity and have the additional benefit of increasing the stoichiometry of trypsin proteins immobilized on polylysine- coated polycarbonate membranes by the sulfo-SANPAH technique. Preferably, the oligomeric trypsin is obtained by treating 10 mg/ml trypsin in 0.05 M sodium borate (pH 8.0), 20 mM Ca Cl₂, with 0.05% glutaraldehyde for 10 minutes at 4°C with end-over-end mixing. Prior to the formation of any insoluble trypsin, reactions were stopped by addition of Tris/HCl (pH 7.5) to a final concentration of 50mM Tris. Glutaraldehyde derivatized trypsin is then dialyzed against 50 mM sodium phosphate (pH 7.5) ImM CaCl₂ and coupled to a polylysine-coated polycarbonate membrane by the sulfo-SANPAH method.

TARGET REPLICA SURFACE

[0051] A third component of the present invention is a novel target surface for peptide replicas compatible with vacuum deposition and MS.

I • - '

[0052] The target surface comprises a porous plate which will not substantially ionize or volatize in the charge conditions of a mass spectrometer. Preferably, the porous plate is made of stainless steel. The porosity of the plate provides additional surface area for sample spotting relative to conventional solid target surfaces and further permits capture of analytes by filtration of larger sample volumes than on a solid target surface. The porous nature of the target surface permits the ionization of substantially larger amounts of analyte than conventional solid target surfaces. The porous plate is coated with a matrix suitable for MALDI MS. Preferably, the porous plate is coated with organic matrix of α-cyano-4-hydroxycinnamic acid (α-CHCA) or 2- (4-hydroxyphenylazo)-benzoic acid (HABA).

[0053] The relatively poor MALDI ionization signals observed for peptides captured on nitrocellulose or PVDF membranes necessitated the development of a novel porous surface that could efficiently capture tryptic protein digests while maintaining good volatility and ionization potential of the deposited peptides. A prototypic surface that was developed was a sintered stainless steel plate. The target surface of the present invention improves the capture for MS of protein fragments from the proteolytic membrane of the present invention. A custom porous sintered stainless steel plate was manufactured (Applied Porous Technologies, Tarrifville, CT) to specifications that would allow it to be substituted onto any standard Voyager plate holder (Applied Biosystems, Foster City, CA, part # V700698). Two different porosities of plates were produced (0.05 and 0.10 μm). Plates having a 10-15% void area, meaning that 85% of the surface is solid stainless steel are suitable. It will be understood, however, that the surface area available for capture of analyte is a function of the porosity (pore size and pores per area).

[0054] Preferably, the porous plate is pre-coated with nitrocellulose (a suitable material is nitrocellulose membrane obtained from Schleicher and Schuell, GmbH, Dassel, Germany, catalogue # BA85) dissolved in acetone either co-applied with organic matrix or followed by the application of organic matrix.

[0055] Preferably, peptides are deposited onto the porous plate target surface using vacuum deposition. The use of nitrocellulose increases the bonding and stability of the organic matrix on the porous surface and assists in peptide capture. High concentrations of nifrocellulose are however likely responsible for reducing the volatility and subsequent ionization of proteins and peptides from the MALDI target surface. Enough nitrocellulose should be used in pre- coating the porous plate to bind peptides vacuum deposited, but not so much that the nitrocellulose reduces the volatility of the protein fragments in the mass spectrometer. Preferably, 1-2 mg/ml of nitrocellulose is mixed with 20 mg/ml matrix in acetone. Preferably, a coating thickness of about 3.5 ml of this mixture is applied to one side of the porous plate surface.

[0056] A variety of commercially available matrices were tested for their ability to ionize purified peptides from pre-coated porous stainless steel plates in a MALDI TOF mass spectrometer. HABA and α-CHCA obtained from Aldrich Chemical Co. (catalogue # 14,803-2 and # 47,687-0, respectively) were found to be particularly suitable. Higher purity matrices are associated with better ionization of the protein fragments .

[0057] Two types of porous plate coatings were formulated and tested with these organic matrices: Mixed Coating and Sequential Coating.

[0058] To achieve Mixed Coating of the porous plates, 3 ml of nitrocellulose (0.5 mg/ml) and α-CHCA or HABA matrices (20 mg/ml) mixtures are sprayed to cover the entire surface of each plate and these are allowed to dry in a fume hood. A simple automotive airbrush sprayer driven by pressurized nitrogen gas may be used to achieve a uniform coating. Attempts to spin coat plates were less successful at producing uniform applications of nitrocellulose and matrix.

[0059] To achieve Sequential Coating of the porous plates, substantially the entire surface of each plate is initially sprayed with 3 ml of nitrocellulose dissolved in acetone (0.5 mg/ml). The plates are then allowed to dry in a fume hood and are then sprayed with 3 ml of saturated α-CHCA or HABA dissolved in methanol. [0060] Conventionally, deposition of peptides onto a conventional (solid) target surface is achieved by the method of dried droplet co-crystallization. Dried droplet co-crystallization involves the mixing of a quantity of matrix dissolved in solvent with the peptides and spotting the mixture onto a conventional MS target surface usually consisting of a non-ionizable solid stainless steel plate. As such, conventional technology can only establish a sample spot on the solid target that has a minimal area dependent on the surface tension of the liquid droplet originally deposited. Various hydrophobic surfaces (for example Perkin Elmer' s/Sciex's MALDICHIP™ and Bruker's ANCHORCfflP™) have been developed with increased surface^' tensions that act to concentrate analyte but they produce only incremental increases in dried analyte concentrations. As mentioned above, deposition onto the porous plate target surface is preferably by vacuum deposition. Sample spots generated by this porous plate technology have areas that are essentially scribed out by the liquid stream applied to the surface during evacuation. Capillary applications of sample to such evacuated porous surfaces can therefore produce high density arrays that are fully competent for analysis by a variety of devices including a mass spectrometer.

[0061] Of the two organic matrices tested, (HABA and α-CHCA), HABA was consistently more stable and less sensitive to dissolution under a variety of buffer conditions used for sample vacuum deposition. However, α-CHCA (Sigma Chemical Co., St. Louis, MO, catalogue # C-8982) provided stronger peptide ionization signals which more than compensated for its lower stability than HABA on pre-coated porous plates.

[0062] HABA pre-coated porous plates are particularly useful since they can readily capture and ionize large proteins as well as peptide digests. Although BSA ionization signals were lower using nitrocellulose/HABA pre-coated porous stainless steel plates than those obtained with the dried droplet co-crystallization method using a commercial solid stainless steel surface, the sample volumes that could be vacuum deposited onto the former (100-200 μl) were much greater than for a solid target plate (1-2 μl). This was advantageous since some proteins are only sparingly soluble in low ionic strength aqueous buffers and very dilute samples normally need to be concentrated and desalted prior to MS. This concentrating and desalting process (usually performed through reverse phase absorption and elution) often leads to significant losses of proteins and peptides by surface adsorption on microtubes and pipette tips during sample cleanup. The porous plate technology of this invention eliminates the need for reverse phase extraction of peptide digests. It therefore significantly reduces costs and increases sample throughput particularly in robotic operations designed to array proteins or peptides for MS analysis. Another advantage to the porous plate vacuum deposition technology of this invention is its ability to overcome the traditional dependence on organic solvents used both to elute proteins from reverse phase resins and maintain matrix solubility prior to cp-crystallization. Organic solvents dehydrate proteins and this leads to their denaturation thereby reducing the likelihood that good co-crystals (whereby a' uniform lattice structure is obtained through adduction of the matrix molecules with analyte), will form with most organic MALDI matrices. Acidic solutions (1-2% trifluoroacetic acid or trifluoroacetate (TFA) or acetic acid) also denature many proteins, this problem is substantially reduced for samples vacuum deposited in slightly acidic (pH 4-6) buffers on nitrocellulose/matrix pre-coated porous plates. Inclusion of saturating concentrations of matrices within aqueous sample solutions improves protein and peptide ionization signals especially when more neutral buffers are used for vacuum deposition. It is likely the mechanism for this effect is replacement of fresh matrix molecules within the crystal lattice as previously deposited matrix begins to titrate, solubilize and leach from the surface.

HIGH VACUUM STAGE

[0063] A fourth component of the present invention relates to a thermostatically regulated high vacuum stage.

[0064] Thus, with reference to Figure 1, there is illustrated a thermostatically regulated high vacuum stage 2 for procuring peptide replicas. High vacuum stage 2 enables the efficient transfer, of proteins from a depolymerizing gel matrix to a proteolytic membrane and deposit of the resulting peptides on a porous plate target surface that will undergo MALDI MS. Stage 2 provides a vacuum of at least 30 inches of Hg at a porous plate surface following overlay with a proteolytic membrane and gel. This vacuum is applied to enable sufficient buffer flow through the gel matrix. A reduced counter vacuum pressure (not exceeding 3 inches Hg) above the porous plate surface helps form a good gasket seal around the gel and reduces air leaks through the gasket into the high vacuum region. Stage 2 also has an acrylic cover 6 capable of withstanding such pressures while enabling some visibility of buffer wick overlays during the replication process to ensure they do not become dry. Within stage 2, beneath each porous plate 4a and 4b there is a lower chamber. The lower chamber provides a large surface area for vacuum exposure of porous plate 4.

[0065] Preferably, stage 2 is thermostatically regulated to provide an ideal temperature for tryptic activity (37-39°C). While a Pelletier controlled heating device would be ideal for this purpose, an aquarium gravel filled dry bath heater 8 can be used to mount a machined aluminum stage 2 in.

