WO1999005309A1

WO1999005309A1 - Multiple target screening of molecular libraries by mass spectrometry

Info

Publication number: WO1999005309A1
Application number: PCT/US1998/015112
Authority: WO
Inventors: Yinliang F. Hsieh
Original assignee: Millennium Pharmaceuticals Inc
Current assignee: Millennium Pharmaceuticals Inc
Priority date: 1997-07-23
Filing date: 1998-07-22
Publication date: 1999-02-04
Anticipated expiration: 2000-01-23
Also published as: AU8414798A

Abstract

The invention is based on the discovery that mass spectrometry (MS) can be used for screening a library of drug candidates for activity against multiple targets simultaneously. The new methods allow both the drugs and the targets involved to be rapidly and accurately identified.

Description

MULTIPLE TARGET SCREENING OF MOLECULAR LIBRARIES BY MASS SPECTROMETRY

Background of the Invention The invention relates to methods for high throughput screening of molecular libraries by mass spectrometry . Specifically, the invention relates to methods for simultaneously identifying compounds that inhibit multiple enzyme targets.

As technology improves researchers' ability to identify biological targets for fighting disease, the demand will continue to increase for high throughput methods of screening compounds that can bind to these targets .

Existing methods for screening molecular libraries allow identification of ligands that bind to a receptor molecule. For example, Hsieh et al . , Molecular Diversi ty, 2_..189-196 (1997), discloses a method for screening a library of potential ligands. In this procedure, the receptor of interest is first mixed with the library, and the resulting mixture is subjected to size exclusion chromatography (SEC) to separate complexes from unbound compounds. The ligands are then eluted from their receptors by reverse phase chromatography (RPC) and subjected to electrospray ionization mass spectrometry (ESI -MS) to determine the molecular weight of the ligands that were bound. Because it is necessary to elute the ligand from the receptor prior to MS analysis, this method is applicable only to the screening of one receptor at a time. The method would not be suitable if multiple receptors were present, as the determination of which ligand was bound to which receptor prior to elution would be difficult. Furthermore, if the ligand-receptor interaction were instead covalent, it would be difficult to identify any of the ligands because of the much larger molecular weight of the associated receptors. Another screening method is described in Nedved et al . (Anal . Chem . , 8_.:4226-4236 , 1996), in which a mixture is formed that contains an antibody and a library of drug candidates. In this system, the mixture is passed through a protein G immunoaffinity column to remove unbound drugs, and then the non-covalent antibody/drug complexes are eluted from a reverse phase column and subjected to ESI -MS.

Summary of the Invention The invention is based on the discovery that mass spectrometry (MS) can be used to screen a library of drug candidates for activity against multiple targets simultaneously. For example the new methods can be used to screen for antagonists of enzyme catalyzed reactions. The new methods allow both the drugs and the targets, e.g., enzymes, to be rapidly and accurately identified.

In general, the invention features a method of screening a molecular library of compounds for individual compounds that affect (e.g., inhibit) the ability of one or more of a plurality of target molecules (e.g., enzymes or other polypeptides) to catalyze the conversion of corresponding peptide substrates (e.g., peptides, polypeptides, or proteins) to products, or the reaction of a substrate with a target molecule to yield a product . The method includes the steps of combining each compound in the library with the plurality of target enzymes to form a set of inhibition mixtures; adding to each of the inhibition mixtures a peptide substrate corresponding to each target enzyme to form a set of experimental mixtures (e.g., by mixing the components at the time of use); obtaining a control mixture including the plurality of target enzymes and the corresponding peptide substrates; sampling the control mixture and the set of experimental mixtures at a first time point to obtain control aliquots and experimental aliquots; analyzing the control and experimental aliquots using a mass spectrometer to produce mass spectra (i.e., qualitatively or quantitatively) ; and comparing the mass spectrum of the control aliquot to the mass spectra of the experimental aliquots to detect individual compounds in the library that affect (e.g., inhibit) the ability of one or more of the plurality of target molecules (e.g., enzymes) to catalyze the conversion of corresponding peptide substrates to products.

A "corresponding" substrate is a substrate (e.g., a peptide or protein) that normally reacts with, or is converted into a product by, a given target molecule, e.g., an enzyme, under control conditions. Second aliquots of the control and experimental mixtures can be obtained at a second time point.

In a specific embodiment, the new method can be used to screen the molecular library of compounds for individual compounds that inhibit the ability of only one of the plurality of target enzymes to catalyze the conversion of a corresponding peptide substrate to a product, and that do not inhibit the ability of the remaining target enzymes to catalyze the conversion of the corresponding peptide substrates to products. Each of the reaction mixtures can include a plurality of synthetic, semi-synthetic, or natural compounds. The aliquots can be chromatographed prior to the analyzing step. The mass spectrometer can be a matrix-assisted laser desorption ionization (MALDI) mass spectrometer, a electrospray ionization (ESI) mass spectrometer, a liquid chromatograph mass spectrometer (LC-MS; can be quantitative), a liquid chromatograph tandem mass spectrometer (LC-MS/MS; can be quantitative) , a capillary electrophoresis mass spectrometer (CE-MS; can employ microchannel chips) , or a capillary electrophoresis tandem mass spectrometer (CE-MS/MS; can employ microchannel chips) .

