WO2000058735A2

WO2000058735A2 - High throughput screening apparatus and methods for arraying microparticles

Info

Publication number: WO2000058735A2
Application number: PCT/US2000/008314
Authority: WO
Inventors: Amit Kumar
Original assignee: Optigon Technologies
Priority date: 1999-03-31
Filing date: 2000-03-29
Publication date: 2000-10-05
Also published as: AU4180500A; WO2000058735A3

Abstract

A device that directs the rapid arraying of a determinate number of microparticles, such as polymeric beads, from bulk suspension, and that is particularly suited for high throughput screening of combinatorial libraries synthesized on polymer microbeads, is presented. The device further dispenses, in highly parallel fashion, predefined small volumes of a fluid. In a preferred embodiment, the device has disposed upon its surface a rectangular array of microparticle retention elements, each comprising a fluid retention surface and a blind depression. Microparticles are arrayed and a fluid is dispensed by sequential wetting and dewetting of the active surface. Corresponding methods are presented.

Description

High Throughput Screening Apparatus and Methods for Arraying Microparticles

FIELD OF THE INVENTION

The present invention relates to apparatus and methods for partitioning microparticles from bulk suspension. In particular, the present invention relates to apparatus and methods for the rapid, uniform, and highly parallel arraying of microparticles for high throughput screening of combinatorial chemical libraries synthesized on polymeric microbeads.

BACKGROUND OF THE INVENTION

Increasingly, the process of new drug discovery is being automated. Accordingly, the speed with which new pharmaceutical agents can be developed and brought to market now depends in large measure on the speed with which synthesis and then testing of new chemical agents can be automated, integrated, and miniaturized.

Synthesis itself presents few problems. Recent advances in combinatorial synthesis chemistries now permit millions of compounds to be synthesized rapidly and efficiently on polymeric microbeads. In the split-pool (split and mix) approach, each bead uniformly displays a unigue, tethered small molecule ligand.

For analysis, such compounds must then be introduced into assays that report an effect of the agent on a chosen biological target. Some assays may usefully be performed with the ligand still attached to the synthesis bead; others, however, require that the ligand be released from the bead before assay. These latter assays may be accommodated by using synthesis methods that tether the ligand to the synthesis bead by a cleavable linker; the ligand is introduced to the target still attached to the synthesis bead, with subsequent cleavage of the linker releasing the ligand to interact with the target in fluid phase. Czarnik et al . (eds.), A Practical Guide to Combinatorial Chemistry, American Chemical Soc'y (1997); Chaiken et al . (eds.), Molecular Diversity and Combinatorial Chemistry: Libraries and Drug Discovery, American Chemical Soc'y (1996).

Off-bead assays obviate the constraints on ligand-target interaction occasioned by physical attachment to the bead. Off-bead assays also approximate more closely the effects to be expected from administration of the ligand as a pharmaceutical agent. However, each bead in the library typically contains 100 pmol or less of the compound; thus, to achieve sufficient concentration in the fluid phase, the ligand must be released into a small fluid volume. And since the released ligand diffuses rapidly, these small fluid volumes must be physically segregated to prevent assays performed in parallel from cross- contaminating one another. Although tagging methods exist that permit individual microbeads to be identified after admixture in bulk suspension, biological assay still requires that the beads be assayed individually or, if serial screens are to be performed, in progressively less complex admixture. Even in initial screens, assays require predetermined small numbers of beads. High throughput assay of combinatorial libraries synthesized on polymer microbeads thus depends upon the massively parallel partitioning, or arraying, of such beads into individually identifiable and physically segregated small volume reaction vessels. Maximal efficiency is achieved when the distribution variance is small — that is, when the percentage of reaction wells containing the desired number of beads is maximized.

Most existing approaches to bead partitioning rely upon entraining the beads in a fluid stream. This permits the bead to be dispensed concurrently with the solvent required for subsequent fluid phase assay, and further permits existing fluidic devices readily to be adapted to particle arraying — so long as beads can reliably be entrained at known density in a fluid stream, methods and devices designed for automated dispensation of small fluid volumes may be adapted to distribute particles to wells.

Thus, Kirk et al . , U.S. Patent No. 5,649,576, describe a partitioning device for arraying small objects into a plurality of containers, such as individual wells of a microtiter dish, by entraining the objects in a fluid stream. The apparatus comprises: a vessel containing the small objects in fluid suspension; a tube, the proximal end of which is immersed continuously in the suspension; a mechanical member connected to the distal end of the tube, for directing gravitationally-formed fluid drops downward from the distal end of the tube into a plurality of containers positioned below the tube's distal end; a small-objects detector, adapted for detecting the small objects in the fluid within the tube, disposed near the distal end of the tube; a drop detector, attached to the member below the distal end; means for controlling flow of the suspension through the tube, preferably by pressurizing the vessel; means for discriminating between signals from the small-objects detector caused by the small objects and signals not caused by the objects; means for comparing signals from the small- objects detector caused by the small objects with signals from the drop detector; means for determining the number of objects deposited into each of the containers; means for re-positioning the distal end of the tube from one of the containers to another; and means for timing the repositioning with respect to signals from the drop detector, small objects detector, or the combination thereof.

The device described by Kirk et al . distributes beads serially into reaction wells. The dual detectors permit the regulable and determinate distribution of microparticles into the wells. Present drug discovery goals in the pharmaceutical industry, however, contemplate screening millions of compounds against scores of different targets each year, rendering such serial approach problematic. Both WO 98/52047 (Aurora Biosciences) and Astle, U.S. Patent No. 5,525,302, describe devices in which such liquid dispensation is performed in modestly parallel fashion. However, the mechanical complexity of pipetting approaches calls into question the ability of any such device reliably to perform millions of assays without failure.

Furthermore, pipetting approaches — including, more generally, dispensation by ink jet printing, Leπvmo et al . , Curr. Opin . Biotechnol .

9:615-617 (1998) — require that the dispensing tip be dimensioned to accommodate passage of the entrained microparticles. This, in turn, places constraints on the minimum size of the dispensed fluid droplet, compromising the ability freely to adjust the reagent volume, and thus ligand concentration, of the assay.

Borchardt et al . , Chemistry & Biol ogy 4:961- 968 (1997) describe an alternative approach. A mixture of beads evenly dispersed in fluid medium is slowly injected into a stream of air, forming a mist. When deposited onto a surface, the mist forms into spatially segregated nanodroplets . The average volume of the droplets is controlled by the amount of liquid applied to the surface; the fraction of droplets containing beads depends on the density of beads in the medium prior to spraying.

Two problems render this technique difficult to integrate into an efficiently automated, robotic drug discovery system. First, the technique partitions beads in a spatially nonaddressable fashion, frustrating efforts to integrate the assay step with subsequent icrofluidic manipulations . Second, because the number of beads per assay spot distributes stochastically rather than deter inately, the number of beads per droplet can only crudely be adjusted: to achieve a distribution in which a high percentage of bead- containing droplets contain a single bead, a high percentage of droplets must be formed that have no beads at all. To achieve a distribution in which 88% of bead-containing drops contained a single bead, 90% of Borchardt's droplets contained no beads.

You et al . , Chemistry & Biology 4:969-975 (1997) present a stochastic partitioning approach adapted to a spatially arrayed format. A polymeric substrate with arrays of dimple-like indentations (wells) was constructed from polydimethylsiloxane (PDMS) . Cells and beads uniformly suspended in liquid medium were applied by wetting the substrate surface uniformly; excess fluid was then drained by pipetting. Nanodroplets of uniform volume remained within the wells owing to surface tension, providing a formatted grid of spatially defined, miniaturized assay sites. As with Borchardt, however, beads distribute stochastically in this latter method. Thus, wetting- dewetting the fashioned substrate with 80 μm diameter beads suspended at a concentration of 10 mg per ml resulted in 18% of wells remaining empty, 44% containing one or two beads, 18% having three beads, and the remaining wells containing four or more beads. More diluted bead mixtures resulted in most wells containing one or two beads per well, but with a concomitant increase in the number of empty wells. More concentrated bead mixtures resulted in nearly all wells containing multiple beads.

In a related approach, Schullek et al . , Analyti cal Biochemi stry 246:20-29 (1997) fabricated a spatial array of 0.37 μl nanowells in a polyacrylonitrile butadiene styrene (ABS) plastic sheet using a high-speed drill. Beads (130 μm diameter) suspended in molten agarose were then partitioned into the wells by spreading the slurry repeatedly over the surface of the array using a straightedge tool. When 1.0 equivalents of beads were added to 1.0 equivalents of nanowells, 23% of the wells were empty, 48% contained one bead, 24% two beads, and over 5% three or more beads. When 1.5 equivalents of beads were used, the distribution was further skewed toward multiple beads per nanowell.