[0066] In a preferred embodiment, stage 2 houses two pre-coated porous plates 4a and 4b and a high vacuum shutoff valve 10 enables the use of just one plate when necessary. Clearly, however, the system could be adapted to house a single or multiple porous plates. While the specific dimensions of stage 2 can vary, an embodiment of stage 2 was constructed as follows.

[0067] The surface of a ^ΛA inch aluminum plate (9.5 cm x 20 cm) was routed with a high torque grinder (Craftsman, model #315.25840) mounted on a robotic machining device (minirobo HI, Super-Tech, Phoemx & Associates, AZ) to provide two housings for porous plates 4a and 4b. Two octagonal shapes (5.8 cm x 5.7 cm with 0.8 cm beveled corners) corresponding to the outside dimensions of the custom manufactured porous plates 4a and 4b were cut to a depth of 0.15 cm. A lower chamber was routed down an additional 0.1 cm centered within 0.2 cm of the original surface cut. This provided a ledge for the 0.15 cm thick porous plates 4a and 4b to rest flush with the top surface of stage 2. A hole was tapped to accommodate a ³/s inch brass pipe elbow fitting n the middle of the lower chamber and a ^lA inch copper pipe was plumbed using compression fittings to provide vacuum to the lower chamber. The two vacuum lines from stage 2 were plumbed to two brass ball valve shutoff fittings using compression fittings. The ball valves were plumbed to a common ^lA inch copper pipe and a particulate air filter 12 (Princess Auto, Mississauga, ON) was installed to protect the vacuum pump from any solids that might escape the porous plate 4. The line was connected to a larger 1 inch copper pipe that contained a cold finger 14 (model CC-60 Cryocool, Neslabs) to act as a condenser for volatile fluids. A 1/16 inch thick silicone gasket was cut to the dimensions of the stage 2 (9.5 cm x 20 cm) and two square windows (4 cm X 3.5 cm) were removed to allow buffer wicks to be inserted above the gel(s). These windows were centered to match the positions of the two vacuum chambers and their slightly smaller dimensions than porous plates 4a and 4b enabled the gasket to form a good seal around the overlaid gel(s). Acrylic cover 6 was constructed from 5/16 inch thick acrylic plastic and it was connected to a diaphragm pump (KNF Neuberger) and valved to provide a vacuum pressure of no more than 3 inches of Hg in the upper chamber. While this upper vacuum chamber helps to reduce any air leakage through windows cut in the silicone sealing gasket, it is not essential to effecting vacuum transfer of proteins to a proteolytic membrane and subsequently to a target surface. In another, preferred embodiment, the lid was replaced with an acrylic buffer reservoir chamber containing an adhesive, Vs inch thick foam sealing gasket attached to the bottom or the reservoir. Two windows corresponding to the dimensions of the sealing gasket were centred and cut through the acrylic reservoir base to provide a continual exposure of the top gel surface to depolymerization buffer. The reservoir was C-clamped to stage 2 after the vacuum seal was formed around the gels by the silicone gasket. The reservoir was subsequently filled with 500 ml of DTT-containing depolymerization/digestion buffer. This reservoir eliminated the need for buffer wicks which provided less than optimal hydration of the gel surface during longer term vacuum transfers.

[0068] While the system of this invention is unique in its application to proteins separated by PAGE and the analysis of peptides generated thereby, it may be used in other applications relating to the analysis of other biological polymers (eg., carbohydrates, nucleic acids), or sample targets conducive to replica technology (eg., intact cells, tissues or other biological material or non-biological materials) would similarly benefit from the current technology. The current system is therefore readily applicable to any target where the juxtaposition relationship between various molecular entities must be conserved. In addition, the porous plate technology of the present invention is readily applicable to the high throughput analysis of a variety of analytes by MS (including small molecules such as pharmaceuticals or other biological or non-biological molecules of interest).

[0069] Embodiments of the present invention will be further described with reference to the following Examples which are provided for illustrative purposes and should not be used to construe or limit the scope of the invention.

EXAMPLE 1 - Tricine-BAC Mini-Gel [0070] A 1 mm thick mini-gel containing 2 ml 16.5% acrylamide/BAC, 2 ml 10% acrylamide/BAC spacer gel and 1.5 ml 4% acrylamide/Bis stacking gel was constructed as in table 1.

[0071] To prevent gel slippage during electrophoresis, 5 inch hydfated dialysis tubing was sliced open and attached to the bottoms of glass plates with waterproof tape.

[0072] While pre-electrophoresis of the separating and spacer gels did not significantly improve the resolution of proteins, when it was performed, it was conducted for 12 hours at 25

V.

[0073] To avoid depolymerization of gels by mercaptans used to reduce proteins in sample buffer prior to electrophoresis, samples were treated with a 10 fold molar excess of iodoacetamide prior to electrophoresis. Samples were reduced with 0.5 mM DTT in 10% SDS, 50 mM Tris/HCl (pH 6.8), 10% glycerol, 0.01% bromophenol blue for a minimum of 1 hour at room temperature then fresh iodoacetamide (500 mM stock solution) was added to a final concentration of 5 mM. Alkylation of samples was carried out for 1 h in the dark. Electrophoresis was carried out at 40 mA per gel for approximately 3 h or until the dye front exited the bottom of the gel. Table 1 shows the composition of tricine/BAC polyacrylamide gel solutions used for preparing reversibly cross-linked mini-gels and Phast system gels.

Table 1

Separating Gel 16.50 %

5 ml 10 ml

Acrylamide (25%) 6C (23.4 g: 1.5 g BAC) 3.300 6.600

Tris (Base) 0.605 g 1.211 g

SDS 0.05 g 0.100 g

Glycerol 0.483 0.967

HC1 0.200 0.400

APS (0.08% Final) 0.004 0.008

TEMED (1% Final) 0.050 0.100

H₂0 0.963 1.925 Spacer Gel 10.00 %

5 ml 10 ml

Acrylamide (25%) 3C (24.25 g: 0.75 g BAC) 2.000 4.000

Tris (Base) 0.605 g 1.211 g

SDS 0.050 g 0.1 g

Glycerol - -

HCl 0.200 0.400

APS (0.08% Final) 0.004 0.008

TEMED (1% Final) 0.050 0.100

H₂0 2.746 5.492

Stacking Gel 4 %

5 ml 10 ml

Acrylamide 49.5T-3C (48 g : 1.5 g BIS) 0.404 0.808

Tris (Base) 0.605 g 1.211 g

SDS 0.05 g 0.100 g

H₂0 4.541. 9.082

APS (10% Final) 0.050 0.100

TEMED (0.1% Final) 0.005 0.010

EXAMPLE 2 - Tricine-BAC Mini Gel Reϋlica

[0074] A test vacuum transfer set-up was used to test the efficacy of a Tricine-BAC Gel at efficiently releasing proteins to a proteolytic membrane surface. This test vacuum transfer setup was developed through modification of a thermostated gel dryer (model 240, Hoefer Scientific Instruments, San Francisco, CA). The adjustable thermostat in the instrument was set to 37°C to provide an optimal temperature for tryptic activity. Normally only 1-2 gels were set up for replication on one gel dryer. A 10 cm x 10 cm piece of 0.45 μm pore size nitrocellulose membrane (Biotrace-Pall Corporation, Ann Arbor MI) was hydrated with distilled water and incubated for 1 hour at 4°C with 25 ml of 0.1 mg/ml bovine pancreatic trypsin (ICN) in TBS. The trypsin membrane was washed 4 times with 25 ml of distilled water and overlaid onto a hydrated, 10 cm x 10 cm, 0.05 μm pore size nitrocellulose membrane (Schleicher & Schuell GmbH, Dassel, Germany) positioned on the smooth surface side of a sintered plastic sheet (Porex Technologies Inc., Fairburn, GA). The gel was overlaid onto the trypsin sheet and a 1/32 inch silcone gasket cut to' provide a window at least 3-5 mm smaller than the dimensions of the gel was used to form a tight seal between the upper and lower vacuum chambers of the modified gel dryer. This setup provided a focused flow of digestion buffer (10 mM DTT, 0.1 mM CaCli, in TBS) through the pores of the polyacrylamide gel from wicks consisting of a single piece of blotting paper (Schleicher & Schuell GmbH, Dassel, Germany) and a humidifier wick (Bionaire) overlaid onto the top surface of the gel. The lower and upper chambers were respectively evacuated (30 μm lower, 3 μm upper) using a rotary vane vacuum, pump (Leybold Vakuum GmbH, Cologne, Germany) connected to a lyophilizer cold trap (Labconco, Kansas City, MO).