If a MALDI mass spectrometer is used, the method additionally includes the steps of combining a MALDI matrix with the aliquots from the control mixture to form MALDI control samples; combining a MALDI matrix with the aliquots from the experimental mixtures to form MALDI library samples; and analyzing the MALDI control samples and MALDI library samples using the mass spectrometer to produce mass spectra. The MALDI matrix can be α-cyano-4- hydroxycinnamic acid (αCHCA) or 6-aza-2-thiothymine (ATT) , for example.

In the present context, a "library" is a heterogeneous multiplicity of compounds that can be combined together in a single mixture, divided into several mixtures, or separated into discrete, single- component fractions. A library can include any compounds, including, but not limited to, oligopeptides (e.g., derived from natural or synthetic amino acid monomers, or both) , polypeptides, proteins, peptide nucleic acids (PNA) , oligonucleotides (e.g., natural or synthetic DNA or R A) , small molecules (i.e., including any natural products or synthetic compounds) , or any combination of these components. A library can be dissolved or suspended in a solvent, buffer, or other carrier, or can exist neat. A library can also include a natural product, such as a bodily fluid, organic tissue, or secretion of an organic tissue, including blood, skin, urine, sweat, milk, liver tissue, kidney tissue, hair, thymus tissue, pancreatic tissue, brain tissue, a cell culture (e.g., bacterial, fungal, or yeast; the culture can additionally include a growth medium) , a cell extract, the supernatant of a centrifuged tissue suspension, plant matter, or other organic matter; the tissue can be processed or can be used in a natural state. The components of a library can be soluble or insoluble in a given screening mixture. The library can include 1, 10, 100, or more compounds. Additionally, compressed libraries can be used. The target molecules can be enzymes (e.g., proteases, peptidases, phosphatases, exonucleases, endonucleases, glycosidases, transferases, kinases, and polymerases) , or other proteins or polypeptides. "Multiple target screening" refers to the simultaneous presence of more than one target molecule in an experimental reaction mixture. In each screening reaction, there can be a single compound or a multiplicity of compounds that are screened against the plurality of target molecules. The target molecule can be any molecule (such as an enzyme) that catalyzes the conversion of a corresponding substrate to a product, or that reacts with a corresponding substrate to yield a product .

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All publications, patent applications, patents, technical manuals, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present application, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

Advantages of the new methods include the ability to determine the specific location on the target where the compound binds, the capacity of the new methods for high throughput screening, and the ability to simultaneously screen a library of compounds against multiple targets. Additionally, because both the substrates and products of enzymatic reactions are detected in the new methods, it is not a problem if multiple compounds from the library inhibit the same target; the library can be deconvoluted later to determine which compounds are acting as inhibitors.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Brief Description of the Drawings Fig. 1 is a mass spectrum of a substrate mixture containing angiotensin I, and substrates for N-myristyl- transferase (NMT) and protein tyrosine kinase (PTPase) . Fig. 2 is a mass spectrum of a reaction mixture that included both the substrate mixture of Fig. 1 and an enzyme mixture containing angiotensin converting enzyme (ACE) , NMT, and PTPase, sampled 15 minutes after mixing. Fig. 3 is a mass spectrum of the reaction mixture of Fig. 2 with ammonium vanadate added, sampled 15 minutes after adding the enzyme/inhibitor mixture to the substrate mixture.

Fig. 4 is a mass spectrum of the reaction mixture of Fig. 2 with lisinopril added, sampled 15 minutes after adding the enzyme/inhibitor mixture to the substrate mixture .

Fig. 5 is a mass spectrum of the reaction mixture of Fig. 2 with N-myristyltransferase inhibitor (NMTI) added, sampled 15 minutes after adding the enzyme/inhibitor mixture to the substrate mixture.

Fig. 6 is a mass spectrum of the reaction mixture of Fig. 2 with ammonium vanadate, lisinopril, and NMTI added, sampled 15 minutes after adding the enzyme/inhibitor mixture to the substrate mixture.

Detailed Description The invention is a highly effective mass spectrometric method for the identification and characterization of compounds that bind to a target molecule, such as an enzyme, present in a mixture of targets. The method is based on the detection of substrates and products of an enzymatic reaction catalyzed by the target enzymes, or the substrates and products of any reaction affected by the target molecules, and not necessarily the aforementioned compounds themselves.

General Methodology A typical procedure for carrying out the new methods is as follows.