There thus still exists a need in the art for spatially addressable partitioning methods that obviate pipetting while concurrently providing determinate distribution of beads with low well-to-well variance.

Burbaum, Drug Discov. Today 3(7):313-318 (1998), mentions a high density microtiter plate (9600 wells of 0.2 μl volume) the wells of which have a pyramidal shape: according to Burbaum, the well shape, together with specially fashioned intrawell boundaries, segregates beads into defined wells, apparently by mechanical sorting. No further data are given.

Smethers et al . , U.S. Patent No. 5,382,512, similarly describes an assay device that captures particle reagents mechanically by virtue of well shape, albeit particles having a preferred size between about 1 - 7 mm, preferably about 2 - 5 mm. The assay device uses flexible retaining means that may be deformed to admit a particle into the well, but that, once returned to undeformed shape, retain the particle in a captured position within a well.

WO 98/40726 (Trustees of Tufts College) describes optically interrogatable fiber optic sensors that are constructed by capture of chemically- derivatized microspheres individually from bulk suspension by arrayed icrowells. The microwells, formed in two-dimensional array by anisotropic etching of the end of an optical fiber bundle, capture and retain beads from bulk suspension by electrostatic retention; the wells are dimensioned just slightly larger in diameter than the beads themselves. In an alternative embodiment, microsphere swelling is used to entrap each microsphere in its corresponding microwell. The fiber tip is then dried.

None of the three aforementioned mechanical sorting and bead capture devices apparently provides for the concurrent dispensation of controlled volumes of fluid assay media. There thus still exists a need in the art for devices and methods that partition microparticles from bulk suspension and array them in spatially addressable fashion concurrently with dispensation of defined volumes of fluid medium.

SUMMARY OF THE INVENTION

The present invention solves these and other needs in the art by providing a simple device that automatically partitions microparticles from bulk suspension, arraying the particles in spatially addressable assay locations at determinate density in defined volumes of fluid medium.

The device is based upon the novel marriage of two physical principles: patterns of hydrophobic and hydrophilic areas compel fluid retention in segregated droplets of defined volume on the device surface; depressions in the fluid retention areas then capture beads suspended in the fluid droplets by electrostatic interaction. To array microparticles, the device is simply wetted and then dewetted with a suspension of microparticles: empty assay sites may be filled by repetition of the procedure; microparticles present in excess in any drop may be removed by washing. Wetting and dewetting may be accomplished simply by immersing the device in a fluid suspension of particles and then withdrawing the device from the solution. The device contains no moving parts and is capable in a single step of arraying microparticles at high density, thus proving particularly useful in automated screening of combinatorial libraries synthesized on polymer microbeads .

Thus, in a first aspect, the invention provides a microparticle arraying device, comprising a substrate having a plurality of fluidly non- communicating microparticle retention elements . Each of the microparticle retention elements has a fluid retaining surface with at least one depression; the depression is so dimensioned as to retain at least one of the microparticles via non-covalent interactions.

The area of the fluid retaining surface serves to define the volume of fluid retained. If the microparticles are suspended in a hydrophobic or nonpolar solvent, the fluid retaining surfaces are designed to be hydrophobic; the hydrophobic areas are rendered mutually noncommunicating by hydrophilic areas of the device substrate that intervene between the plural microparticle retention elements. If the microparticles are suspended in a hydrophilic solvent, the fluid retaining surfaces are hydrophilic; intervening hydrophobic areas of the device substrate then define a pattern of mutually noncommunicating hydrophilic fluid-retaining areas. The device substrate may be formed from any solid or semisolid material, such as glass, silicon, quartz, ceramic, metal, or plastic. In preferred embodiments, the substrate is glass, with the device conveniently constructed from a glass slide, or alternatively composed of plastic. In embodiments useful for high throughput screening assays, the device substrate, whatever its composition, is desirably substantially transparent to electromagnetic radiation. The intervening, pattern-defining areas of the device substrate may be integral to the device substrate or overlaid upon it. Thus, when the fluid retaining surfaces of the microparticle retaining elements are hydrophilic, the intervening hydrophobic areas may be integral to the device substrate or overlaid thereupon. In the latter embodiments, the hydrophobic layer may be selected from the group consisting of polytetrafluoroethylene, alkane thiol molecules, silane molecules, and straight-chain organic molecules. In preferred embodiments, the hydrophobic layer is polytetrafluoroethylene (Teflon^®) .

Analogously, the fluid retention surface may be contributed by the device substrate or applied thereto. Where the fluid retention surface is hydrophilic, it may be contributed by a glass surface of the device substrate.

In preferred embodiments, the fluid retention surface retains about 0.1 microliters to 20 microliters, preferably about 0.5 microliters to 5 microliters, most preferably about 1.5 microliters to 3.0 microliters.

The depression in each microparticle retaining element is so dimensioned as to retain at least one of the microparticles via non-covalent interactions: both the size of the depressions, relative to the microparticle size, and the number of depressions in each of the fluid retaining surfaces serve to define the number and nature of beads that will be retained. In embodiments useful for screening combinatorial libraries synthesized on polymeric microbeads, the microparticle capturing depression preferably has a maximal depth of at least about 0.01 microns to 1000 microns and, in its longest dimension coplanar with said hydrophilic fluid retention surface, extends about 0.01 microns to 1000 microns. Embodiments intended for screening combinatorial libraries preferably have depressions that are so- dimensioned as to retain no more than one of said microparticles .

Embodiments of the device easily integrable into current automated drug discovery systems pattern the fluid retaining surfaces into rectangular arrays, preferably rectangular arrays of 96, 384, 864, 1536, 3456, 6144, or 9600 elements.

In another aspect, the invention provides a method of making a microparticle arraying device, comprising the steps of: patterning a substrate to define at least two non-communicating fluid retaining surfaces; and fabricating at least one depression in each of said fluid retaining surfaces; wherein the depressions are so-dimensioned as to retain at least one microparticle via non-covalent intermolecular interactions.

The patterning may, in some embodiments, be effected by overlaying a hydrophilic surface of the substrate with a patterned hydrophobic layer, for example by a screen printing process, a spraying process, or an imprinting process. Alternatively, in other embodiments the patterning is effected by selectively modifying a hydrophilic surface of the substrate to render portions thereof hydrophobic, or, conversely, by selectively modifying a hydrophobic surface of said substrate to render portions thereof hydrophilic.

The depression may be fabricated before or after hydrophilic/hydrophobic patterning, so long as the depressions are so placed as ultimately to lie within the fluid retaining areas. The depressions may be fabricated by molding of the substrate, by a photolithographic process, by a laser machining or drilling process, or by a mechanical machining or drilling process.

Although particularly described as depressions that capture microparticles by electrostatic retention, various means for retaining microparticles in a predefined volume of fluid on the fluid retention surface are contemplated by the invention.

In another aspect, the invention provides methods of using the subject device. In one such aspect, the invention provides a method of arraying microparticles, comprising the steps of: wetting the microparticle arraying device with a fluid suspension of microparticles; and then dewetting said device. As noted above, the procedure may be repeated to fill all, or substantially all, microparticle retention elements. The microparticles may be suspended in substantially aqueous solution, and may be polymeric beads . Alternatively, the microparticles may be cells, including eukaryotic cells — such as mammalian cells or yeast cells — and prokaryotic cells. In yet another aspect, the invention provides a method of identifying chemical compounds with desired activity, comprising the steps of: arraying a plurality of microparticles in microparticle retention elements of the device, at least one of the microparticles bearing an identifiable chemical compound; and then detecting a desired activity of the compound on a target concurrently present in said retention element. In one embodiment, the chemical compounds, after arraying said microparticles and prior to detecting said activity, are detached from said microparticles. The target may be a cell, such as a eukaryotic or prokaryotic cell.