[0075] The ability to generate and capture peptides from electrophoretically resolved, reversibly polymerized BAC protein gels was assessed using ³⁵S-methionine/cysteine labelled COS cell extracts. Cells were grown to confluence in a T70 flask (Corning) and labelled with 100 μCi/ml methiomne/cysteine translabel (Perkin Elmer, Wellesley, MA) in 10 ml methionine/cysteine free minimal essential media for 2 hours at 37°C. As an autoradiograph shows in Figure 2, the transfer of S-radiolabelled material from a reversibly cross-linked BAC gel to a 0.05 μm membrane target was enhanced by an interposed trypsin-embedded 0.45 μm membrane sheet. While some low molecular weight radiolabelled proteins were recovered on the 0.05 μm membrane target when a 0.45 μm control nitrocellulose membrane lacking trypsin was interposed, increased levels of radioactivity, particularly those associated with higher molecular weight proteins, were observed on the 0.05 μm membrane target when the interposed trypsin-embedded membrane sheet was used. Table 2 shows the recoveries of tryptic peptides generated from 5 high abundance ³⁵S-methionine labelled COS cell proteins captured on a nitrocellulose peptide replica. Table 3 shows the total tryptic peptide recoveries from ³⁵S- methionine labelled COS cell proteins captured on a nitrocellulose peptide replica. Table 2

Gel ] Densitometry μg Protein Band l Band 2 Band 3 Band 4 Band 5

80 3049 1802 1642 4098 2488

40 3168 1883 1647 4086 2296

20 2637 1936 1491 3676 1794

10 1831 1146 1244 3025 1248

5 1520 871 943 2374 997

Replica Densitometry μg Protein Band Band 2 Band 3 Band 4 Band 5

80 2126 1552 1442 1867 1629

40 2119 1304 1088 1625 1372

20 1898 850 747 1113 824

10 1385 685 593 821 674

5 936 541 478 641 577

% Recovery μg Protein Band l Band 2 Band 3 Band 4 Band 5

80 0.70 0.86 0.88 0.46 0.65

40 0.67 0.69 0.66 0.40 0.60

20 0.72 0.44 0.50 0.30 0.46

10 0.76 0.60 0.48 0.27 0.54

5 0.62 0.62 0.51 0.27 0.58

Table 3

Tricine-BAC Mini Gel μg Protein Dried Gel Replica Recovery

80 31907 28754 0.90

40 30961 22244 0.72

20 27486 14927 0.54 10 25941 6305 0.24

5 19263 3101 0.16 0.51 Mean 0.31 SD 0.14 SE

Tricine-BAC Phast Gel μg Protein Gel Replica Recovery •

2.5 ^■ 1722 950 0.55

2.5 1598 1616 1.01

1.25 1329 1346 1.01

1.25 479 406 0.85

0.625 1092 519 0.48

0.625 757 241 0.32

0.70 Mean

0.29 SD

0.12 SE

[0076] It was demonstrated through testing that a trypsin-impregnated nitrocellulose membrane binds significant amounts of peptides generated from proteins exiting a depolymerising gel. This made it an unsuitable proteolytic membrane for generating peptide replicas. >

[0077] Autoradiographs of gels and 0.05 μm nitrocellulose membrane targets were obtained to determine the efficiencies of transfer and capture of peptides generated during the replica process. Reversibly polymerized BAC gels were fixed, dried and autoradiographed after electrophoresis of different sample loads of ³⁵S-methionine labelled COS cell extracts. An identical set of samples run on a parallel gel were subjected to the replica process and the 0.05 μm target membrane was autoradiographed for the same time period as the untransferred gel. Films were scanned using a ScanJet 5370C (Hewlett Packard, Palo Alto, CA) using the transparency imaging device. Scion Image software (Scion Corporation, Frederick, MD) was used to quantitate the relative amounts of radiolabelled protein band densities detected by the film. As shown in Figure 3, the resolution of radiolabelled peptides on the target membrane was comparable to that observed for the same proteins autoradiographed in the gel matrix immediately following electrophoresis. The recoveries of 5 different radiolabelled protein bands as peptides on the replica membrane surface varied from 27-88% (Table 2). Measurement of the total lane densities on the autoradiographs revealed a more dramatic reduction (90%-16%) in the recoveries of peptides as sample loads decreased (Table 3). Peptide recoveries on the target membrane surface were highest for more abundant proteins and increased as greater amounts of protein were electrophoresed. The possibility that the interposed 0.45 μm trypsin impregnated nitrocellulose membrane was binding peptides as they were being generated from proteins exiting the depolymerizing gel was tested and confirmed to be the case. Attempts to reduce this problem by blocking binding sites with compounds like ethanolamine after trypsin impregnation failed to increase yields and resulted in the loss of resolution of protein bands on the 0.05 μm target membrane.

[0078] While the resolution of radiolabelled peptide bands on the replica surface using an interposed trypsin-impregnated 0.45 μm nitrocellulose membrane was reasonably good, the recovery of peptides associated with lower abundance proteins was less than optimal. This was readily recognized from the higher peptide recoveries measured for the 5 major radiolabelled protein bands compared to the total sample peptide recoveries on the replica surface at the 5 different gel loadings. The reduced recoveries of peptides from lower abundance proteins could not be fully accounted for by their failure to transfer out of the depolymerizing gel matrix. Rather, their retention by the interposed trypsin-impregnated 0.45 μm nitrocellulose sheet appeared to be the most likely cause for their failure to be recovered on the replica surface. Attempts to block peptide binding sites on the trypsin-impregnated 0.45 μm nitrocellulose sheet with compounds like ethanolamine increased the recovery of peptides on the replica surface but significantly reduced the resolution of protein bands.

EXAMPLE 3 - Tricine-BAC Phast-Gel & Replica

[0079] 41 mm wide x 50 mm long Phast system gels were made using the acrylamide/BAC, Tris, glycerol and SDS concentrations in Table 1. [0080] Two pieces of Gelbond (Sigma Chemical Co., St. Louis, MO) were attached to an outer mini-gel plate at four corners using 2 mm square tabs of two sided tape (3M Scotch Brand) with their hydrophobic sides fastened to the glass plate, 3mm from the bottom of the glass plate. Gel-off (ICN) was applied to the inner glass plate and allowed to dry to a haze (5 min) then wiped off. ' .

[0081] The mini-gel cassette was assembled using 1 mm spacers (resulting gel thickness was 0.55 mm). The two Tricine-BAC cross-linked Phast gels were then poured, using 2 ml of 16.5% separating gel buffer (20 mm) followed by 2 ml (20 mm) spacer gel buffer. The acrylamide was overlaid with 2 ml distilled water and allowed to polymerize for 15 min at 28°C (actual polymerization time is 4-5 min). The water was removed and 1.5 ml of 4% acrylamide/bis stacking gel buffer was poured and overlaid with distilled water to create a 15 mm stacking gel. After polymerization (about 15 min) the water was removed and the Gel-off- treated inner plate was carefully lifted off the formed gel. Using a scalpel blade, the excess separating, spacer and stacking gels were cut away from the edges of the Gelbond. The gels were carefully removed from the tape holding them to the outer glass plate, covered with plastic wϊap and stored at 4°C until use.

[0082] Buffer strips were formed from the following composition: Anode [0.4M Tris (pH

8.9, 20% Glycerol, 3% (w/v) IEF agarose] and Cathode [0.2M Tris, 0.2M Tricine (pH 8.4), 20% (v/v) Glycerol, 3% (w/v) IEF agarose, 0.1% SDS].

[0083] The agarosp was heated in boiling water until it dissolved and 2.5 ml/strip were poured into the Phast gel buffer strip moulds (Amersham Biosciences, Uppsala, Sweden). The agarose was allowed to set and the strips were covered with plastic wrap and stored at 4°C until use.

[0084] Prior to running gels, the Phast system apparatus (Amersham Biosciences,

Uppsala, Sweden) was allowed to cool to 15°C. Fifty microlitres of distilled water was placed on the gel template, the gels were placed on the template in the defined area carefully avoiding air bubbles. Excess water was wicked away using blotting paper. The buffer strips were placed in their respective holders; The samples to be applied were solubilized in the same low 0.5 mM DTT, 10% SDS sample buffer used for BAC cross-linked mini-gels and treated with a final concentration of 5 mM iodacetamide as described in Example 1. Two microlitres of sample were pipetted into parafilm sample wells made using the Phast system sample- well stamp (Amersham Biosciences, Uppsala, Sweden). The samples were drawn up into the 8/1 applicator capillary combs (Amersham Biosciences, Uppsala, Sweden) and placed into the appropriate slot. From time to time, the run was paused to check for fluid accumulation around the buffer strips. Blotting paper was used to wick away excess fluid. As shown in Figure 4, when a S- (Methionine/cysteine) labelled COS cell extract was electrophoresed on this system, then subjected to peptide replication onto nitrocellulose paper and autoradiographed, an acceptable resolution of a wide molecular weight range of proteins was obtained.

EXAMPLE 4

[0085] To assess the coupling efficiency of polylysine to a polycarbonate membrane, the amine-reactive compound, fluorescamine (catalogue # F-9015, Sigma Chemical Co., St. Louis, MO) was used.

[0086] Fluorescamine reacts strongly with polylysine and can be used to quantitatively measure the free polylysine content of reaction mixtures. Fluorescamine can also be used directly to estimate the degree of polylysine coupling to membrane surfaces following alkaline conjugation. Fluorescamine modification of polylysine side chains is readily detected by exposing derivatized membranes to an ultra-violet (UV) light source.

[0087] In one test to determine the degree of polylysine coupling to a polycarbonate membrane surface, small circular disks (0.476 cm diameter) of polycarbonate membrane (Osmonics Laboratory Products, Minnetonka, MN) were cut using a hole punch and 5 disks were mixed end-over-end for 24 hrs at 4°C in a 1.5 ml polypropylene microtube containing 200 ml of 10-50-mg/ml poly-L-lysine (80-150 kDa, catalogue # P-1274, Sigma Chemical Co., St. Louis, MO) dissolved in 50 mM sodium carbonate (pH 11). The membranes were subsequently rinsed with deionized water and reacted with fluorescamine followed by visualizaton with a UV. lamp to assess the degree of polylysine attachment to the membrane surface. As a composite test, the amounts of polylysine remaining unreactive with membrane disks following the alkaline incubation were quantitated by reaction with fluorescamine. Aliquots of 40 μl of reaction mixtures that had either been exposed to the polycarbonate membranes or not were removed from the microtubes and placed into a 96-well microtiter plate. Each well also received 135 μl of 50 mM sodium carbonate (pH 11) and 25 μl of 6 mg/ml fluorescamine dissolved in DMSO. Calibration curves were prepared similarly by diluting a stock solution of 1 mg/ml polylysine to encompass a concentration range of 0.1 to 0.5 mg/ml.