A mixture of target enzymes and their respective substrates are obtained. The enzymes are combined with the members of a library of compounds in containers (e.g., a combinatorial drug library or a library of natural products, contained, for example, in the wells of a 96-well plate) . The entire mixture of enzymes is added to each container, while the individual members or groups of members of the library are added to distinct containers, in the same or different concentrations. It can be desirable for high throughput screening to include multiple members of the library in each of the wells. Each well can contain 1, 10, 20, 50, 100, or even 500 or more members. Thus, for example, a library of up to 9600 compounds can be screened against an enzyme library in a single 96-well plate by loading each well with 100 library members. Finally, the substrates are added to each container to form reaction mixtures. At least one reaction mixture is set up with the target enzyme and substrate mixtures, but without the inclusion of any members of the molecular library. Such a mixture is referred to as a control mixture and is used to monitor the native activity of the enzymes (or other target molecules) in the absence of any compounds other than the substrates .

Samples can be drawn from the reaction mixtures at various times (e.g., at 5, 10, and 20 minutes) after the addition of the substrates to the enzymes. The samples can be quenched, e.g., by addition of a low pH solution, and analyzed by mass spectrometric analysis.

Matrix-Assisted Laser-Desorption Ionization Mass Spectrometry (MALDI MS) Because of ease of sample preparation, tolerance of high salt concentrations, substantial mass range, resolution, and sensitivity, matrix-assisted laser- desorption ionization (MALDI) is a suitable mass spectrometric technique for use with the new methods . For MALDI analysis, the reactions proceeding in the sample are quenched, for example, by addition of a low-pH matrix compound, such as a 10 mg/ml solution of a- cyano-4-hydroxycinnamic acid (αCHCA) in a 1:1 mixture of methanol and water or a 1:1 mixture of acetonitrile and water. The matrix includes a conjugated, amphoteric molecule that not only quenches the reactions, but can also absorb laser energy and transform it into excitation energy (e.g., via electron-phonon coupling) in the molecules trapped within the matrix, disintegrating, and desorbing at least some of the top molecular layers of the sample.

Other low-pH matrix molecules can also be used, including: sinapinic acid, 2 , 5-dihydroxybenzoic acid (2,5-DHB), 2- (4-hydroxyphenylazo)benzoic acid (HABA) , anthranilic acid, 3-hydroxypicolinic acid, picolinic acid, and nicotinic acid. These matrix molecules are better suited for different ones of peptides, proteins, and oligonucleotides . For analysis of peptides, for example, α-CHCA is suitable. For DNA or RNA substrates, 3-hydroxypicolinic acid and picolinic acid are both satisfactory.

Non-covalent complexes, held together via hydrogen bonds, van der aals forces, or ionic interactions, dissociate in the presence of the low-pH MALDI matrices. Neutral MALDI matrices, such as 6-aza-2-thiothymine (ATT), dithranol, and 2 , 4 , 6-trihydroxyacetophenone (THAP) , however, can be used to analyze non-covalent complexes . A portion of the quenched reaction mixture is transferred to a metal MALDI plate (also called a "MALDI probe") that fits within a MALDI MS instrument. In general, any metal plate adapted to fit into the instrument can be used. The plate can have a number of wells drilled into it to accept the quenched reaction mixture. Plates without wells can also be used; the samples can simply be spotted onto the surface of a flat, polished plate. Each MALDI plate can have from about a hundred, microliter-scale samples to several thousand, picoliter-scale samples. In some cases, the plates are precoated with matrix solution, allowing the quenching and spotting steps to be combined. Precoated plates are commercially available, for example, from Lab Connection Co. (Marlborough, MA) . MALDI has sufficiently high resolution that the isotopic composition of a sample can be monitored. This provides an internal standard that can be used for better mass accuracy in complex sample mixtures.

MALDI also promotes high throughput screening, because only 15-20 seconds are typically required per sample for data acquisition, and multiple target enzymes can be screened simultaneously. For each sample, the laser ionizes the compound and the matrix. The MALDI plate is pulsed with a current (e.g., 20 kV) for a fraction of a second (e.g., about 50 to 100 ns) . The pulse results in the desorption of at least some of the top molecular layers of the sample, which then fly toward a detector. The time of flight is generally about 100 to 200 μs. The sample is detected by the detector, which sends a signal to a computer for data collection and processing. The entire process takes about one-third of a second.

To enhance the signal intensity and the signal-to- noise ratio, the ionization/desorption process can be repeated 8, 16, 32, 64, 128, or more times for each sample. Accumulation of 32 data points, for example, would take just under 11 seconds at a rate of one scan every third of a second.

Electrospray Ionization Mass Spectrometry (ESI -MS) Other mass spectrometry ionization techniques such as electrospray, ion spray (pneumatically assisted electrospray) , and nanospray allow the formation of multiply charged gas phase macromolecular ions directly from solution at atmospheric pressure via protonation and ion evaporation. These techniques, for example, provide a rapid, straightforward method for the analysis of target molecules with bound drugs.

The procedure used for screening with ESI is similar to that used for MALDI. Unlike MALDI, however, ESI does not require the use of a matrix compound.