BRIEF DESCRIPTION OF THE DRAWINGS

These, and other aspects of the present invention, can best be understood by reference to the drawings, in which:

FIG. 1 is a top perspective view of a detached corner region of a microparticle arraying device of the present invention, also termed a "HydroPlate^™. " Shown are several mutually noncommunicating microparticle retention elements, each having a fluid retaining surface with a depression. One retention element is shown with a bead present in a retained fluid droplet;

FIG. 2 is a partial side sectional view of the HydroPlate. FIG. 2A shows the structure of a fresh, unused HydroPlate. FIG. 2B shows the HydroPlate after wetting and initial dewetting with a fluid suspension of microparticles. FIG. 2C shows microparticles uniformly arrayed in the retention elements after subsequent wash; FIG. 3 is a simplified top view of a HydroPlate showing microparticle retention elements in rectangular array;

FIG. 4 schematically illustrates use of the HydroPlate in a high-throughput drug screening protocol. FIG. 4A is a partial side sectional view of a HydroPlate arrayed with beads carrying a combinatorial chemical library, analyte cocktail, and having been placed into the analysis device. FIG. 4B shows the device reading background fluorescence from the HydroPlate. FIG. 4C shows the release of the test compound by ultraviolet illumination. FIG. 4D shows identification by the device of those microparticle retention elements containing a test compound interacting with the target. FIG. 4E shows the retrieval of a bead for further analysis of the test compound; and

FIG. 5 shows an alternative embodiment of the HydroPlate. FIG. 5A is a partial side sectional view of a HydroPlate designed to retain two microparticles of dissimilar size in each microparticle retaining element. FIG. 5B is a top view of this embodiment.

DETAILED DESCRIPTION OF THE INVENTION

In order that the invention herein described may be fully understood, the following detailed description is set forth. Definitions

Unless otherwise defined, all technical and scientific terms used herein have the meaning commonly understood by one of ordinary skill in the art to which this invention belongs. Definitions of such terms may be found, inter alia, in Czarnik et al . (eds.), A Practical Guide to Combinatorial Chemistry, American Chemical Soc'y (1997); and Chaiken et al . (eds.), Molecular Diversity and Combinatorial Chemistry: Libraries and Drug Discovery, American Chemical Soc'y (1996) .

As used particularly herein, the phrase "fluidly non-communicating" or its equivalents is understood to describe areas or structures that may only transiently, ephemerally, or under specific environmental conditions be in fluid noncommunication with one another. As further described below, the microparticle arraying device (HydroPlate) of the present invention will in typical use first be wetted with a suspension of microparticles and then dewetted, at which time the patterning of hydrophilic and hydrophobic areas will compel retention of fluidly- segregated droplets. During wetting, of course, the variously patterned areas of the device surface will be in fluid communication with one another. As used herein, the term fluidly non-communicating refers to the condition wherein, after wetting and dewetting, fluids retained are spatially segregated and in fluid noncommunication with one another. As used herein, the term "patterning" does not necessarily connote concurrent physical modification, but may also include the determination of physical locations that are subsequently to be modified. As further described below, the microparticle arraying device of the present invention may be fabricated in two steps. In a first approach, a substrate is first patterned to define at least two non-communicating fluid retaining surfaces; thereafter, at least one depression is fabricated in each of the defined fluid retaining surfaces. Alternatively, however, the device may be made by first fabricating depressions, then patterning fluid retaining surfaces thereabout. In the latter case, the organization of depressions predicts the form of the pattern yet to be imposed on the substrate.

Structure and Manufacture

The microparticle arraying device of the present invention, also termed a HydroPlate, is based upon the novel marriage of two physical principles: patterns of hydrophobic and hydrophilic areas compel fluid retention in segregated droplets of defined volume on the device surface; depressions in the fluid retention areas then capture beads suspended in the fluid droplets by electrostatic or other noncovalent or covalent interaction.

If the microparticles are suspended in a hydrophobic or nonpolar solvent, the fluid retaining surfaces are designed to be hydrophobic; the plural hydrophobic areas are rendered mutually noncommunicating by hydrophilic areas of the device substrate that intervene therebetween. If the microparticles instead are suspended in a hydrophilic solvent, the fluid retaining surfaces are designed to be hydrophilic; intervening hydrophobic areas of the device substrate then define a pattern of mutually noncommunicating hydrophilic fluid-retaining areas . The latter geometry — plural hydrophilic fluid retaining surfaces bounded by hydrophobic areas of the substrate — will be the more typical in devices designed for high throughput assays that use biologic targets, and will thus now be particularly described as exemplary.

FIG. 1 presents a top perspective view of a detached corner region of "HydroPlate" 10; a plurality of microparticle retention elements (MREs) 28 are shown in full or in part. In this embodiment intended for hydrophilic fluid retention, each MRE 28 has a hydrophilic fluid retaining surface 24. Each of the individual hydrophilic fluid retaining surfaces 24 is rendered mutually noncommunicating with that of other MREs 28 by intervening hydrophobic areas 22 of device substrate 20. That is, the disposition of fluid retaining surfaces 24 on device substrate 20 defines a pattern of hydrophilic areas rendered mutually noncommunicating by intervening hydrophobic areas of the device substrate. Each MRE has at least one bead capturing depression 26; the depression is so dimensioned as to retain at least one microparticle by noncovalent interactions, typically electrostatic interaction. In use, as further described below, the

HydroPlate is wetted with a fluid suspension of microparticles 30, then dewetted. After dewetting, each fluid retaining surface 24 retains a spatially segregated droplet 32 of fluid, which preferably contains one or more microparticles 30.

FIG. 2A presents a side sectional view of HydroPlate 10. In this particular embodiment, hydrophobic areas 22 are shown as elevated relative to hydrophilic fluid retaining surfaces 24; as further described below, this embodiment is fabricated by overlaying a patterned hydrophobic layer 22 upon hydrophilic device substrate 20; this overlay creates a pattern of hydrophilic fluid retaining surfaces 24 segregated and rendered mutually noncommunicating by intervening hydrophobic areas 22.

FIG. 2B shows the HydroPlate of FIG. 2A after wetting in a hydrophilic fluid suspension of microparticles and then dewetting. The fluid retaining surface 24 of each microparticle retention element 28 retains a droplet of fluid 32; the intervening hydrophobic areas 22 overlaid upon substrate 20 maintain the droplets in fluid noncommunication with one another. As shown, the suspension of microparticles is of sufficient density stochastically to ensure the presence in each droplet of a plurality of microparticles 30. Some microparticles are in free suspension, some are in contact with the fluid retaining surface 24, and some have fallen into depression 26.

After washing, as by rinsing in buffer, only microparticles 30 that have fallen into depression 26 are retained, thus causing a determinate number of microparticles to be retained per microparticle retention element 28. Wetting and dewetting the HydroPlate in fluid lacking microparticles thereafter places each retained microparticle 30 in a fluid droplet of defined volume (not shown) . As would be understood, for those embodiments intended for use with hydrophobic fluids, the disposition of hydrophilic areas and hydrophobic areas would be reversed, so that fluid retaining surface 24 of microparticle retention element 28 would be hydrophobic, and would be rendered noncommunicating with the fluid retaining surface of other MREs by interposition of hydrophilic surface 22 on device substrate 20. In this manner, microparticles suspended in a nonpolar fluid, such as hexane, a chlorofluorocarbon, or carbon tetrachloride can be arrayed in the MREs. In such embodiments, the substrate itself typically is hydrophobic (e.g., polystyrene) and a hydrophilic material is either overlaid to create the pattern necessary for spatially- segregated fluid retention, or created by surface derivatization, as by exposure, through patterned mask, to corona or plasma discharge. FIGS. 1 and 2 present the device as a rectangular plate. Although preferred for purposes of robotic integration, as further described below, this shape is not required; so long as the device presents a surface that may be patterned into hydrophilic and hydrophobic areas and depressions fabricated therein, the device may be made to function to array microparticles. Furthermore, although presented in FIGS. 1 and 2 with a single active surface external to the device, the device may have plural microparticle arraying surfaces, some or all of which may be internal. In those embodiments with internal particle arraying surfaces, the internal surface must of course be accessible to the fluid suspension of microparticles. One goal of these alternative configurations is to increase the active surface area of the device. In one such approach, the device incorporates pleats, in a form reminiscent of a fan- fold, or fins, such as those found in heatsinks or radiators .

In any chosen shape, the HydroPlate must be thick enough to permit partial-thickness depressions 26 of the appropriate depth and diameter, as further discussed below. At the same time, for some embodiments, the plate should not be so thick as to interfere with desired optical properties.