[0088] After allowing for 30 minutes reaction time between polylysine and fluorescamine, fluorescence readings were obtained on a microplate fluorimeter (Fluorescence Concentration Analyzer, Pandex Corp., LL). Readings were obtained using excitation and. emission wavelengths of 400 and 450 nm respectively. Based on the fluorescence differences between the samples that had previously been incubated in the presence and absence of the polycarbonate membrane disks it was estimated that 0.1 μm polycarbonate membranes bound approximately 126 pmole/cm² of polylysine (assuming an average molecular weight of 80 kDa) or 10.1 μg/cm².

[0089] Reaction times, temperatures and concentrations were optimized to the extent that a suitable derivatization of the polycarbonate surface was achieved for membranes having a Q.l μm pore diameter, irregardless of whether they were polyvinylpyrrolidone (PVP) coated (catalogue #K01CP320FX, Osmonics Laboratory Products, Minnetonka, MN) or not (PVPF - catalogue #K01SH810FX, Osmonics Laboratory Products, Minnetonka, MN) to increase their wettability. Less satisfactory results were obtained with membranes having larger, 8 μm (catalogue #K80CP81030, Osmonics Laboratory Products, Minnetonka, MN) or 20 μm (catalogue #K22CP810FX, Osmonics Laboratory Products, Minnetonka, MN) pore sizes.

[0090] It was uncertain why better results were consistently achieved with the 0.1 μm pore membranes. While not wishing to be bound by any theory or particular mode of action, there may be some fundamental differences in the track etch method used to generate larger diameter pores. The use of elevated temperatures and longer exposure times to caustic alkali to etch larger pores may reduce the reactivity of polycarbonate towards polylysine thereafter.

EXAMPLE 5 [0091] One possible coupling methodology to assemble a functional proteolytic membrane using a polylysine-coated polycarbonate surface involves the assembly of avidin/biotinylated trypsin on biotinylated polylysine-coated membranes.

[0092] The following example of this coupling methodology was performed using small circular disks (0.476 cm diameter) of polycarbonate membrane (Osmonics Laboratory Products, Minnetonka, MN) cut using a hole punch. These disks provided a uniform membrane surface with a well defined area for the attachment of trypsin. However, it will be apparent that the same coupling methodology can be scaled-up for larger membrane surfaces. Following the covalent attachment of polylysine to polycarbonate membrane disks under alkaline conditions, the polylysine modified polycarbonate membrane disks were biotinylated by incubation with 0.05 mg/ml n-hydroxysuccinimide (NHS)-biotin (Sigma Chemical Co. St. Louis, MO, catalogue #H- 1759) dissolved in 10 mM phosphate buffer (pH 7.7). Membranes disks were incubated at 4°C for 2 hrs, followed by extensive rinsing with deionized water. Membrane disks that served as controls were not biotinylated. While the polylysine-coated membrane disks were being biotinylated, 1.95 nmoles of biotinylated trypsin (46.8 μg) was mixed with 1.56 nmoles of avidin (100 μg) and diluted to a final volume of 1 ml with 70 mM citrate-phosphate buffer (pH 4.1) in a separate vial. The vials were rotated for 1 hr at 4°C to allow the avidin-biotinylated trypsin complex to form. Biotinylated polylysine-coated membrane disks were then washed extensively with deionized water and then incubated with the avidin/biotinylated trypsin solution for 24 hours, he membrane disks were then rinsed extensively with deionized water to remove non- specifically bound complexes. Membrane-associated tryptic activity can be monitored using the synthetic colorimetric reagent, nα-benzoyl-DL-arginine-p-nitroanilide hydrochloride (DL- BAPA) (Sigma Chemical Co. St. Louis, MO) and solution digestions of BSA by the assembled tryptic membranes can be evaluated by electrophoresis and Coomassie Blue staining of gels.

EXAMPLE 6

[0093] The following test illustrates the ability to biotinylate trypsin without a subsequent loss in enzymatic activity without prior protection of the active site with soybean trypsin inhibitor. Using the HABA/Avidin detection reagent (catalogue # H2153, Sigma Chemical Co., St. Louis, MO) it was determined that 4-5 out of the 6 lysines in bovine pancreatic trypsin were derivatized at a high biotin to protein weight ratio (eg., 1:1). As shown in Figure 5 this high level of derivatization did not inactivate the enzyme but reproducibly increased enzymatic activity.

[0094] A stock solution of trypsin (5-10 mg/ml) was prepared in 5 mM sodium phosphate (pH 8.0) and mixed with different amounts of NHS-biotin (catalogue #H1759 Sigma Chemical Co., St. Louis, MO) dissolved as a 10 mg/ml stock solution in DMSO. The relative weight ratios tested ranged from 1 mg trypsin (as determined by optical density assuming a 1. mg/ml solution gives an absorbance of 1.43 at 280 nm) per mg NHS biotin to 512 mg trypsin per mg NHS biotin. At the end of 1 hour on ice, 50 μl aliquots were mixed with 50 μl of 1 mg/ml of BSA in TBS containing 1 mM CaCl₂ and incubated at 37°C for 1 hour. Digestions were terminated by addition of 100 μl electrophoresis sample buffer and boiling for 1 minute. Nόn- biotinylated trypsin at the same relative concentration was used as a comparison of the digestion efficiency. Note that high ratios of biotimtrypsin (eg., 1:1) produced a more active or stable protease with slightly better digestion efficiency than non-biotinylated trypsin.

EXAMPLE 7

[0095] Recognition by trypsin of the C-terminal lysine peptide bond structure in polylysine and how this might competitively inhibit the proteolysis of protein substrates was examined. While it was previously reported that high molecular weight polylysines do not inhibit immobilized trypsin [25] this potential problem was directly investigated by incubating trypsin with BSA in the presence of different concentrations of the particular polylysine used for derivatization of the polycarbonate membranes. As shown in Figure 6 there was no diminution of the proteolysis of BSA by trypsin even at the highest concentrations of polylysine tested.

EXAMPLE 8

[0096] A test was performed to verify the fidelity of the biotinylation reaction. High levels of biotin, including those covalently bound to a biological target through the action of derivatization agents like NHS-biotin can competitively displace other biotinylated proteins from avidin. A quick method to check the fidelity of the biotinylation reaction was to biotinylate polylysine-coated polycarbonate membrane disks then incubate them with the avidin/biotinylated peroxidase reagent, Vectastain Elite ABC, (catalogue # PK6100, Vector Laboratories, Burlingame, CA). The disks were then exposed to the chemiluminescent substrate, SuperSignal (catalogue # 34080, Pierce Chemical Co., Rockford, LL) to evaluate the extent of complexation of the peroxidase enzyme with the biotinylated polylysine. As shown in Figure 7, polylysine coated membranes that were biotinylated with 0.05 mg/ml NHS biotin for 2 hours bound significant amounts of avidin/biotinylated peroxidase as evidenced by their high chemiluminescent signal relative to membranes that were not previously biotinylated.

EXAMPLE 9

[0097] To better quantify the amount of biotinylated peroxidase incorporated into the complex with avidin and the biotinylated polylysine on the polycarbonate membrane surface, the soluble peroxidase substrate, 3,3',5,5'-tetramethyl benzidine (1-Step Turbo TMB-ELISA, catalogue #34024, Pierce Chemical Co., Rockford, IL) was used to monitor membrane bound peroxidase activity. Membrane disks that had been coated with polylysine were biotinylated with various concentrations of NHS biotin in 100 μl PBS each for various periods of time. The disks were then incubated with VectaStain™ reagent for 3 hours. The amount of bound peroxidase was measured by incubating the disks in microtitre plate wells containing the colorimetric substrate and monitoring the absorbance at 450 nm of the reactions using a Titertek Multiskan MCC microplate reader (Eflab Oy, Helsinki, Finland).

[0098] As set out in Table 4, high levels of peroxidase activity were complexed to both

PVP and PVPF polycarbonate disks when the polylysine was biotinylated with less than 1 mg/ml NHS-biotin. High concentrations of NHS-biotin (>0.1 mg/ml) used to biotinylate polylysine side chains reduced the amount of peroxidase activity complexed to the polycarbonate disks presumably by displacing the biotinylated enzyme from the avidin complex. Biotinylation of polylysine on polycarbonate disks occurred rapidly with longer reaction times not producing significantly increased levels of avidin and biotinylated peroxidase binding to the membranes over the 1 hour time point. Biotinylation of polylysine coated, PVP-containing polycarbonate disks provided slightly higher levels of avidin and biotinylated peroxidase binding than PVPF membranes. This may be attributed to the better wetting properties of PVP-treated membranes which subsequently might allow better polylysine conjugation at the surface. Table 4

EXAMPLE 10

[0099] A quick test of the functionality of the assembled proteolytic membrane following biotinylation of the polylysine-coated surface and incubation with biotinylated trypsin and avidin was to overlay a gelatin-impregnated polyacrylamide gel with the proteolytic membranes. Tb.e loss of protein staining of the gel following an incubation period would be indicative of trypsin proteolysis of the gelatin substrate.