Aliquots from the control and experimental mixtures can be injected directly into the ESI-MS. Although not necessary, it can be desirable to quench the reactions prior to analysis. Quenching can be accomplished, for example, by using a solution of formic acid or trifluoroacetic acid.

ESI-MS has the additional advantage of being highly quantitative and ESI data is sufficiently accurate to be used, for example, in FDA drug trials.

Substrates and Enzymes

In general, MALDI is suitable for monitoring molecules having molecular weights over 500; the MALDI matrix molecules currently used all have a molecular weight lower than 500 and thus do not interfere with the mass-to-charge ratio (m/z) of larger molecules. Apart from the requirement that the analyte be larger than m/z 500, nearly all compounds are amenable to analysis with MALDI . Examples of such compounds include the substrates of numerous enzymes, such as glycosidases, proteases, nucleases, phosphatases, kinases, and other transferases . Some specific examples of substrates include angiotensin I, PTPase substrates, and NMT substrates. There are no restrictions on the lower limit of molecular weight for ESI analysis, since a matrix is not necessary. Thus, in addition to the substrates listed above, small molecules can also be monitored with ESI.

ESI-MS is an excellent quantitation method for monitoring the appearance and disappearance of substrates, products, and even noncovalent intermediates for performing kinetic analyses of enzymatic reactions.

High Throughput Screening

The beam of the MALDI laser typically has a diameter of 100 to 200 μm. Consequently, the sample wells on a MALDI plate can be of a similar size, and are generally in the range of 100 μm to about 2 mm.

MALDI plates having up to 100, 4,000, or even more samples, e.g., on a 5 cm x 5 cm sample area can be used. The sample volume required for this high-density spotting is on the order of microliters (μl; e.g., for 100 sample plates) down to picoliters (pi; e.g., for 4,000 well plates) . As described in greater detail below, the spotting of so many of such tiny samples onto a single plate is facilitated by a robot that can transfer the samples from the reaction containers to the MALDI plate.

An advantage of spotting several thousand samples onto a single plate is that hundreds of thousands of compounds can be screened against multiple targets in several hours. The rate of throughput is limited only by the rate of automated sample introduction, as the MALDI mass spectrometer can accumulate data points for each sample and then prepare to analyze the next sample, all within just a few seconds. Commercial mass spectrometry software packages to control both lasing and the movement of the laser are available from many of the manufacturers of MALDI mass spectrometers (e.g., Voyager™ from PerSeptive Biosystems, Inc., Framingham, MA). The use of compressed libraries, wherein each well of the reaction plate could have up to 50 compounds, or more, also enables faster screening.

The new methods provide an advantage over other methods in the screening of compressed libraries. For example, the compounds contained in whichever wells yield multiple "hits" can be screened independently to identify exactly which compounds are responsible for the activity. In some cases, several compounds in a single reaction mixture can have activity against the same target . Unlike screening methods that rely on the detection of a ligand after it is eluted away from its receptor, the new methods allow identification of all compounds in a mixture that bind to the target, not just the strongest or the fastest. Robotics

A robot can be used to introduce the samples onto the MALDI plate. Suitable robots are available from, for example, Tecan U.S., Inc. (Durham, NC) , and can be controlled using available software and standard techniques. Robots can also be custom made to fit a particular application (e.g., for high density spotting. For low density spotting (i.e., about 100 samples per plate) each sample volume can be about 1 μl . A Tecan robot (Genesis™ Model 1020) can be used to transfer a solution containing multi-enzyme targets to a 96-well sample mixing plate. A compressed combinatorial library or natural library (e.g., from bacterial or fungal fermentation and extraction) is selected and robotically mixed with the multi-enzyme target solution in the mixing plate. A multi-enzyme substrate mixture is transferred and combined with the enzyme-drug library in the mixing plate by the robot. A sample of the enzyme-library- substrate mixture is taken out and mixed with a MALDI matrix solution in a second 96-well plate by the robot. The robot dispenses a 1 μL aliquot of the sample mixture with the matrix solution onto a MALDI plate.

For high density spotting (i.e., about 4000 samples per MALDI plate) each sample volume is on the order of one picoliter. In general, a robot capable of spotting several thousand samples onto a MALDI plate for qualitative analysis requires an error tolerance of 100 to 200 μm. The robot transfers a solution containing multi-enzyme targets to a 96-well sample mixing plate. A compressed combinatorial library or natural library is selected and robotically mixed with the multi-enzyme target solution in the mixing plate. A multi-enzyme substrate mixture is transferred and combined with the enzyme-drug library in the mixing plate by robot. The robot spots a sample of the enzyme-drug-substrate mixture onto a MALDI plate precoated with the matrix solution.

Deconvolution and Structural Analysis Those reaction mixtures that yield mass spectra differing from the control mixture spectra can be further analyzed.