Device substrate 20 may be formed from any solid or semisolid material that may be patterned with hydrophilic and hydrophobic areas. The HydroPlate substrate will typically be a mechanically rigid, hydrophilic solid, such as glass, quartz, silicon, ceramic, metal, or plastic. Alternatively, however, the substrate may be composed of a solid non-rigid material, such as the elastomeric polymer polydimethylsiloxane . The fabrication method is chosen to be suitable to the substrate composition. Such methods include casting a molten or liquid material, such as glass, metal or plastic; forging; extruding; stamping; molding; or cutting from a billet.

In embodiments useful for high throughput screening assays, as further described below, the device substrate, whatever its composition, is desirably substantially transparent to one or more wavelengths of electromagnetic radiation, such as visible light, infrared (IR) , ultraviolet (UV) , or microwave radiation. In this manner, the HydroPlate may be transilluminated for the purposes of, for example, cleaving UV sensitive covalent bonds; stimulating fluorescent or other light responsive reporters, such as dyes, via a laser or other light sources; collecting fluorescent light emission from such stimulated chemical entities; or examining by light microscopy microparticles, such as mammalian cells, disposed on the substrate or the MREs.

Some applications of the HydroPlate may require that it be free from biotic contamination. Sterilization of the HydroPlate will permit its use in applications that require a sterile environment, such as some mammalian cell-based bioassays. In this circumstance, the substrate is chosen based, in part, on resistance to, and tolerance of, means for inactivating contaminating microorganisms, including viruses, bacteria and yeasts. Such means include dry heating, wet heating under pressure (autoclaving) , exposure to chemical agents, such as ethylene oxide, and exposure to high frequency electromagnetic radiation, including ultraviolet, x-ray or gamma ray radiation. For example, it is known that glass and polydimethylsiloxane can be autoclaved.

In preferred embodiments, the substrate is glass, with the device conveniently constructed from a glass slide.

The intervening, pattern-defining areas 22 of the device substrate 22 may be integral to the device substrate or overlaid upon it. Thus, when the fluid retaining surfaces of the microparticle retaining elements are hydrophilic, the intervening hydrophobic areas may be integral to the device substrate or overlaid thereupon. In the latter embodiments, the hydrophobic layer may be selected from the group consisting of polytetrafluoroethylene, alkane thiol molecules, silane molecules, and straight-chain organic molecules. In preferred embodiments, the hydrophobic layer is polytetrafluoroethylene (Teflon^®) . Analogously, the fluid retention surface 24 may be contributed by the device substrate or applied thereto. Where the fluid retention surface is hydrophilic, it may be contributed by a glass surface of the device substrate.

Alternatively, the substrate surface may contribute to both hydrophobic and hydrophilic areas, the pattern therebetween being created by surface derivatization. The area of fluid retaining surface 24 serves to define the volume of fluid retained. If the fluid retaining areas 24 are of a consistent size throughout the HydroPlate, then the volume of fluid retained will also be consistent; if the fluid retaining surfaces 24 differ in size on a single HydroPlate, commensurately different volumes of fluid will be retained by the microparticle retention elements 28 on the plate. This latter approach, including areas of disparate size, permits assays readily to be conducted at varying concentrations. In preferred embodiments, the fluid retaining surfaces 24 are essentially circular and so- dimensioned as to retain approximately 0.1 μl - 20.0 μl, more preferably 0.5 μl - 5.0 μl, and most preferably 1.5 μl - 3.0 μl . The depression 26 in each microparticle retaining element 28 is so dimensioned as to retain at least one of the microparticles via non-covalent interactions: both the size of the depressions, relative to the microparticle size, and the number of depressions in each fluid retaining surface 24 serve to define the number of beads that will be retained in each microparticle retention element 28. In embodiments preferred for use in screening combinatorial chemical libraries constructed on polymer microbeads, each depression is configured to hold a single spherical microparticle, or microbead, of predefined average diameter. Considered three-dimensionally, each depression typically approximates a portion of a sphere, an oblate spheroid, or prolate spheroid. Alternatively, the depression may be substantially cone-shaped, or cylindrical, with the bottom being flat. Regardless of the internal configuration, there must be sufficient contact area between the walls of the depression and a bead residing therein for electrostatic forces to retain the bead during subsequent manipulation and processing of the HydroPlate. When viewed from the side, as shown in cross section in FIG. 2A, the angle formed by the intersection of the wall of the depression with the surface of the plate will typically be 90 degrees or greater. The depressions partially penetrate the surface of the substrate and do not communicate with the opposite side. In preferred embodiments, the depression has a maximal depth of at least about 0.01 μm - 0.1 μm, more preferably 0.1 μm - 1 μm, typically 1 μm - 10 μm, but also may have maximal depths of 10 μm - 100 μm, and 100 μm - 1000 μm. In embodiments useful for screening combinatorial libraries synthesized on polymeric microbeads, the microparticle capturing depression preferably has a maximal depth of at least about 0.01 microns to 1000 microns .

The entry to the depression — that is, the portion coplanar with the surface of the substrate — typically is substantially circular. When viewed from above, as shown in FIG. 3, fluid retention surface 24 thus appears as an annulus surrounding depression 26, with the hydrophobic field surrounding the annulus. Although shown as centered within the fluid retaining surface, the depression may alternatively be disposed asymmetrically therein.

The entry to the depression has a diameter that is just slightly larger (approximately 1% - 30%, more preferably 5% - 15%, most preferably 5% - 10% larger) than the average diameter of the beads to be retained. In this way, minor variance in bead diameter in the population of beads to be arrayed may be accommodated. For spherical beads of typical average diameter of approximately 150 μm, for example, the depression entryway diameter would thus preferably be about 165 μm. In its longest dimension coplanar with the fluid retention surface, the depression is at least about 0.01 μm - 1000 μm, preferably 100 - 500 μm, most preferably 150 μm. Depressions may be fabricated in various ways, the choice dictated in part by the composition of the substrate, as would be understood in the art. In general, the depressions may be fabricated into the device substrate by molding, laser drilling, mechanical drilling, non-drilled machining, etching by a photolithographic process, or by other means of chemical etching, such as treating mask protected glass with hydrofluoric acid. Depressions can be formed first, followed by hydrophobic/hydrophilic patterning of the substrate surface, or vi ce versa .

Although particularly described as having a single depression, one or more of the microparticle retention elements 28 on the device substrate may have more than one depression. Such retention elements are capable of retaining a plurality of microparticles of determinate number.

If the plural depressions are identical in size and form, then multiple beads of the same size will be retained, maximally up to the total number of depressions per MRE. Plural depressions would permit multiple, yet determinate numbers of beads to be assayed at one time; this presents advantages in the initial screens of libraries against new targets.

Thereafter, beads plurally present in MREs producing positive signals are assayed in less complex admixture, ultimately to be assayed individually.

It is also possible to fabricate in a single microparticle retention element a plurality of depressions of differing volumes, depths, three- dimensional form, and entry dimensions and shapes, in this way selectively retaining in any single MRE microparticles of different sizes and shapes. For example, FIG. 5 presents, both in side cross-sectional view (FIG. 5A) and top plan view (FIG. 5B) , a HydroPlate in which each microparticle retention element 28 has two depressions 26 of disparate size, a larger depression and a smaller depression. This permits two different size beads to be arrayed on the same HydroPlate; to ensure uniform arraying, the larger beads would be arrayed first, the smaller beads subsequently. This embodiment permits the selective mixing of chemical compounds bound to the different sized beads, providing flexibility in assay design. In preferred embodiments of the invention, microparticles are retained within depression 26 solely by electrostatic interactions between the particle 30 and the walls of the depression formed in the device substrate 20. Other methods of retaining microparticles in retention elements are possible, however. For example, rather than relying exclusively on electrostatic interactions to retain the microparticle, beads doped with a paramagnetic or superpara agnetic material may be retained by magnetic areas disposed within each MRE. The size and/or shape of the magnetic area is adjusted as needed to prevent retention of multiple or indeterminate numbers of beads .

In another example, the bead is retained by frictional forces after swelling. In this method, the beads are arrayed into depressions in the HydroPlate in a "non-swelling" solvent, and are then exposed to a "swelling" solvent, increasing their volume. Given a sufficiently small difference in volume of the depression and that of the unswollen bead, the bead will expand to fill the depression, making firm contact with the walls and thereby retaining the bead in the depression. For example, "Tentagel" , a styrene- polyethylene glycol copolymer, is often used in combinatorial chemical synthesis. It is unswollen in nonpolar solvents such as hexane, and swells approximately 20%-40% in volume upon exposure to more polar solvents or aqueous media.