[00100] The method of Makowski and Ramsby [24] was used to conduct the test with a

7.5% bisacrylamide cross-linked polyacrylamide gel containing 1 mg/ml gelatin, 4 mM CaCl₂, 50 mM Tris/HCl (pH 7.6) incubated for 12 hr at 37°C with the overlaid membranes. As Figure 8 demonstrates, only membranes that had been assembled with biotinylated trypsin (either 1:16 or 1:32 weight ratios biotin/trypsin), avidin and biotinylated polylysine coated surfaces were capable of degrading gelatin imbedded in the polyacrylamide matrix. This was evidenced by the loss of Coomasie Blue staining in areas corresponding to where the membranes had been overlaid. Membranes that were assembled without prior biotinylation of the polylysine surface or without prior biotinylation of trypsin failed to assemble a proteolytic reactor capable of degrading gelatin. There was no corresponding loss of Coomassie Blue staining in areas of the gel where these membranes had previously been overlaid. EXAMPLE 11 .

[00101] To optimize the covalent attachment of trypsin to polylysine using the sulfo-

SANPAH cross-linker, polycarbonate membrane disks were used. These disks provided a uniform membrane surface with a well defined area for the attachment of trypsin. However, it will be apparent that the same coupling methodology can be scaled-up for use with larger membranes. Small circular disks (0.476 cm diameter) of polycarbonate membrane (Osmonics Laboratory Products, Minnetonka, MN) were cut using a hole punch. From a stock solution of 50 mg/ml of sulfo-SANPAH dissolved in dimethyl sulfoxide (DMSO), 0.1 mg/ml solutions of sulfo-SANPAH in 50 mM phosphate buffer (pH 7.5) were prepared. The polylysine-coated polycarbonate membrane disks were incubated with 0.1 mg/ml sulfo-SANPAH at 4°C in an aluminum foil covered 1.5 ml microfuge tube with end-over-end mixing. After 1 hour of incubation, the membrane disks were rinsed with deionized water and placed individually into wells of a 96-well microtiter plate. Each well contained 200 μl of 1 mg/ml of trypsin dissolved in 50 mM phosphate/lmM CaCl₂ (pH 7.5). A light box (λ_max=312 nm, model TFL-20M, France) was used to irradiate the microtiter plate for 20 minutes. The disks were then rinsed with deionized water followed by 100 mM phosphate buffer (pH 7.7).

EXAMPLE 12

[00102] To assess whether the proteolytic activities provided by two different conjugation methods, avidin-biotin and sulfo-SANPAH, were stably associated with polycarbonate membranes complexed with trypsin by the two conjugation methods, disks having undergone the two conjugations methods were first incubated with DL-BAPA for a period of time and the absorbance changes were then recorded as shown in Figures 9 and 10.

[00103] Figure 9 shows the digestion of DL-BAPA by biotinylated trypsin complexed to polylysine-coated polycarbonate membrane disks with avidin.

[00104] Figure 10 shows digestion of DL-BAPA by trypsin coupled to polylysine-coated polycarbonate membrane disks with sulfo-SANPAH. The membranes were then removed from the substrate solution and placed in new microtitre wells containing fresh DL-BAPA. [00105] As shown in Figure 11, only membranes that had undergone trypsin coupling using sulfo-SANPAH continued to hydrolyse the DL-BAPA at similar rates when transferred to fresh substrate solution. The initial rates of DL-BAPA hydrolysis by membranes containing trypsin immobilized by the biotin avidin technique became significantly reduced when disks were transferred to fresh substrate solution. Avidin contains several lysine and arginine residues within its primary sequence and these could potentially be acted on by the biotinylated trypsin bound to it. Proteolytic cleavage of avidin by trypsin may have been responsible for releasing these proteins from the polylysine-coated membrane surface. Continued monitoring of DL- BAPA hydrolysis following removal of membranes initially derivatized using the biotin/avidin chemistry indicated that significant amounts of soluble tryptic activity had been generated. Thus to increase the fidelity of this type of membrane there would be a need to generate a more trypsin-resistant avidin molecule either by site-directed mutagenesis or chemical derivatization of sensitive lysine. and/or arginine residues.

EXAMPLE 13

Λ

[00106] To increase the stoichiometry of sulfo-SANPAH coupling of trypsin to polylysine, in order to increase the proteolytic activity of the polycarbonate membranes so coupled, polymerization of trypsin using glutaraldehyde was performed.

[00107] While the sulfo-SANPAH reaction produces a stable covalent crosslink between trypsin and the polylysine-coated polycarbonate surface, increasing the coupling stoichiometry would be advantageous suice it would endow membranes with a greater proteolytic capacity. This in turn would permit larger amounts of proteins to be digested more efficiently during vacuum transfer through membrane pores en route to the target surface. High proteolytic activity (0.9 μg/cm²) was previously measured for trypsin covalently bound to immobilon membranes [10] and this was considered to be sufficient to digest Coomassie Brilliant Blue stained proteins during electrotransfer to PVDF membranes.

[00108] Glutaraldehyde modification of lysine residues in trypsin stabilizes the protein from unfolding, reduces autolysis and improves its hydration state [26]. These favourable properties serve to increase its proteolytic activity. To also increase the stoichiometry of sulfo- SANPAH coupling of trypsin to polylysine a controlled polymerization of trypsin using glutaraldehyde was undertaken. 10 mg/ml trypsin was prepared in 0.05 M sodium borate (pH 8.0), 20 mM CaCl₂ and treated with various amounts of 50% glutaraldehyde for 10 minutes at 4°C with end-over-end mixing. Reactions were stopped by addition of a 1/10 volume of 0.5 M Tris/HCl (pH 7.5). Modified trypsin samples that remained soluble (those treated with less than 0.1% and greater than 1% glutaraldehyde) were dialyzed overnight against 1 mM HCl and analyzed by gel electrophoresis. Enzyme activities were also measured on glutaraldehyde- modified trypsin samples using the DL-BAPA assay. Similar to results obtained for biotinylated trypsin, mildly increased enzymatic activities were observed for samples treated with 0.05% glutaraldehyde relative to unmodified trypsin. Moderate glutaraldehyde concentrations (>0.1%<1.0%) produced a heavily crosslinked, insoluble enzyme and at higher glutaraldehyde concentrations (>1%) a soluble protein with substantially reduced proteolytic activity. Low glutaraldehyde concentrations (> 0.005%<0.10%) produced an activated, soluble, crosslinked trypsin as judged by the DL-BAPA assay and SDS-polyacrylamide gel electrophoresis respectively. This trypsin preparation was deemed optimal for coupling to polylysine-coated, polycarbonate membranes.

[00109] As shown in Figure 12, when glutaraldehyde-polymerized trypsin was crosslinked to polylysine by the sulfo-SANPAH reaction, a 4-8 fold increase in the proteolytic activity of polycarbonate membrane disks was observed relative to unmodified trypsin. The solid curves represent that initial hydrolysis rates of DL-BAPA for membranes during the first incubation cycle which was conducted for 3 hours in the presence of substrate at 37°C. Thereafter, the membranes were removed, rinsed and placed into wells containing fresh solution of DL-BAPA for a second incubation cycle (dash curves). Figure 12 demonstrates the higher rates of DL- BAPA hydrolysis by membrane disks containing glutaraldehyde-conjugated trypsin.

[00110] The reduced proteolytic activity of membranes containing glutaraldehyde- crosslmked trypsin during the second cycle was attributed to loss of non-covalently bound enzyme which had become absorbed to polycarbonate membranes non-specifically during the sulfo-SANPAH reaction. Glutaraldehyde-crosslmked trypsin showed higher levels of absorption to unmodified polycarbonate membranes and required more stringent conditions to effect its removal. Estimates of the enzymatic activity of membrane disks containing glutaraldehyde- crosslmked trypsin based on standard curves established by incubating unmodified trypsin in the DL-BAPA assay indicated more than 1 μg/cm² of trypsin activity had become conjugated to the polylysine-coated surfaces. These results compared favourably with estimates previously reported for immobilon membranes containing covalently bound trypsin [10].

EXAMPLE 14 ^' .

[00111] Two assays were used to ascertain the extent of coupling of trypsin to polylysine coated, polycarbonate membrane disks.

[00112] The first assay involved the digestion of DL-BAPA. DL-BAPA was dissolved in

DMSO to achieve a concentration of 50 mg/ml. This solution was diluted to 0.5 mg/ml with 50 mM Tris/HCl, 20 mM CaCl₂ (pH 8.2). Using a 96 well microtiter plate, 200 μl of 0.5 mg/ml DL- ' BAPA solution was added to each well. An additional 50 μl of 50 mM Tris/HCl, 20 mM CaCl₂ (pH 8.2) was added to wells that were to receive polylysine-coated polycarbonate membrane disks. The quantities of free biotinylated trypsin serving as standards for in-solution digestion were 1, 10, 50, 100, 200 and 500 nanograms. Volumes of 50 mM Tris/HCl, 20 mM CaCl₂ (pH 8.2) added to standard wells were adjusted accordingly to achieve a total volume of 250 μl with a final concentration of DL-BAPA equivalent to 0.4 mg/ml. Absorbance of the solutions owing to cleavage of the DL-BAPA substrate by trypsin was monitored at 414 nm using a Titertek Multiskan MCC microplate reader (Eflab Oy, Helsinki, Finland).