The new methods can be combined with enzymatic digestion (e.g., with trypsin, chymotrypsin, or subtilisin) to determine the specific location of the binding site if a drug covalently binds to a target. For example, the mass spectrum obtained after trypsin digestion of a target-drug complex can be compared with the corresponding mass spectrum obtained after digestion of the target itself. The peaks that differ between the two spectra indicate the precise segment to which the drug binds .

To determine the specific amino acid to which a drug binds a proteinaceous target, the sample can be analyzed with, for example, ESI liquid chromatography tandem mass spectrometry (LC-MS/MS) . See, for example, Anal . Chem . , 68.:455-462 (1996).

Time course studies can be carried out with the new methods. For example, time course studies can be used to evaluate the product-to-substrate ratio at various time points, relative to a control sample, to provide an understanding of the kinetic inhibition of the enzymes by the drugs. The corresponding inhibition constants can be thereby determined. Time course studies can also be used to ascertain the time required for partial or total inhibition. Finally, the pharmacological kinetic behavior of the drug-enzyme system can be determined via time course studies. Similarly, the new methods can be used in concentration/ effect studies. By collecting mass spectra corresponding to various quantities of a single compound in a series of reaction mixtures, dosage information can be derived.

The invention will be further described in the following examples, which do not limit the scope of the invention described in the claims.

EXAMPLES Example 1 - Screening of a Molecular Library for Inhibitors of Angiotensin Converting Enzyme (ACE) by MALDI Angiotensin converting enzyme (ACE) is an important target in research aimed at the design of anti- hypertensive drugs. This example illustrates a method for identifying an inhibitor of ACE from a molecular library. The procedure employed in this example can also be extended to the simultaneous screening of multiple targets .

Enzymatic reactions: 2 μl of ACE (305 fmol) in a

20 mM Tris buffer solution (pH 7.5) was added to each of

3 wells of a 96 well plate. To the first well was added a mixture containing 60 pmol of each of the following nine compounds (molecular library) : ampicillin, lincomycin, trimethoprim, rapamycin, captopril, piperacillin, penicillin G, bisindolylmaleimide, and vancomycin, to determine whether any or all of these compounds interferes with the enzymatic reaction. No inhibitor was added to the second well; this well was used as a control well. To the third well was added 60 pmol lisinopril, a known inhibitor of ACE (positive control) . 20 μl of angiotensin I (4 pmol/μl) was then added to each of the 3 wells.

Sample preparation: 1 μl samples were removed from each well at various time points (i.e., 0, 5, 10, 20, 30, and 50 minutes) . 1 μl of a MALDI matrix solution (1.0% αCHCA and 0.1% trifluoroacetic acid (TFA) in 50% acetonitrile/50% water solvent mixture) was added to the sample, thereby quenching the reaction. 1 μl of the resulting solution was applied with a micropipettor to the surface of a stainless-steel 100-well MALDI plate (PerSeptive Biosystems, Inc., Framingham, MA). The plate was then air-dried and inserted into the sample introduction port of the Voyager™ Elite Biospectrometry™ MALDI time-of-flight (TOF) mass spectrometer (PerSeptive Biosystems, Inc., Framingham, MA). MALDI -TOF MS analysis: The Voyager was operated in reflector mode using the delayed extraction function mode with an accelerating voltage of 20 kV from the ion source; these conditions resulted in better resolution compared with the continuous extraction function mode. The software packaged with the Voyager was used for controlling the laser and for collecting and analyzing the data.

The Voyager™-Elite Biospectrometry™ Workstation with Voyager-DE™ technology can be operated in continuous or delayed extraction (DE) mode. In the continuous extraction mode, accelerating voltage is continuously applied; a potential gradient exists when the sample is ionized; and ions are immediately accelerated. In the delayed extraction mode, a potential gradient does not exist when the sample is ionized (i.e., the sample plate and grid are at similar potentials) ; the energy of the ions is stabilized before acceleration; and accelerating voltage is pulsed after the potential gradient is applied and the ions are accelerated. Voyager-DE™ (delayed extraction) technology allows control of field strength and delays in the acceleration of ions, thereby minimizing ion energy loss due to collision and extraction field strength. In the Voyager- DE™ system, ions are formed in a weak electrical field, and then extracted by applying a high voltage pulse after a predetermined time delay.

Delayed extraction of ions improves performance over that obtained with continuous extraction of ions by improving resolution and mass accuracy. It also improves the quality of MALDI mass spectra by suppressing matrix background, reducing chemical noise, and minimizing the effect of laser intensity on overall performance.