Alternatively, a suitable, fluid-compatible "adhesive" may wholly replace the depression without unduly diminishing the essential function of the HydroPlate. As used herein, the term "adhesive" intends substances that non-covalently bond two surfaces. Such adhesives could augment the function of the depression by surrounding the depression; surrounding and being applied to the internal walls of the depression; or being applied to the internal walls of the depression only.

For example, streptavidin could be applied to the fluid retention surface and beads could display biotin on their surface, or vice versa . The noncovalent interaction between streptavidin and biotin is extremely strong, and if a sufficient number of such interactions occur between surface and bead, then the bead could be stably retained on the surface. Antibodies and their cognate ligands (antigens) could be used for the same purpose of creating specific, complementary, non-covalent adhesions. If necessary, reactive chemical compounds could be used to create more stable, covalent linkages between the bead and the surface.

In other embodiments, a first means of retaining microparticles could be combined with a second, different, means to enable the selective retention of microparticles with different defined properties in each MRE. For example, instead of combining different sized depressions only, it is possible to combine one or more depressions with one or more alternative means for retention. As discussed above, the alternative means could be a specific adhesive system, wherein one adhesive component in the MRE partners with a complementary adhesive component on the microparticle. In another embodiment, depressions could be dispensed with entirely and replaced with multiple alternative means for retention within the elements .

Preferred embodiments of the HydroPlate are intended to be easily integrable into existing drug discovery efforts, and thus are preferably compatible in shape, size, and dimension with existing robotic devices .

The HydroPlate may be dimensioned, for example, similarly to a standard glass microscope slide, approximately 1.0 mm high, 25.0 mm wide and 75.0 mm long. However, it will especially advantageous to dimension the HydroPlate identically to a standard microtiter dish: approximately 16 mm high, 85 mm wide and 128 mm long. In addition to its gross dimensions, the

HydroPlate will preferably have microparticle retention elements disposed in patterns that are readily addressable by existing devices. Thus, although the microparticle retention elements may be disposed in any regular or irregular fluidly noncommunicating pattern on the device substrate, embodiments of the microparticle arraying device that are easily integrable into current automated drug discovery systems pattern the fluid retaining surfaces into rectangular arrays, preferably rectangular arrays of 96, 384, 864, 1536, 3456, 6144, or 9600 elements. These arrays will typically replicate the geometry of those commercially available for use in high-throughput assays, including 96 wells (8 by 12 wells) , 384 wells (16 by 36 wells), and 1536 wells (32 by 48 wells).

Presently, so called "nanowell" formats are being developed that will have 3456 wells (48 by 72 wells), 6144 wells and 9600 wells. For continued compatibility with the original 96-well format, increasing well density typically follows the geometric series defined by y = n ² * 96, where "y" is the total number of wells on a plate and ^vn" is an integer. As total well number increases, the well volumes become proportionally smaller, thereby contributing to savings of valuable reagents, overall cost, space for equipment, and time. As further geometries become standard, the HydroPlate may advantageously be identically so dimensioned. As suggested by the above, approaches to manufacture of the HydroPlate will be as protean as the device's physical embodiments, and the desired form of the device will in large measure dictate the desired manufacturing approach. Present techniques in materials science, well known to the skilled artisan, will suffice.

One preferred approach to patterning the device surface, as described further in Example 1, below, is to overlay a hydrophilic glass substrate with a patterned hydrophobic layer, such as polytetrafluoroethylene (PTFE or Teflon^®) , by screen printing. Alternatively, the entire surface of the substrate may be rendered hydrophobic, as for example by dipping the substrate into a liquid polymer, curing the polymer, and thereafter removing from the applied layer defined areas or patches by, for example, laser ablation or chemical treatment, to reveal the hydrophilic surface beneath, thus creating the hydrophilic-hydrophobic pattern. To create the hydrophobic layer in other embodiments, alkane thiol molecules could be imprinted on metals, especially gold, silver and copper; silane molecules could be imprinted on glass, silicon dioxide, and other oxide surfaces; and straight-chain organic molecules could be imprinted on a hydrophilic substrate, wherein there is a functional group on one end that binds to a hydrophilic surface, and another functional group on the other end of the molecule that imparts hydrophobic characteristics. Advantages

Because the HydroPlate relies on physical principles — such as surface tension and electrostatic attraction — for its operation, the HydroPlate offers three significant advantages.

The first, of course, is simplicity and reliability. Because the HydroPlate is arrayed with beads and filled with fluid by means of a passive, fully automatic mechanism, the baroque fluidics of present-day devices are obviated. And because the HydroPlate itself has no moving parts, chance of failure is low. As the need for increased throughput increases, so too will this advantage.

Second, the HydroPlate both arrays microbeads and dispenses fluid in a single step, in massively parallel fashion. This markedly diminishes the time necessary to accomplish these steps.

Third, because the distribution of beads may be decoupled from the dispensation of fluids, valuable reagents may be used at higher concentration in smaller volumes, thereby reducing waste and saving costs. That is, the HydroPlate may be dried after arraying of beads and then rewet with an assay buffer without dilution thereof. In addition to high-throughput drug screening, it is expected that applications that have heretofore relied on the fluid and microparticle sorting abilities of fluorescence activated cell sorting (FACS^®) technology can be adapted to use the HydroPlate. Use and Function of the HydroPlate

To realize these advantages, it is necessary first to wet an "active" surface of the HydroPlate — that is, to wet one or more surfaces of the device upon which are disposed microparticle retention elements in fluid noncommunication.

To array microparticles, the particles are first suspended in a fluid; the HydroPlate is then wetted with this suspension. Microparticles are suspended in fluids chosen for chemical compatibility with the particle composition and chosen to have a specific gravity equal to or lower than that of the particle. The latter requirement prevents particles from floating to the surface of the fluid. The suspension will ideally be monodisperse prior to wetting the HydroPlate. Aggregation of particles may be prevented by vigorous agitation, for example by sonicating the beads in dimethylformamide (DMF) , then in methanol, followed by washing the beads in water to remove residual organic solvents. Even dispersion of the disaggregated particles may be maintained by increasing the viscosity of the fluid, as by adding viscosity-enhancing agents such as solubilized agar, methylcellulose, or glycerol. Alternatively, the particle suspension may simply be mixed, for example using a magnetic stir bar.

Although only part of the surface may be wetted, it is expected that more typically the entirety of the device surface will be wetted in each step. Uniform and complete wetting can be accomplished by various means. For example, the fluid may be poured over the HydroPlate or, alternatively, dispensed from a wide bore pipet . Preferably, however, the HydroPlate is simply dipped into a fluid reservoir. In one such approach, the HydroPlate is immersed, edge first, by penetrating the surface of the fluid until the device is completely submerged, and then dewetted by slowly removing the HydroPlate, allowing gravity to drain the bulk fluid back into the reservoir. This method requires that there be sufficient volume of fluid in the reservoir that the HydroPlate can be completely submerged. Such complete immersion is preferred if the HydroPlate has more than one active surface. To minimize adherence of fluid to irrelevant surfaces, the irrelevant surfaces may be rendered fluid-repelling: for hydrophilic fluids, irrelevant surfaces of the device may be rendered hydrophobic; for hydrophobic fluids, irrelevant surfaces may be rendered hydrophilic.

Complete immersion may be disfavored for arraying combinatorial libraries, however. Combinatorial chemical libraries built on polymeric beads are commonly regarded as expensive to produce, and are carefully husbanded. Thus, rather than immersing the HydroPlate in its entirety into a suspension of such beads, it may be preferable to contact only the active surface of the device to the reservoir. In this way, the fluid does not make contact with irrelevant surfaces of the HydroPlate, minimizing wastage.

Following wetting, the surface must be dewetted of the bulk fluid to permit the formation of discrete, constant volume droplets over the fluid retention surfaces of the MREs. As is necessary for any particular application, the HydroPlate may be wetted and dewetted iteratively. To dewet, the HydroPlate may be angled to enable the excess bulk fluid to drain off. Alternatively, a pipet may be used to aspirate the excess fluid. For embodiments in which the HydroPlate fluid retention surfaces are hydrophobic, and the surrounding pattern is hydrophilic, it may be necessary to complete the dewetting step within a hydrophilic fluid, such as water, rather than in air. For example, if the HydroPlate is wetted with the hydrophobic fluid styrene and then dewetted by removal to air, most of the fluid will be disposed, as desired, over the hydrophobic fluid retention surfaces. However, there may be a slight tendency for some of the styrene to flow over the boundary between the hydrophobic and hydrophilic areas. To prevent this from occurring, the HydroPlate may be removed from styrene to water. The water facilitates the spatial segregation of hydrophobic droplets . After wetting the HydroPlate with a bead suspension and dewetting, droplets retained over the MREs of a HydroPlate are typically allowed to evaporate. Evaporation may be accomplished passively, or may be facilitated by heating or by placing the HydroPlate in a desiccating environment, or both. As the fluid dries, beads contained therein contact and come to rest upon the fluid retaining surface of the microparticle retention element, including the depression therein. Inside the depression, the large surface area of contact between the bead and the walls of the depression increases electrostatic forces sufficiently to hold the bead in the depression. In contrast, beads disposed elsewhere on the fluid retention surfaces of the MRE do not make as extensive contacts and will be held weakly by electrostatic forces. The disparity in the strength of the attraction permits beads not retained within a depression to be removed by washing without dislodging the retained bead.