[00113] The second assay involved the digestion of BSA. Trypsin-immobilized polylysine-coated, polycarbonate membrane disks were individually immersed in 200 μl of 0.1 mg/ml BSA dissolved in 50 mM Tris HCl, 20 mM CaCl₂ (pH 8.2). Digestion was allowed to proceed at 37°C and 20 μl aliquots were removed at different times for SDS-PAGE. Each 20 μl aliquot was mixed with 6.67 μl of 4x electrophoresis sample buffer. Samples were electrophoresed on a 1 mm thick, 4% stacking/10% separating mini-gel at constant current of 40 mA per gel by the method of Laemmli [3]. Gels were stained with 0.25% Coomassie Blue R250 dissolved in 25% isopropanol, 10% acetic acid and destained with 30% methanol, 10% acetic acid. Gels were equilibrated with 2% glycerol, 10% acetic acid overnight, dried and scanned with a ScanJet 5370C (Hewlett Packard, Palo Alto, CA). Densitometry of the 66 kDa BSA protein band was performed with Scion Image software (Scion Corporation, Frederick, MD). EXAMPLE 15

[00114] As shown in Figure 13, HABA pre-coated porous plates are particularly suitable target surfaces since they can readily capture and ionize large peptides as well as peptide digests. Although BSA ionization signals were lower using nitrocellulose HABA pre-coated porous stainless steel plates than those obtained with the dried droplet co-crystallization method using a commercial solid stainless steel plate, the sample volumes that could be vacuum deposited onto the former (100-200 μl) were much greater than for a solid target plate (1-2 μl).

[00115] Figure 13 shows MALDI MS TOF of 1 μg BSA. Specifically, it shows (A) 1 ml of sample was spotted on a solid stainless steel MALDI plate as a dried droplet in a 1:1 mixture with 10 mg/ml a-CHCA dissolved in 65% acetonitrile/0.1% acetic acid. (B) 100 ml of sample was vacuum deposited on a nitrocellulose/a-CHCA sequentially coated, 0.10 mm porous MALDI plate. (C) 1 ml of sample was spotted on a solid stainless steel MALDI plate as a dried droplet in a 1:1 mixture with 10 mg/ml HABA dissolved in 65% acetonitrile/0.1% acetic acid. (D) 100 ml of sample was vacuum deposited on a nitrocellulose/HABA sequentially coated, 0.10 mm porous MALDI plate. Spectra were acquired on a Perseptive Biosystems Voyager DE STR operating in linear mode. 100 shots per spectrum were accumulated operating with a laser intensity of 2142 (A & B) and 2242 (C & D).

EXAMPLE 16 - Effect of Salt

[00116] The effect of salt suppression of the total ion count (TIC) for peptides derived by tryptic digestion of BSA was assessed for both the α-CHCA and HABA matrices. The restoration of peptide ionization signals as a consequence of sample cleanup of digests on a PRP- 3 polymeric reverse phase resin (Hamilton Company) was also investigated for the dried droplet co-crystallization technique using a solid stainless steel MALDI target surface. As Figure 14 shows there is a dramatic enhancement of BSA peptide TICs as a consequence of PRP-3 sample cleanup when α-CHCA is used. Specifically, Figure 14 shows MALDI MS TOF analysis of BSA tryptic digests spotted on a solid stainless steel MALDI plate. A and C show digests not reverse phased with PRP-3 B and D. B and D digests reverse phased with PRP-3. 9.5 picomoles (A,B) and 0.19 picomoles (C,D) of BSA tryptic peptides were spotted as a 1:1 mixture with 10 mg/ml a-CHCA dissolved in 65% acetonitrile/0.1% acetic acid. Spectra were accumulated in reflectron mode and represent an average of 50 laser shots at a laser intensity of 2057 Comparable TICs were obtained when nitrocellulose/HABA sequentially pre-coated porous plates were used to analyse the same amounts of BSA tryptic peptides. While α-CHCA pre- coated porous plates produced lower TICs for BSA peptides (Figure 15 A,C) than HABA plates

(Figure 15 B,D), they were considerably greater than the TICs obtained for samples that were not reverse phased with PRP-3 and co-crystallized with α-CHCA on solid MALDI plate surfaces by the dried droplet method (Figure 14, A,C). Specifically, Figure 15 shows MALDI MS TOF analysis of BSA tryptic peptides vacuum deposited on porous stainless steel MALDI plates pre- coated with nitrocellulose and either α-CHCA (A,C) or HABA (B,D). 9.5 picomoles (A,B) and 0.19 picomoles (C,D) of BSA tryptic peptides were spotted. Spectra were accumulated in reflectron mode and represent an average of 50 laser shots at a laser intensity of 2157.

EXAMPLE 17 - Optimal Nitrocellulose and Matrix Coating Conditions

[00117] The ionization efficiencies of BSA tryptic peptides from porous stainless steel

MALDI plates pre-coated with either a mixed nitrocellulose/matrix formulation or sequentially applied nitrocellulose followed by matrix were assessed. As shown in Table 5, the TICs for three prominent BSA peptides were highest for the sequentially pre-coated plates particularly when HABA was used. Table 5 shows a comparison of Nitrocellulose/Matrix Precoating methods on BSA Tryptic Peptide ionization efficiency from porous stainless steel plates.

[00118] TICs were measured for selected BSA tryptic peptides spotted onto porous stainless steel MALDI plates. Plates were pre-coated either by spraying with a mixture of 4 ml of nitrocellulose dissolved in acetone (1 mg/ml) and saturated with α-CHCA or HABA or by sequentially spraying first with 4 ml of nitrocellulose dissolved in acetone (1 mg/ml), followed by saturated α-CHCA or HABA in methanol (4 ml). The amount of BSA tryptic peptides spotted was prepared from a stock BSA tryptic digest solution (125 ng/μl) diluted to 10 μL in 0.1% acetic acid. TIC were calculated from mass spectra averaged from 50 laser shots using a laser intensity for α-CHCA and HABA respectively were 2057 and 2157.

[00119] While sequentially coated nitrocellulόse/HABA porous stainless plates produced favourable ionizations of peptides, a mixed coating was more effective when α-CHCA was used as a matrix. A range of nitrocellulose concentrations used to pre-coat porous plates was assessed to evaluate their effects on peptide TICs.

[00120] As shown in Figure 16, BSA tryptic peptide ionization signals were maximal when nitrocellulose pre-coatings employed mixtures of 1-2 mg/ml of mtrocellulose in 20 mg/ml α-CHCA. Specifically, Figure 16 shows MALDI TOF MS analysis of 10 picomoles of BSA tryptic peptides vacuum deposited on porous stainless steel MALDI plates pre-coated with 10 mg/ml (A), 2 mg/ml (B), 1 mg/ml (C) and 0.01 mg/ml (D) of nitrocellulose in and 20 mg/ml α- CHCA. Spectra were accumulated in reflectron mode and represent an average of 50 laser shots at a laser intensity of 2157. It is likely that lower concentrations of nitrocellulose fail to provide a significant binding surface for vacuum deposited peptides while higher concentrations reduce the peptide volatility.

Table 5

Mixed Nitrocellulose/Matrix Precoating α-CHCA HABA

BSA digest TIC TIC TIC TIC TIC TIC

(Picamole) m/z 1439 m/z 1479 m/z 1881 m/z 1439 m/z 1479 m/z 1881

9.5 9100 5600 5300 14400 11000 12300

3.8 14450 8840 3400 ND ND ND

1.9 9350 7990 3060 ND ND ND

0.19 4830 5040 ND ND ND ^' ND

Sequential Nitrocellulose/Matrix Precoating α-CHCA HABA

BSA digest TIC TIC TIC TIC TIC TIC

(Picamole) m/z 1439 m/z 1479 m/z 1881 m/z 1439 m/z 1479 m/z 1881

9.5 16490 15980 10370 24000 16800 23520

3.8 15360 24890 6480 15360 10720 15360

1.9 7140 8120 4200 14000 10920 13860

0.19 ND ND ND 180 ND ND EXAMPLE 18 T- Dynamic Range

[00121] To assess the dynamic range of peptide ionization signals, porous plates were precoated with 2 mg/ml nitrocellulose in 20 mg/ml α-CHCA then various amounts of BSA tryptic peptides were vacuum deposited.

[00122] As demonstrated in Figure 17, this modified porous surface provided a broad range of ionization signals beginning with a limit of detection of less than 1 pmole of BSA tryptic peptides increasing linearly in intensity beyond 15 pmoles of peptides Specifically, Figure 17 shows MALDI TOF MS analysis of 15 (A), 3.8 (B), 0.8 (C) and 0.2 (D) pmoles of BSA tryptic peptides vacuum deposited on porous stainless steel MALDI plates pre-coated with 2 mg/ml nitrocellulose in 20 mg/ml a-CHCA. Spectra were accumulated in reflectron mode and represent an average of 50 laser shots at a laser intensity of 2157.

EXAMPLE 19 - Volume of Sample Effect

[00123] The considerable differences in sample deposition volumes between the dried droplet co-crystallization method and the porous plate vacuum deposition technique prompted an investigation into whether peptide TICs could be maintained over a wide range of sample sizes when the porous plate vacuum deposition technique procedure was employed. As shown in Table 6, the TICs for two relatively abundant BSA tryptic peptides were relatively constant over a wide sample volume deposition range using both the α-CHCA and HABA pre-coated surfaces. These results clearly demonstrated the effectiveness of the pre-coated porous MALDI plate

I target surface as an efficient peptide capture surface. Furthermore this type of target surface can provide a sufficient ionization potential to a broad range of peptide concentrations present in samples and subsequently arrayed within a reasonably small area.

[00124] Table 6 shows the effect of sample spotting volume on BSA tryptic peptide ionization efficiency from nitrocellulose/matrix sequentially pre-coated, porous stainless steel plates. Table 6 shows the TICs (± SD) for selected BSA tryptic peptides. 9.5 picomole samples . of the digest were spotted onto a nitrocellulose/α-CHCA or nitrocellulose/HABA sequentially coated porous MALDI plate in different buffer volumes. Laser intensities were 2057 and 2157 for the α-CHCA and HABA coated plates respectively and 50 laser shots were averaged for each spectrum.