The samples were ionized with 128 pulses from a 337 nm nitrogen laser, using the Voyager's oscilloscope to rapidly acquire data. The resulting signal was digitized at 2 GHz frequency (i.e., 0.5 ns resolution) and the data was downloaded to the Voyager's data system. The peptides were isotopically resolved. Results : The MS spectrum for the "control" sample in the second well showed the disappearance of the angiotensin I (DRVYIHPFHL; SEQ ID NO:l; MW 1296.68) and the concomitant appearance of the product, angiotensin II (DRVYIHPF; SEQ ID NO : 2 ; MW 1046.52). In the mass spectrum corresponding to an aliquot taken from the control well just after combining the enzyme and substrate, the predominant species was the angiotensin I, at m/z 1296.68. Smaller, but still finite, quantities of other species were also present at m/z 931.47, 1046.52, and 1181.65. The peak at 1181.65 corresponds to an impurity present in the angiotensin I solution (RVΥIHPFHL; SEQ ID NO: 3) ; the impurity is identical to angiotensin I (SEQ ID NO:l) except that it lacks the N-terminal aspartic acid residue. The peak at 1046.52 is angiotensin II (SEQ ID N0:2), as described above. The peak at 931.47 (RVYIHPF; SEQ ID NO: 4) is the conversion product of the impurity; it is thus identical to angiotensin II (SEQ ID NO: 2), except that it lacks the N-terminal aspartic acid residue. Aliquots were taken from the control well at 0, 5, 10, 20, 30, and 50 minutes. In the mass spectra of these aliquots, the intensity of the angiotensin II (SEQ ID NO: 2) peak increased over the first 20 minutes as the intensity of the angiotensin I (SEQ ID NO:l) peak declined. At 20 minutes, nearly all of the starting material had been converted. At 30 and 50 minutes, there was little change from the 20 minute time point. In the well containing the experimental library, the transformation was slowed significantly.

In samples taken from the well containing the lisinopril, no conversion was detected at all.

In another set of experiments, the inhibitors were screened individually against ACE. In the samples containing captopril, sampled at 0, 5, 20, and 30 minutes, the transformation was significantly slowed by the captopril relative to the reaction in the control well. The intensity of the angiotensin I peak decreased and the intensity of the product peaks increased over the 30 minute period, but approximately 50% of the starting material remained after 30 minutes. Thus, the reaction was partially inhibited by captopril relative to the control sample .

The reaction mixtures containing ampicillin, lincomycin, trimethoprim, rapamycin, piperacillin, penicillin G, bisindolylmaleimide, and vancomycin each yielded time course plots similar to those of the control sample. These compounds apparently did not significantly inhibit ACE. The experiments were repeated for captopril and lisinopril, using various quantities of the compounds in the reaction mixtures: 1, 10, 20, 50, 100, 250, 500 pmol, and 1 nmol . Mass spectra were acquired 30 minutes after addition of the substrates to the enzymes. As the concentration of captopril was increased, the quantity of the products formed decreased. Thus, 1 pmol captopril did not inhibit the reaction significantly, whereas 10 to 500 pmol did cause partial, steadily increasing inhibition. Using 1 nmol captopril caused nearly complete cessation of the reaction.

For lisinopril, 1 pmol was sufficient to substantially inhibit the reaction.

Although this example describes a method for identifying compounds that inhibit only a single target enzyme, the same procedure can be used to screen multiple targets, as described in the following example.

Example 2 - Simultaneous Screening of Multiple Targets by MALDI MS Several compounds were individually screened against three target enzymes: ACE, N-myristyltransferase (NMT) , and protein tyrosine kinase (PTPase) . It would be desirable to identify an inhibitor of NMT from, for example, Candida albicans or Cryptococcus neoformans that does not also inhibit human NMT, as such inhibitors could be non-toxic medications against yeast infections. PTPase can be a target in the development of anti- obesity, or weight control, drugs.

Buffer: A buffer solution was prepared (pH 7.4), containing 20 mM Tris HC1 and 3 mM dithiothreitol .

Enzyme mixture : A 2:1:1 mixture of ACE (final concentration, 305 fmol/μl) , PTPase (0.39 fm/μl) , and NMT (1.98 pmol/μl) was prepared in the buffer solution. Substrate mixture : A 2:2:1:1 mixture of angiotensin I (SEQ ID NO:l; 2 pmol/μl), PTPase substrate (TRDIXETDYYRK (X = Y-P0₃H₂) ; SEQ ID NO : 5 ; 4 pmol/μl), the peptide GNAASARR-NH₂ (SEQ ID NO: 6) (3 pmol/μl), and myristyl-CoA (3 pmol/μl) was prepared. Reaction: 4 μl of the enzyme mixture and 30 μl of the substrate mixture were combined in each of six wells of a 96-well plate. Potential inhibitors were added to the wells according to the following scheme:

At 15 minutes and 30 minutes, 2 μl samples were taken from each well and mixed with 2 μl matrix solution. 1 μl of each of the resulting mixtures was applied to a stainless-steel MALDI plate. The samples were analyzed by MALDI -TOF, as described in Example 1. Results :

Fig. 1 is a mass spectrum (MS profile) of the mixture of the three substrates (myristyl-CoA is not a substrate, but a cofactor of NMT) , at mass-to-charge ratios (m/z) of 801.35 (peptide; SEQ ID NO:6), 1296.68 (angiotensin I; SEQ ID NO:l), and 1702.88 (PTPase substrate; SEQ ID N0:5). The peaks at m/z 1181.66 and 1587.81 corresponded, respectively, to impurities introduced along with the angiotensin I (SEQ ID NO: 3) and the PTPase substrate (TRDIYETYYRK; SEQ ID NO: 8) .