The HydroPlate is preferably washed after beads are arrayed and before any analyte cocktail is applied. Washing removes from microparticle retention elements those beads that are not captured in depressions or in their functional equivalents, ensuring that the microparticle retention elements retain a determinate number of particles. For devices with hydrophilic fluid retention surfaces, such wash solution may be water, salt-containing buffers, or the like. For devices with hydrophobic fluid retention elements, the wash buffer may be a suitable nonpolar solvent .

Washing is accomplished by wetting, as above, using a wash solution. For hydrophilic fluid retaining surfaces, the wash may be water or other hydrophilic solution. For devices with hydrophobic fluid retaining surfaces, the wash solution is a suitable nonpolar solvent. If necessary, the plate can be washed in consecutive reservoirs of washing fluid to reduce the chance that any adventitious, unretained beads will remain on the plate. Alternatively, washing can be effected by a stream of wash fluid.

Beads removed by washing will preferably be recycled. The beads are captured and concentrated, either by centrifuging or filtering, or if paramagnetic by use of magnets. Such recycled beads can be used later to array other HydroPlates .

If after performing the steps for arraying beads into a HydroPlate, some of the MREs remain empty, wetting, dewetting, evaporation and washing may be repeated as necessary.

After particles have been arrayed, the HydroPlate may be dried for storage, or wetted to dispense fluids for chosen applications. These assay fluids will often be cocktails containing various analytes, reporter substrates, targets, and other components necessary for assay.

Typically the analyte cocktail will be an aqueous solution containing dissolved solids, such as salts, pH buffers and other components. The analyte cocktail will also contain the so-called "target", whose physico-chemical interaction with the test compound is to be determined. According to the design of the assay, a positive interaction is evidence that the test compound may have some useful biological activity that would make it suitable to serve as a therapeutic agent. For bioassays, the target could be a known, characterized protein, nucleic acid, carbohydrate or any other macromolecule of biological interest, solubilized in the fluid of the analyte cocktail. The target could also be incorporated into some form of suspended carrier, such as a liposome, if the target requires a lipid environment for its correct structure and function.

The target could also be expressed by a cell, either eukaryotic or prokaryotic. Among eukaryotic cell types usefully employed are: mammalian cells (particularly including human cells and rodent cells, the latter including mouse, rat, or hamster cells) , yeast cells (particularly including Saccharomyces cerevisiae and Schi zosaccharomyces po be cells) , roundworm cells, such as those from Caenorhabdi tis el egans , plant cells, including Arabidopsis thaliana, and insect cells, including those from Drosophila melanogaster. Where the target is expressed by or present in a cell, the cells would be suspended in an analyte fluid capable of supporting their viability. Such targets would be naturally occurring within the cell, or the cell could be genetically engineered to express atypical (e.g., supranormal) quantities of an endogenous target, or alternatively, to express a target not normally expressed by that cell type.

During assay, it may be desired to prevent evaporation of the small volumes of fluid disposed on the surface; thus, it may be desired to store the HydroPlate in a high humidity environment. This will particularly be desired when volume changes would sufficiently alter solute concentrations as adversely to affect the performance of an assay, or if the HydroPlate is arrayed with cells that must be kept viable . As illustrated in Example 3, below, the

HydroPlate proves particularly useful in drug screening for the pharmaceutical industry. In this case, beads are polymers, such as polystyrene, wherein each bead carries a unique test compound that is a member of a combinatorial chemical library, synthesized using the split-pool technique. The analyte fluid contains a target whose interaction with the test compounds is to be determined. The assay is designed such that an interaction between the test compound and the target elicits a signal that can be detected. Such assays are well known in the art.

For some assays it is possible for an interaction between the target and the test compound to occur while the compound is bound to the bead. However, it will usually be desirable to study iterations in the liquid phase. This requires that the compound be released from the polymer support. A common strategy is to cleave a photolysable linker that covalently links the compound to the support, permitting the compound to diffuse into the ambient fluid. Photolysis is effected by transilluminating a HydroPlate from below, if the HydroPlate is fashioned of light-transmissible material, or by illuminating the HydroPlate incidentally from above.

For assays where the targets are components of viable cells, interaction between the target and a test compound can result in a detectable change in cellular physiology. For example, changes. in intracellular calcium ion concentration can be measured by loading cells with a calcium sensitive fluorescent reporter dye.

The HydroPlate is also readily adapted to competition assays, where the interaction between the target and a known agonist can be altered by the presence of a test compound. Applications include identifying antagonists of receptors and enzymes.

Such competition assays may advantageously be performed using a HydroPlate design in which each MRE has both a large and small depression, permitting beads of two sizes to be arrayed. The larger beads result from split-mix combinatorial chemical synthesis, each displaying a unique test compound. The smaller beads all carry an identical compound, known to bind specifically to the target, conjugated to a fluorophore moiety. The photolysable linkers are different for the two bead populations. The large beads are arrayed first, followed by the smaller beads. Subsequently, the analyte fluid containing the target is dispensed by wetting and dewetting. The test compound is released by photolysis first, and allowed to equilibrate with the target. Then, the tagged agonist is released by light of a different wavelength. Binding in the presence of the test compounds is determined using fluorescence polarization (FP) , fluorescence resonance energy transfer (FRET) or another method, as further described below. If extent of binding is polled at different time points, the kinetics of binding can be determined.

Due to the strength of electrostatic attraction retaining beads, the HydroPlate can be subjected to multiple rounds of wetting and dewetting. Thus, by modulating the degree of linker photolysis to release a fractional quantity of the test compound bound to beads, it is possible to study the interaction of test compounds with multiple targets contained in a different analyte cocktails. In many assays currently in use, the signal is fluorescent light emission, stimulated by electromagnetic, or nuclear decay radiation, or another energy source. Methodologies for fluorescent signal generation and detection include lanthanide time- resolved fluorescence (LnTRF) , fluorescence resonance energy transfer (FRET) , fluorescence polarization (FP) , homogeneous time-resolved fluorescence (HTRF) , and chemiluminescence (CL) .

After sufficient time has elapsed to permit an interaction between the test compound and the target, fluorescence is stimulated by light of an appropriate wavelength. Fluorescent signal emission is monitored using an electro-optical detection system over a time course of microseconds to hours. With reference to appropriate controls, computer software programs analyze and report the location of MREs where a fluorescence signal exceeding the background level occurs. After washing the plate to remove the analyte fluid, the beads contained in all MREs where positive fluorescence signals were detected are retrieved and the identity of the test compounds determined. Identification of these compounds (deconvolution) can proceed according to typical analytical chemistry techniques including magnetic resonance spectroscopy, infrared spectroscopy, mass spectroscopy, and high performance liquid chromatography . Additionally, or alternatively, compound identification can proceed by reference to chemical or other types of tags attached to the bead during synthesis of the combinatorial chemical library, which uniquely identify the chemical synthetic steps that built the test compounds of interest .

Many alternative applications exist that advantageously use the HydroPlate.

To sequence DNA, the DNA strand to be analyzed is the target and oligo-deoxyribonucleotides of predefined length are synthesized by the split-pool technique on beads. Interaction between oligos on the beads and the target is detected, followed by identification of the oligo sequences that specifically bound to the target strand. Sequence is reconstructed from such hybridization reactions by known techniques, such as those set forth in U.S. Patent Nos . 5,492,806; 5,525,464; 5,695,940; 5,202,231; 5,695,934; 5,795,716; the disclosures of which are incorporated herein by reference .

Compounds with catalytic activity can be discovered using the HydroPlate in a high-throughput format. For this application, potential organic, inorganic, and organo-metallic catalysts are built by the split-pool technique on beads; the catalytic substrate is the target. Catalytic interaction between the test compound and the substrate is detected by monitoring heat production, viscosity changes, color changes, or by other means.