Table 6 α-CHCA BSA Digest Peptide Masses

Volume (μL) M/z 1439 TIC M z 1880 TIC

10 10491 ± 2590 3857 ± 566

25 9043 ± 2966 5539 ± 1094

50 13027 ± 3468 4484 ± 509

100 15948 ± 770 5758 ± 878

200 9552 ± 2680 3158 ± 163

HABA BSA Digest Peptide Masses

Volume (μL) M/z 1439 TIC M z 1880 TIC

10 7678 ± 1177 16135 ± 1332

25 7423 ± 809 11378 ± 2130

50 18763 ± 2694 10762 ± 1066

100 7173 ± 1915 7790 ± 1186

200 6582 ± 1462 5428 ± 796

EXAMPLE 20 - Uniformitv

[00125] Another important requirement for a successful matrix pre-coated porous plate as a replica target surface is the ability to uniformly ionize peptides across its entire surface. This was tested in a series of experiments which evaluated the standard deviation in TICs obtained for angiotensin and ACTH peptides evenly vacuum deposited across the surfaces of α-CHCA and HABA pre-coated porous plates.

[00126] As shown in Table 7, the variation in TICs for these 2 peptides was greatest for the larger ACTH molecule and increased with higher laser intensity. Both matrices demonstrated similar ionization reproducibility which varied between 5 and 40% SD depending on the firing conditions and peptide analyzed. [00127] Ionization .reproducibility across a porous MALDI plate was investigated. An equal mixture of Angiotensin I and ACTH 18-39 (5 μg each) in 2 ml of 0.1 % acetic acid was spotted on a sequentially coated nitrocellulose-matrix porous MALDI plate (area = 5 cm²). For each matrix, 20 spectra were obtained at random locations, with two laser intensity (LI) settings and 100 shots per spectrum averaged. The average TICs and relative percent standard deviation was calculated using N=20. Table 7 shows the peptide ionization uniformity from porous stainless steel plates.

Table 7

Nitrocellulose/α-CHCA Nitrocellulose/HABA

Peptide LI 1857 LI 1957 LI 2007 LI 2057 .

Angiotensin I 37919 58657 40030 61473

Average TIC ACTH 18-39 13021 30924 11193 30531

Relative Percent Angiotensin I 16 10 27 6 SD ACTH 18-39 37 20 34 18 ^

EXAMPLE 21

[00128] The relative efficiencies of peptide ionizations in the MALDI TOF using the two different matrix pre-coated surfaces (α-CHCA and HABA) and the dried droplet cocrystallization technique Is were measured for different amounts of BSA tryptic peptide digest on both porous and solid stainless steel plates. The effect of reverse phase sample cleanup on the ionization efficiencies of three abundant BSA tryptic peptides was assessed using the dried droplet method to co-crystallize either the α-CHCA or HABA matrix.

[00129] Aliquots of a BSA tryptic digest (125 ng/μl in 0.1 M ammonium biocarbonate) were diluted to 50 μl with 0.1% acetic acid then vortexed for 1 hour with 1 μl of a 1:1 suspension of PRP-3 in 65% acetonitrile/1% acetic acid. After washing resins by centrifugation twice with 250 μl of 2% acetonitrile/1% acetic acid, peptides were eluted with 5 μl of 65% acetonitrile/1% acetic acid. Alternatively, the peptide digest was not subjected to reverse phase cleanup and samples were diluted to an appropriate concentration in 0.1 M ammonium bicarbonate for direct spotting with matrix on a solid stainless MALDI plate.

[00130] As shown in Table 8A, both α-CHCA and HABA pre-coated porous plates capture and ionize BSA tryptic peptides relatively well even when relatively low sample amounts (0.19 pmoles) are vacuum deposited. Table 8 A shows TIC for a BSA digest spotted on porous MALDI plates that were sprayed first with 4 ml of nitrocellulose dissolved in acetone (1 mg/ml), followed by saturated α-CHCA or HABA in methanol (4 ml). The amount of BSA digest spotted was prepared from a stock BSA digest solution (125 ng/μl) and was diluted to 10 μl with a saturated solution of α-CHCA or HABA in 0.1% acetic acid. TICs were calculated from mass spectra averaged from 100 laser shots using respective laser intensities for α-CHCA and HABA of 2057 and 2157.

[00131] In comparison, samples spotted on solid stainless plates by the dried droplet co- crystallization method ionized poorly unless they underwent prior cleaning by reverse phasing on PRP-3. Peptide digests co-crystallized with HABA (Table 8C) were less dependent on prior reverse phasing than their α-CHCA counterparts (Table 8B).

[00132] Table 8B shows TIC for a BSA digest spotted as a dried droplet co-crystal with α-

CHCA on a solid stainless steel MALDI plate. The amount of BSA digest spotted was prepared from a stock BSA digest solution (125 ng/μl). Individual samples of the digest were reverse- phased using PRP-3 and peptides were eluted with 65% acetonitrile/1%) acetic acid or left untreated. Samples were spotted as a 1:1 mixture of 10 mg/ml α-CHCA dissolved in 65% acetonitrile/0.1% acetic acid. TICs were calculated from mass spectra averaged from 50 laser shots using a laser intensity of 2057.

[00133] Table 8C shows TIC for a BSA digest spotted as a dried droplet co-crystal with

HABA on a solid stainless steel MALDI plate. The amount of BSA digest spotted was prepared from a stock BSA digest solution (125 ng/μl). Individual samples of the digest were reverse- phased with PRP-3 and peptides luted with 65% acetonitrile/1.0% acetic acid or left untreated. Samples were spotted as a 1:1 mixture of 10 mg/ml HABA dissolved in 65% acetonitrile/0.1% acetic acid. TICs were calculated from mass spectra averaged from 50 laser shots using a laser intensity of 2057.

[00134] After PRP-3 cleanup, peptide ionization signals from the lowest amounts of BSA digest spotted (1.9 and 0.19 pmol) were still much weaker using the dried droplet co- crystallization technique than those obtained with comparable amounts vacuum deposited peptide digest on matrix pre-coated porous plates. The greater ability of the matrix pre-coated porous plate for producing peptide ions at low sample loads could in part be attributed to the> reduced losses associated with direct capture of peptide samples on this surface. Porous plates pre-coated with either α-CHCA or HABA did not appear to suffer the decrimental decreases in ionization signal intensities characteristic of the dried droplet co-crystallization technique especially at low sample loads. This finding could presumably be accounted for in part by the higher losses of material incurred during reverse phase sample cleanup of small concentrations of peptides. The results affirmed the matrix pre-coated porous plate as a sensitive target surface for peptide capture and ionization in the gel replica process.

[00135] Subsequent improvements in reagent quality improved peptide signal intensities with corresponding reductions in trace contaminant ion signals. Electron microscopy grade collodion (catalogue #12620-50, Electron Microscopy Sciences) which is 2% nitrocellulose in amyl acetate, was evaporated to dryness and reconstituted to the applicable concentration in acetone as specified above.

[00136] While this invention has been described with reference to illustrative embodiments and examples, the description is not intended to be construed in a limiting sense. Thus, various modifications of the illustrative embodiments, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to this description. It is therefore contemplated that the appended claims will cover any such modifications or embodiments.

[00137] All publications, patents and patent applications referred to herein are incorporated by reference in their entirety to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference in its entirety. Table 8A

^• A α-CHCA HABA

BSA digest TIC TIC TIC TIC TIC TIC (picomole) m/z 1439 m/z 1479 m z 1881 m/z 1439 m/z 1479 m/z 1881

9.5 21781 10673 7118 7696 7542 4618

3.8 21236 14228 11255 13615 9803 5854

1.9 13352 14513 6676 8561 4452 3596

0.19 6417 6296 3027 4017 3610 1525

Table 8B

B No PRP-3 PRP-3 SA digest TIC TIC TIC TIC TIC TIC

(picomole) m/z 1439 m/z 1479 m z 1881 m/z 1439 m/z 1479 m/z 1881

9.5 6700 9900 5400 13320 35640 29520

3.8 ND ND ND 7360 3360 13920

1.9 ND ND ND 2148 4525 5062

0.19 ND ND ND ND ND ND

Table 8C

C No PRP-3 PRP-3

BSA digest TIC TIC TIC TIC TIC TIC

(picomole) m/z 1439 m/z 1479 m/z 1881 m/z 1439 m/z 1479 m/z 1881

9.5 8433 5313 1940 18000 8100 1800

3.8 7783 4436 1401 4387 7563 2723

1.9 9025 6555 1805 2380 2940 ND

0.19 4414 3001 1192 ND ND ND References

[1] Liebler, D.C. (2002) Introduction to Proteomics. Tools for the New Biology.

Humana Press, Totowa, NJ.

[2] Ornstein, L. Disc Electrophoresis - 1 Ann. New York Acad. Sci. 121 , 321 -349,

1964.

[3] Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-695, 1970.

[4] O'Farrell, P.H. High resolution two-dimensional electrophoresis of proteins. J. Biol.

Chem. 250, 4007-4021, 1975.

[5] Towbin, H., Staehelin, T., and Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Nat. Acad. Sci. USA 76, 4350-4354, 1979.

[6] Gibson, W. Protease-facilitated transfer of high-molecular- weight proteins during electrotransfer to nitrocellulose. Anal. Biochem. 118, 1-3, 1981.