Fig. 2 is the MS profile of the reaction mixture of well 1 after 15 minutes. All three substrates appeared to have been fully consumed. Four new peaks developed, corresponding to the products formed by reaction of the enzymes with the substrates and impurities, at m/z 1011.58 (N-myristyl-GNAASARR-NH₂; SEQ ID NO:9), 931.45 (SEQ ID NO : 4 ; product formed by reaction of the impurity in the angiotensin I), 1046.51 (SEQ ID NO:2; angiotensin II), and 1622 (TRDIYETDYYRK; SEQ ID NO:10; dephosphorylated tyrosine-bearing peptide). It was also noted that not all of the angiotensin I had been consumed within the 15 minutes.

To determine the correlation between the substrates and products for the NMT, PTPase, and ACE catalyzed reactions, each of the substrates was independently analyzed by MS before and after addition of the corresponding enzymes. Thus, the NMT substrate was analyzed by MS before (m/z 801.55) and after addition of NMT. Addition of NMT caused the substrate to be converted to a product of MW 1011.78 (SEQ ID NO: 9) .

Likewise, the PTPase substrate was analyzed before (m/z 1703.05, impurities at 1588.00 and 1608.40) and after addition of PTPase. The starting materials in this case became products having molecular weights of 1623.11 (SEQ ID NO:10), 1508.04, and 1606.41, respectively. Lastly, angiotensin I was analyzed before (m/z 1296.68, impurity at 1181.63) and after (m/z 1046.49 and 931.44, respectively) addition of ACE.

As shown in Fig. 3, the MS profile of the reaction mixture in well 2 after 15 minutes showed that ammonium vanadate was not a very potent inhibitor of any of the three enzymes. The reactions of all of the substrates proceeded to about the same extent as the reactions in the control well shown in Fig. 2. The mass spectrum for the reaction of well 3 after 15 minutes is shown in Fig. 4. Lisinopril completely inhibited the enzymatic activity of ACE (no peaks at 931.48 or 1046.52), but had virtually no effect on the conversion of the NMT and PTPase substrates to the corresponding products . The mass spectrum in Fig. 5 shows that NMTI inhibited the activity of NMT but had little effect on the activity of either ACE or PTPase.

The combination of the three inhibitors in well 5 had an effect that was similar, but not identical, to the independent inhibitors. Fig. 6 indicates that there was no conversion of the ACE and NMT substrates (m/z 801, 1296, and 1181) , whereas the PTPase substrates (m/z 1702) reacted completely. The impurity introduced with the PTPase substrate (m/z 1587) also reacted completely, producing a product of m/z 1507 upon loss of a phosphate group .

Thus, the new methods were demonstrated to be successful for screening inhibitors against multiple targets simultaneously.

Example 3 - Multiplex LC-MS Drug Screening

Liquid chromatography followed by mass spectrometry (LC-MS) is used for the kinetic analysis of enzymatic reactions in real time. The products and substrates for drug screening are quantitatively monitored by LC-MS or LC-MS/MS.

To quantitate the multiple enzymatic reactions of targets, a calibration curve of each free substrate is constructed by injecting various concentrations of the substrates into the LC-MS, and then calculating the area under the total ion chromatographic peaks resulting from the LC-MS analysis. The calibration curve is used to determine the concentration of each component substrate in the experimental reaction mixtures. In the LC-MS screening of a library against NMT, the substrate concentration remains much higher with NMTI than after screening with other drug libraries. In the screening of a library against multiple targets, the concentrations of all of the substrates and products are monitored simultaneously and quantitated.

Five hundred reaction mixtures are prepared, each containing NMT, ACE, and PTPase, the respective substrates of these enzymes (SEQ ID NOS:6, 1, and 5), and a library of 500 sets of compressed mixtures with 50 test drug candidates in each set (i.e., a total of 25,000 compounds are screened) . A control mixture is also prepared without any drug candidates. The drug library reaction mixtures, as well as the control mixture, are sampled and analyzed after 30 minutes. Typically, if none of the test drug candidates are inhibitors, the substrates will have been completely converted to the corresponding products (SEQ ID NOS : 9 , 2, and 10) thirty minutes after mixing the target/library mixture with the substrate mixture. However, experimental reaction mixtures ("successful mixtures") in which no angiotensin II or PTPase product is observable at the half hour time point, but in which the NMT substrate has been completely converted, contain one or more test drug candidates that each inhibit ACE and/or PTPase, but not NMT.