Other applications include monitoring the kinetics of chemical reactions, medical diagnostics and materials science applications. Additionally, the spatially addressable format of the HydroPlate, combined with the flexibility of configuring each microparticle retention element to retain multiple beads offers a simplified means for carrying out complex and/or parallel chemical syntheses.

The following examples are offered by way of illustration and not by way of limitation.

EXAMPLES

Example 1. Manufacture of a HydroPlate.

A HydroPlate with a 3 x 10 rectangular array of microparticle retention elements with hydrophilic fluid retaining surfaces was manufactured as follows.

A thin layer of liquid Teflon^® was screen printed (Erie Scientific, Inc., Portsmouth, NH) on a standard glass microscope slide of dimensions 1.0 mm x 25.0 mm x 75.0 mm and allowed to dry. The hydrophobic Teflon^® was patterned to define a 3 x 10 rectangular array of circular voids that reveal the underlying hydrophilic surface of the glass substrate. Each such fluid retention area was 2.0 mm in diameter; the fluid retention areas were spaced approximately 4.0 mm apart, measured from center to center.

A single depression was placed into the center of each hydrophilic area by laser micromachining (Potomac Photonics, Inc., Lanha , MD) . Depressions were 150 μm in depth and 200 μm in width.

Example 2. Use of the HydroPlate to Array Beads in Agueous Suspension.

The HydroPlate manufactured in Example 1 was used to array beads.

ArgoPore^® beads (Argonaut Technology, Inc., San Francisco, CA) 142 ± 10 μm in diameter were dispersed in a solution of IX phosphate buffered saline (PBS) at a concentration of 10 mg beads/ml. The HydroPlate was dipped into the bead suspension and then removed from the bead suspension. Inspection demonstrated capture of beads by the micromachined depressions .

Example 3. Use of a HydroPlate for High-Throughput Drug Screening of a Protein Target

FIG. 4 is a schematic broadly illustrating the use of the HydroPlate for high-throughput screening of a synthetic chemical library to select compounds that have the potential to serve as human or animal therapeutic agents. Although not depicted in this figure, many of the steps herein described are performed automatically by a robot. In this way human error due to fatigue or inattention is minimized and speed of processing is maximized.

In the first step depicted in FIG. 4A, a HydroPlate 100 is arrayed with beads 30 containing a combinatorial chemical library. The HydroPlate is viewed in cross section and is not shown Lu scale.

The HydroPlate has length and width dimensions essentially identical to that of a typical 96-well microtiter plate, and thickness of 1.5 mm, similar to a laboratory microscope slide. The substrate is composed of glass and has disposed on its top surface 1536 circular microparticle retention elements in a 32 by 48 rectangular array. The MREs are formed by the combination of the hydrophilic glass substrate, one centrally located depression per MRE, so-dimensioned as to retain a single 150 μm diameter spherical bead, and a Teflon^® pattern disposed on the top surface of the substrate by a screen printing process that renders MREs fluidly non-communicating. The volume of fluid retained by each fluid retaining surface is approximately 1.5 μl .

The beads, each of which is 150 μm in diameter, are composed of polystyrene. The combinatorial library is synthesized by a "split-pool" protocol, such that each bead carries a single species of compound to be tested in the assay. There is a linker between the polystyrene matrix and the synthetic chemical attached to it which is photolysable by ultraviolet radiation. The library is synthesized with tags that permit subsequent identification of the synthesized compound (deconvolution) .

To array the HydroPlate, beads are disaggregated and evenly suspended in water by mixing with a stirbar. The average concentration of beads is 5.0 beads per 1.5 μl of water. The top surface of the HydroPlate is wetted by immersing it in the bead suspension and dewetted by its removal. The fluid retained over each MRE is then allowed to evaporate to permit a single bead to fall into each depression and be retained there through electrostatic interactions. The HydroPlate is then washed in water to remove beads not retained within depressions. On average 4 non- retained beads occur within each MRE. Such beads are recovered from the wash solution by centrifugation. There is sufficient diversity in the combinatorial library that there is nearly a 100% probability that all of the 1536 beads retained by the HydroPlate carry a unique chemical to be tested. After washing, the HydroPlate is dried in advance of the next step. In the next step, 1.5 μl of the analyte cocktail 38 is dispensed to all of the MREs on the HydroPlate. As for the arraying step, dispensing is accomplished by wetting and dewetting the top surface of the HydroPlate by immersion into and subsequent removal from the analyte cocktail.

The analyte cocktail contains water, salts, pH buffers and solubilized target molecules whose physico-chemical interaction with the 1536 compounds bound to the arrayed beads is to be tested.

The target is a substantially purified protein which is a receptor expressed in human cells, known to interact with a class of medically useful drugs. The target is chemically conjugated to a fluorophore moiety. When the target binds to a ligand, the structure of the target changes, altering the microenvironment of the conjugated fluorophore. This alteration affects the amplitude of the emission spectrum of the fluorophore which is detected and analyzed. A change in amplitude after a target is mixed with a putative ligand is indicative that the ligand binds to and changes the structure of the target . Next, the HydroPlate 100, arrayed with beads and the analyte cocktail containing the target is loaded into a light-tight, temperature controlled, automated machine 120 capable of the simultaneous analysis of all 1536 MREs, a task which is facilitated by the spatially addressable format of the HydroPlate. The device is pressurized with an inert, humidified atmosphere to prevent evaporation of the analyte cocktail. Within the device, two light sources are situated above the HydroPlate and illuminate it with light of uniform intensity. One, a monochromatic light source 60 emits light 62 at a frequency that excites the fluorophore; the second is an ultraviolet (UV) light source 64 that emits UV light 66 at a frequency that cleaves the photolysable linker. An array 68 of high sensitivity charge coupled device (CCD) elements, organized in the same array pattern as the MREs, is situated below the HydroPlate, such that each element collects light from a single MRE.

Steps in the analysis are depicted in FIG. 4B, 4C, and 4D. In FIG. 4B, the monochromatic source 60 briefly illuminates 62 the droplets 38, thereby exciting the fluorophore conjugated to the target. The CCD array 68 then detects and counts fluorescent photons 70 emitted by the fluorophore in the droplet 38 over each MRE. The fluorescence intensity from each MRE is stored in a data analysis computer 72 in association with the location within the grid pattern of the MRE. Next, in FIG. 4C, the UV source 64 transilluminates 66 the beads, cleaving a substantial proportion, but not 100% of the linker between the test compound and the polymeric solid support matrix, such that the compound 44 diffuses out of the bead into the ambient analyte cocktail fluid 38. The HydroPlate 100 is then incubated until any released test compounds 44 capable of binding to them are in equilibrium with the target proteins. As depicted in FIG. 4D, after equilibrium is reached, the monochromatic source 60 briefly illuminates 62 the droplets 38 and fluorescent emission 70 is detected and counted by the array of CCD elements 68. This data is also stored in the computer memory 72. As mentioned above, when a test compound binds to the target protein it changes the protein's structure, altering the microenvironment of the conjugated fluorophore, which affects its fluorescence emission intensity. This change is detected using a computer software program that compares the fluorescence emission from each MRE before and after test compound release. When the difference for any particular MRE exceeds the background level determined using controls, its location is flagged in the computer and output 74 in a form readable by a human operator or robot .

The final stage of analysis is depicted in FIG. 4E. The HydroPlate 100 is removed from the analysis device and placed under a stereoscopic microscope 76 outfitted with an electronically controllable stage 78 and micromanipulator apparatus holding a sample needle 80. A human operator reads the output from the computer to determine which MREs possess a test compound binding to the target. In turn, the stage is moved to bring each of those MREs into view under the microscope. The needle is then manipulated to penetrate the bead captured in the depression and remove it from the MRE. The bead is washed to remove traces of the analyte cocktail and is transferred to a vial containing a small volume of carrier fluid. UV light is then used to transilluminate the bead to cleave the remaining test compound from the solid support such that it diffuses into the carrier fluid. The identity of the compound is then determined by deconvolution, based upon information about the sequential synthesis steps embodied in the tag attached to the microbead.

Alternatively or additionally, a sample of the compound in the carrier is then taken for analysis by mass spectroscopy and high performance liquid chromatography to determine its structure. Knowing the structure, a synthetic chemist produces a larger quantity of each of the compounds that binds to the target, and these compounds are further tested for their suitability to serve as therapeutic agents in animals and humans.