[7] Hansen, J.H., Pheiffer, B.H., and Boehnert, J.A. Chemical and electrophoretic properties of solubilizable disulfide gels. Anal. Biochem. 105, 192-201, 1980.

I

[8] Peferoen, M., Huybrechts, R., and De Loof, A. Vacuum-blotting: a new simple and efficient transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to nitrocellulose. FEBS Lett. 145, 369-372, 1982.

[9] Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Anal. Chem. 68, 850-858, 1996.

[10] Bienvenut, W., Sanchez, J.-C, Karmime, A., Rouge, V., Rose, K., Binz, P.-A., and

Hochstrasser, D.F. Toward a clinical molecular scanner for proteome research: Parallel protein chemical processing before and during Western blot. Anal. Chem. 71, 4800-4807, 1999.

[11] Binz, P.-A., Muller, M., Warmer, D., Bienvenut, W.V., Gras, R., Hoogland, C,

Bouchet, G., Gasteiger, E., Fabbretti, R., Gay, S., Palagi, P., Wilkins, M.R., Rouge, V., Tonella, L., Paesano, S., Rόssellat, G., Karmime, A., Bairoch, A., Sanchez, J.-C, Appel, R.D., and Hochstrasser, D.F. A molecular scanner to automate proteomic research and to display proteome images. Anal. Chem. 71, 4981-4988, 1999.

[12] Muller, M., Gras, R., Appel, R.D. Bienvenut W. V., and Hochstrasser, D.F.

Visualization and analysis of molecular scanner peptide mass spectra. J. Am. Soc. Mass Spectrom. 13, 221-231, 2002. ,

[13] Suye, S.-L, Kumon, Y., and Ishigaki A. Immobilization of glucose oxidase on poly-

L-lysine)-modified polycarbonate membrane. Biotechnol. Appl. Biochem. 27, 245- 248, 1998.

[14] Bienvenut, W.V., Deon, C, Sanchez J.-C, and Hochstrasser, D.F. Enhanced protein recovery after electrotransfer using square wave alternating voltage. Anal. Biochem. 307, 297-303, 2002.

[15] O'Connell, P.B.H. and Brady, C.J. Polyacrylamide gels with modified cross- linkages. Anal. Biochem. 76, 63-73, 1976.

[16] Gelfi, C, and Righetti, G. Polymerization kinetics of polyacrylamide gels I. Effect of different cross-linkers. Electrophoresis 2, 213-219, 1981a

[17] Gelfi, C, and Righetti, G. Polymerization kinetics of polyacrylamide gels II. Effect of temperature. Electrophoresis 2, 220-228, 1981b

[18] Faulkner, R.D., Carraway, R., and Bhatnagar Y.M. Preparative electrophoresis of histones in solubilizable polyacrylamide gels. Biochim. Biophys. Acta 708, 245-252, 1982. [19] Airey, J.A., Rogers, M.J. and Sutko, J.L.. Use of a reversible polyacrylamide gel cross-linker in Western blotting for rapid transfer of a wide size range of polypeptides. Biotechniques 10, 605-608, 1991.

[20] Schagger, H. and von Jagow, G. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins from 1 to 100 kDa. Anal. Biochem. 166, 368-3,79, 1987.

[21] Fleischer, R.L., Price, P.B., Walker, RM. Nuclear Tracks in Solids (Univ. of

California Press, Berkeley, Calif. (1975).

[22] Melrose, J., Rodgers, K., and Ghosh, P. The preparation and use of biotinylated trypsin in Western blotting for the detection of trypsin inhibitory proteins. Anal. Biochem. 222, 34-43, 1994.

[23] Wood, C.L., and O'Dorisio M.S. Covalent cross-linking of vasoactive intestinal polypeptide to its receptors on intact human lymphoblasts. J. Biol. Chem. 260, 1243- 1247, 1985.

[24] Venkatesh, R., and Sundaram, P.V. Modulation of stability properties of bovine trypsin after in vitro structural changes with a variety of chemical modifiers. Protein Engineering 11, 691-698, 1998.

[25] Shtelzer, S., Rappoport, S., Avnir, D., Ottolenghi, M., and Braun, S. Properties of trypsin and of acid phosphatase immobilized in sol-gel glass matrices. Biotechnol. Appl. Biochem. 15, 227-235, 1992.

[26] Makowski,G.S. and Ramsby, M.L. Calibrating gelatin zymograms with human gelatinase standards. Anal. Biochem. 236, 353-356, 1996.

Claims

What is claimed is:

1. A method for generating and capturing peptides on a target surface, comprising:

(a) electrophoresing at least one protein on a high resolution reversibly polymerized polyacrylamide gel;

(b) vacuum depositing said at least one electrophoresed protein on a digestion membrane for digestion into peptides,

(c) said digestion membrane comprising a substantially chemically inert membrane having a proteolytic enzyme immobilized thereon; and

(d) vacuum depositing peptides generated by said digestion membrane on said target surface.

2. The method as in claim 1, wherein said digestion membrane comprises a polycarbonate membrane.

3. The method as in claim 1, wherein said digestion membrane comprises a polyester membrane.

4. The method as in claim 1, wherein said digestion membrane has a thickness of less than about 10 micrometers.

5. The method as in claim 1, wherein said digestion membrane comprises a porous membrane having an average pore diameter between 0.1 micrometers and 10 micrometers.

6. The method as in claim 1, wherein said polyacrylamide gel comprises N,N - bisacrylylcystarnine.

7. The method as in claim 1, wherein said proteolytic enzyme comprises trypsin.

8. The method as in claim 1, wherein biotinylated polylysine is bonded to said digestion membrane and wherein said proteolytic enzyme is bonded to avidin and wherein said proteolytic enzyme is immobilized on said digestion membrane by affinity bonding between avidin and biotin.

9. The method as in claim 1, wherein said proteolytic enzyme is immobilized by means of a heterobifunctional cross-linker.

10. The method as in claim 9, wherein said heterobifunctional cross-linker is sulfo- SANPAH.

11. The method as in claim 7, wherein said trypsin is polymerized using glutaraldehyde.

12. The method as in claim 1, wherein said target surface comprises a porous plate that does not ionize or volatize in the charge conditions of a mass spectrometer and has a coating comprising an organic matrix and nitrocellulose.

13. A system for generating and capturing peptides on a target surface, comprising:

(a) a high resolution reversibly polymerized polyacrylamide gel for electrophoresing at least one protein;

(b) a digestion membrane comprising a substantially chemically inert membrane having a proteolytic enzyme immobilized thereon for digestion of said at least one protein; and

(c) a target surface for capturing peptides generated by said digestion membrane. .

14. The system as in claim 13, wherein said digestion membrane comprises a polycarbonate membrane.

15. A system as in claim 13, wherein said digestion membrane comprises a polyester membrane.

16. The system as in claim 13, wherein said polyacrylamide gel comprises N,N - bisacrylylcystamine.

17. The system as in claim 13, wherein said proteolytic enzyme comprises trypsin.

18. The system as in claim 13, wherein biotinylated polylysine is bonded to said digestion membrane and wherein said proteolytic enzyme is bonded to avidin and wherein said proteolytic enzyme is immobilized on said digestion membrane by affinity bonding between avidin and biotin.

19. The system as in claim 13, wherein said proteolytic enzyme is immobilized on said digestion membrane by means of a heterobifunctional cross-linker.

20. The system as in claim 19, wherein said heterobifunctional cross-linker comprises sulfo-SANPAH.

21. The system as in claim 17, wherein said trypsin is polymerized using glutaraldehyde.

22. The system as in claim 13, wherein said target surface comprises a porous plate that does not ionize or volatize in the charge condition of a mass spectrometer and has a coating comprising an organic matrix and nitrocellulose.

23. A target surface for use in a mass spectrometer, comprising a porous plate that does not ionize or volatize in the charge conditions of the mass spectrometer.

24. The target surface as in claim 23, wherein said porous plate has a coating of organic matrix.

25. The target surface as in claim 24,wherein said coating further comprises nitrocellulose.

26. The target surface as in claim 23, wherein said porous plate is formed of stainless steel.

27. The target surface as in claim 23, wherein said organic matrix comprises α-cyano-4- hydroxycinnamic acid.

28. The target surface as in claim 23, wherein said organic matrix comprises 2-(4- hydroxyphenylazo)-benzoic acid.

29; A high vacuum stage for procuring biopolymers attached to a membrane and depositing them on a porous plate target surface, comprising: a housing having at least one aperture therein for holding said porous plate target surface; said aperture having a depth so as to form a chamber beneath said porous plate; said chamber being connected to a vacuum source to create a vacuum in said chamber, so that when said membrane is placed on said target surface, said vacuum operable to remove the biopolymers attached to the membrane to the target surface.

30. The vacuum stage as in claim 29, wherein the pressure at the porous plate surface when in operation is at least about 30 inches Hg.

31/ The vacuum stage as in claim 29, wherein the pressure between the porous plate surface and said lid is no greater than. about 3 inches Hg.

32. The vacuum stage as in claim 29, wherein said stage is covered with a gasket having at least one aperture corresponding to the aperture in said stage.

33. The vacuum stage as in claim 33, wherein said gasket is made of silicone.

34. The vacuum stage as in claim 29, wherein said stage is made of aluminum.

35. A buffer system for conducting discontinuous polyacrylamide gel electrophoresis, comprising:

(a) an anode buffer containing glycerol and a salt concentration of 0.35 M to 0.4 M;

(b) a cathode buffer containing glycerol and a salt concentration of between 0.35 M to 0.4M;

(c) and a running buffer having a concentration of sodium dodecyl sulface greater than 0.1%.