To determine which one or more of the 50 compounds in the successful mixture was responsible for the inhibition of ACE and which was responsible for the inhibition of PTPase, a second experiment is conducted, wherein 50 reaction mixtures are prepared with the three enzymes, their substrates, and the individual members of the successful mixture. A second control mixture is prepared. This time, most of the reaction mixtures yield results that are the same as in the control mixture.

The reaction mixtures that contain the PTPase and NMT conversion products, but no angiotensin II, contain a drug candidate that inhibits only ACE. If only some angiotensin II is observed, e.g., if about 60% of the angiotensin I remains, then the ACE inhibitor is a weak inhibitor. Other reaction mixtures will give additional information. Because the identity of the individual compounds in each reaction mixture are known, compounds that weakly and strongly inhibit ACE and a compound that moderately inhibits PTPase are identified from the molecular library.

Other Embodiments It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims. For example, although screening molecular libraries for compounds that inhibit enzymes is an important application of the new methods, these methods can also be used to identify compounds that disrupt the binding of ligands to their receptors, or antigens to antibodies.

Additionally, one of the binding partners can be immobilized on the MALDI plate by coating the plate with a layer of gold and covalently attaching the binding partner to the gold surface. For example, an antibody, antigen, ligand, oligonucleotide, or oligopeptide can be immobilized. A similar technique can also be employed for analysis of non-covalent compound-receptor interactions .

Claims

What is claimed is:

1. A method of screening a molecular library of compounds for individual compounds that inhibit the ability of one or more of a plurality of target enzymes to catalyze the conversion of corresponding peptide substrates to products, the method comprising: combining each compound in the library with the plurality of target enzymes to form a set of inhibition mixtures; adding to each of the inhibition mixtures a peptide substrate corresponding to each target enzyme to form a set of experimental mixtures; obtaining a control mixture comprising the plurality of target enzymes and the corresponding peptide substrates; sampling the control mixture and the set of experimental mixtures at a first time point to obtain control aliquots and experimental aliquots; analyzing the control and experimental aliquots using a mass spectrometer to produce mass spectra; and comparing the mass spectrum of the control aliquot to the mass spectra of the experimental aliquots to detect individual compounds in the library that inhibit the ability of one or more of the plurality of target enzymes to catalyze the conversion of corresponding peptide substrates to products.

2. The method of claim 1, wherein the molecular library of compounds is screened for individual compounds that inhibit the ability of only one of a plurality of target enzymes to catalyze the conversion of a corresponding peptide substrate to a product, and that do not inhibit the ability of the remaining target enzymes to catalyze the conversion of the corresponding peptide substrates to products.

3. The method of claim 1, wherein said mass spectrometer is a matrix-assisted laser desorption ionization (MALDI) mass spectrometer, the method further comprising: combining a MALDI matrix with the control aliquots to form MALDI control samples; combining a MALDI matrix with the experimental aliquots to form MALDI library samples; and analyzing the MALDI control samples and MALDI library samples using said mass spectrometer to produce mass spectra.

4. The method of claim 1, wherein the molecular library comprises oligopeptides .

5. The method of claim 1, wherein the molecular library comprises oligonucleotides .

6. The method of claim 1, wherein the molecular library comprises a combinatorial library.

7. The method of claim 1, wherein the molecular library comprises natural products.

8. The method of claim 1, wherein the molecular library comprises organic tissue.

9. The method of claim 1, wherein the molecular library comprises a bodily fluid.

10. The method of claim 1, wherein the target enzymes are selected from the group consisting of proteases, peptidases, phosphatases, exonucleases, endonucleases, glycosidases, transferases, kinases, and polymerases.

11. The method of claim 3, wherein the MALDI matrix comprises α-cyano-4-hydroxycinnamic acid (αCHCA) .

12. The method of claim 3, wherein the MALDI matrix comprises 6-aza-2-thiothymine (ATT) .

13. The method of claim 1, wherein each of the experimental mixtures comprises a plurality of compounds.

14. The method of claim 1, wherein the analysis is quantitative .

15. The method of claim 1, wherein second aliquots of the control and experimental mixtures are obtained at a second time point to provide second control and second experimental aliquots.

16. The method of claim 1, wherein the mass spectrometer is a electrospray ionization (ESI) mass spectrometer.

17. The method of claim 1, wherein the aliquots are chromatographed prior to the analyzing step.

18. The method of claim 17, wherein the mass spectrometer is a liquid chromatograph mass spectrometer (LC-MS) .

19. The method of claim 17, wherein the mass spectrometer is a liquid chromatograph tandem mass spectrometer (LC-MS/MS) .

20. The method of claim 1, wherein the mass spectrometer is a capillary electrophoresis mass spectrometer (CE-MS) .

21. The method of claim 1, wherein the molecular library comprises a mixture of at least about 10 compounds .

22. The method of claim 1, wherein the molecular library comprises a mixture of at least about 100 compounds .

23. The method of claim 1, wherein the molecular library is a compressed combinatorial library.