All patents, patent publications, and other published references mentioned herein are hereby incorporated by reference in their entirety as if each had been individually and specifically incorporated by reference herein. While preferred illustrative embodiments of the present invention are described, it will be apparent to one skilled in the art that various changes and modifications may be made therein without departing from the invention, and it is intended in the appended claims to cover all such changes and modifications that fall within the true spirit and scope of the invention.

Claims

What is claimed is:

1. A microparticle arraying device, comprising: a substrate having a plurality of fluidly noncommunicating microparticle retention elements, each of said elements having a fluid retaining surface with at least one depression, said depression so dimensioned as to retain at least one of said microparticles via noncovalent interactions.

2. The microparticle arraying device of claim 1, wherein each of said fluid retaining surfaces is hydrophobic, and the disposition of said fluid retaining surfaces on said device substrate defines a pattern of hydrophobic areas rendered mutually noncommunicating by intervening hydrophilic areas of said device substrate.

3. The microparticle arraying device of claim 1, wherein each of said fluid retaining surfaces is hydrophilic, and the disposition of said fluid retaining surfaces on said device substrate defines a pattern of hydrophilic areas rendered mutually noncommunicating by intervening hydrophobic areas of said device substrate.

4. The microparticle arraying device of claim 3, wherein said substrate is selected from the group consisting of: glass, silicon, quartz, ceramic, metal, and plastic.

5. The microparticle arraying device of claim 3, wherein said substrate is glass.

6. The microparticle arraying device of claim 5, wherein said glass is selected from the group consisting of: alkali-borosilicate glass, alumina- silicate glass, barium flint glass, barium-borate glass, borosilicate glass, chalcogenide glass, fused silica glass, lanthanum glass, optical glass, phosphate glass, and soda-lime glass.

7. The microparticle arraying device of claim 6, wherein said substrate is borosilicate glass.

8. The microparticle arraying device of claim 3, wherein said substrate is substantially transparent to electromagnetic radiation.

9. The microparticle arraying device of claim 4, wherein said substrate is plastic.

10. The microparticle arraying device of claim 9, wherein said plastic is selected from the group consisting of polyethylene, polypropylene, polyacrylate, polymethylmethacrylate, polyvinylchloride, polytetrafluoroethylene, polystyrene, polycarbonate, polyacetal, polysulfone, celluloseacetate, polydimethylsiloxane, and cellulosenitrate .

11. The microparticle arraying device of claim 3, wherein said hydrophobic areas of said device substrate are integral to said substrate.

12. The microparticle arraying device of claim 3, wherein said hydrophobic areas are overlaid upon said substrate .

13. The microparticle arraying device of claim

12, wherein said hydrophobic layer is selected from the group consisting of: polytetrafluoroethylene, alkane thiol molecules, silane molecules, and straight-chain organic molecules.

14. The microparticle arraying device of claim

13, wherein said hydrophobic layer is polytetrafluoroethylene .

15. The microparticle arraying device of claim 3, wherein said hydrophilic fluid retention surface is contributed by said device substrate.

16. The microparticle arraying device of claim 15, wherein said hydrophilic fluid retention surface is glass .

17. The microparticle arraying device of claim 3, wherein said depression has a maximal depth of at least about 0.01 microns to 0.1 microns.

18. The microparticle arraying device of claim

17, wherein said depression has a maximal depth of at least about 0.1 microns to 1 microns.

19. The microparticle arraying device of claim

18, wherein said depression has a maximal depth of at least about 1 microns to 10 microns.

20. The microparticle arraying device of claim

19, wherein said depression has a maximal depth of at least about 10 microns to 100 microns.

1. The microparticle arraying device of claim 20, wherein said depression has a maximal depth of at least about 100 microns to 1000 microns.

22. The microparticle arraying device of claim 3, wherein said depression, in its longest dimension coplanar with said hydrophilic fluid retention surface, is at least about 0.01 microns to 1000 microns.

23. The microparticle arraying device of claim 3, wherein said depression, in its longest dimension coplanar with said hydrophilic fluid retention surface is at least about 100 microns to 500 microns.

24. The microparticle arraying device of claim 3, wherein said depression, in its longest dimension coplanar with said hydrophilic fluid retention surface is at least about 150 microns.

25. The microparticle arraying device of claim 3, wherein said hydrophilic fluid retention surface retains about 0.1 microliters to 20 microliters.

26. The microparticle arraying device of claim 3, wherein said hydrophilic fluid retention surface retains about 0.5 microliters to 5 microliters.

27. The microparticle arraying device of claim 3, wherein said hydrophilic fluid retention surface retains about 1.5 microliters to 3.0 microliters.

28. The microparticle arraying device of claim 3, wherein the pattern of said hydrophilic fluid retention surfaces is a rectangular array.

29. The microparticle arraying device of claim 28, wherein said rectangular array consists of 96, 384, 864, 1536, 3456, 6144, or 9600 of said hydrophilic fluid retention surfaces.

30. The microparticle arraying device of any one of claims 1, 2, or 3 wherein said depressions are so- dimensioned as to retain no more than one of said microparticles .

31. A method of making a microparticle arraying device, comprising the steps of: patterning a substrate to define at least two fluidly non-communicating fluid retaining surfaces; and fabricating at least one depression in each of said fluid retaining surfaces; wherein said depressions are so-dimensioned as to retain at least one of said microparticles via noncovalent intermolecular interactions.

32. The method of making a microparticle arraying device of claim 31, wherein said patterning is effected by overlaying a hydrophilic surface of said substrate with a patterned hydrophobic layer.

33. The method of making a microparticle arraying device of claim 32, wherein said overlaying is effected by a screen printing process.

34. The method of making a microparticle arraying device of claim 32, wherein said overlaying is effected by a spraying process.

35. The method of making a microparticle arraying device of claim 32, wherein said overlaying is effected by an imprinting process.

36. The method of making a microparticle arraying device of claim 31, wherein said patterning is effected by selectively modifying a hydrophilic surface of said substrate to render portions thereof hydrophobic, or by selectively modifying a hydrophobic surface of said substrate to render portions thereof hydrophilic.

37. The method of claim 31 wherein said depression is fabricated by a photolithographic process .

38. The method of claim 31 wherein said depression is fabricated by a laser machining or drilling process.

39. The method of claim 31 wherein said depression is fabricated by a mechanical machining or drilling process.

40. A microparticle arraying device, comprising: a substrate having a plurality of fluidly noncommunicating microparticle retention elements, each of said elements having means for retaining at least one of said microparticles in a predefined volume of fluid.

41. A microparticle arraying device, comprising: a substrate having a plurality of fluidly noncommunicating microparticle retention elements, each of said elements having a fluid retaining surface and means for retaining at least one of said microparticles on said surface.

42. A method of arraying microparticles, comprising the steps of: wetting the microparticle arraying device of claim 1 with a fluid suspension of microparticles; and then dewetting said device.

43. The method of claim 42, further comprising the step, after said dewetting step, of: drying said device .

44. The method of claim 43, further comprising a subsequent step of washing said device.

45. The method of claim 42 wherein said fluid is substantially aqueous.

46. The method of claim 42 wherein said microparticles are polymeric beads.

47. The method of claim 42 wherein said microparticles are cells.

48. The method of claim 47 wherein said cells are eukaryotic cells.

49. The method of claim 48 wherein said eukaryotic cells are mammalian cells.

50. The method of claim 48 wherein said eukaryotic cells are yeast cells.

51. The method of claim 47 wherein said cells are prokaryotic cells.

52. A method of using the microparticle arraying device of claim 1, comprising: arraying microparticles in a plurality of said microparticle retention elements; and then identifying arrayed microparticles with desired properties.

53. A method of identifying chemical compounds with desired activity, comprising the steps of: arraying a plurality of microparticles in microparticle retention elements of the device of claim 3, at least one of said microparticles bearing an identifiable chemical compound; and then detecting a desired activity of said compound on a target concurrently present in said retention element.

54. The method of 53, wherein said chemical compounds, after arraying said microparticles and prior to detecting said activity, are detached from said microparticles .

55. The method of claim 53, wherein said target is a cell.

56. The method of claim 55, wherein said cell is a eukaryotic cell.

57. The method of claim 56, wherein said cell is a mammalian cell.

58*. The method of claim 57, wherein said cell is a human cell.

59. The method of claim 57, wherein said cell is a yeast cell.

6₀. The method of claim 55, wherein said cell is a prokaryotic cell.